Drop Test
Drop Test vs Cheddar (L)
Drop Test vs Peacemaker (W)
Drop Test vs Wedge 1 Wedge 2 (W)
Drop Test vs Flex (W)
Drop Test vs Thagomizer (L)
Drop Test vs Sidekick (W)
Re-Roll
Re-Roll vs Blue Lion (W)
Re-Roll vs Duct Spartan (W)
|
Drop TestDrop Test vs Cheddar (L)First fight of the tournament for Drop Test. Cheddar was a powerful vertical spinner. From the Brazilian team that showed up. Driven very well and aggressively. I missed an opportunity early to get a good flip in, but it didn't matter. The first major impact that occurred, broke the flipper trigger mechanism and landed me upside down, for a quick knockout. Drop Test vs Peacemaker (W)I really wish I had a video of this fight. One of the best performances I've had with Drop Test. Peacemaker was a Fingertech Viper vertical spinner kit. Drop Test got under them easily and got in a bunch of flips over the course of the fight. At one point I picked up the other bot while resetting the flipper mechanism and showing off for the judges. Easy win by judges decision Drop Test vs Wedge 1 Wedge 2 (W)I was a little nervous for this fight as Hercules had faced Wedge 1 Wedge 2 a year or two ago in LA. They're not damaging, but with good drivers, they're pesky enough to control a fight and win by a decision. Drop Test had a little more "damage" than Hercules does and was able to control the fight a bit better. Close win by judges decision. Drop Test vs Flex (W)Flex is one of m favorite robots of all time. It's builder brought it out of retirement for Robogames this year. The design is a clamping lifter that I at one point tried to coy with Hercules. Very creative design, much like Drop Test that allows for lots of control, and a non damaging fight. Lots of fun, and a split judges decision for Drop Test. Drop Test vs Thagomizer (L)I was dreading this matchup. There was virtually no chance that I'd win this if Thagomizer was working, and he'd had a full day to prepare from his loss. Drop Test had no real side armour, so I had to try and react to his classic spin move in time to block it, for the full match. Obviously this plan falls apart as you start losing wheels. Loss by KO for Drop Test, to knock it out of the tournament. Drop Test vs Sidekick (W)This was a fun grudge match against one of m co-workers who also attended the event. Just for fun, but I still feel like I won. His crab drive antweight was still really impressive and surprisingly difficult to control against. Re-RollRe-Roll vs Blue Lion (W)This was the first fight of the entire tournament. As such, I was a little rushed and didn't have time to record it. But being the first fight, there was a decent crowd, including a news crew there to report on Robogames. As such, they recorded some of the match for their piece, and even interviewed me after about the win. Link to the article Re-Roll vs Duct Spartan (W)Second fight for Re-Roll against a D2 kit. Was actually a little worried about this one, as Re-Roll's drive was lackiing in the first fight. So I knew I'd have to win on damage. The thing that saved the fight for me was taking off one of their wheels. Win for Re-Roll on a judges decision. Link to video Re-Roll vs Badbot (W)Badbot was a Vector horizontal spinner kit bot. Re Roll took out the belt early. When they got stuck on the wall, they tapped out to avoid more damage. Third straight win for Re-Roll. Re-Roll vs Mean Bean (L)Mean Bean is a classic Fingertech Beater Bar spinner. It was a very solid bot, and Re-Roll just couldn't keep up. It was out-classed in weapon and drive speed. If I want to be competitive, I need to be able to beat this type of robot. Definitely some things to improve. But Re-Roll still put up a good fight. Loss by KO Re-Roll vs Killicake (L)Killicake was a gorgeous full body spinner. The entire outer ring was made of 1.5" AR500 and hit like a tank. Only strategy I had was to slowly back it into a corner hoping to get it to stop spinning long enough to damage it's underside. But an early collision took out one of my wheels and the fight was basically over from there. Final loss of the tournament by knockout. Overall I'm very happy with how both bots did. Both finished 3-2 on the weekend. For Re-Roll, I have some upgrades in mind, but will take some time to implement them. The design has a lot of potential, but needs some serious improvements to be competitive in modern beetleweights. For Drop Test, I think I've hit the point of learning enough at the small weight classes to start scaling it up. I plan to work towards creating a 30lb sportsman version that should not only be more competitive, but less likely to receive significant damage, as durability is one of the major downsides of the design.
0 Comments
I have had the goal of competing at Robogames for years. When I moved out to Bay Area in 2019, I was devastated to know that Covid had potentially ended the event permanently. When it was announced that Robogames would be back in 2023, I knew I had to compete. I knew I wanted to bring a few bots. Drop Test had been ready and was basically sitting in a box at this point. But I wanted something bigger, something more competitive, something to prove what I was capable of, something that had a chance of winning the largest robot combat event in the world. It had to be Re-Roll. As a vertical spinner with a wedge, it was right in line with the current meta. Designed to deal some damage, take a beating, and control a fight when necessary. It had to have multiple wedge attachments to effectively battle different opponents. It was perfect. I just had to get it ready in time. Over the years I've gotten worse and worse at taking build photos in the process. I get so into the build, I forget sometimes. An honestly looking back, it doesn't seem worth it. Most of the people reading here only really care about the finished product. In light of that, I'll keep this update short, focusing on the changes that happened from the original design, to the finished product. In general the build went really smoothly. The main issue I had was being underweight. Which is definitely a good issue to have. The issue seemed to be the 3D printed parts. I do my best to keep the weights in check, but definitely missed a few. It's a tedious task to get 3D printed parts accurately. My current system is as follows:
To do this for every part at this scale for every change takes a ton of time and can lead to a lot of mistakes. I think I have a plan to better predict these weights in the future, more on that in another post eventually. But for now, if anyone has a better suggestion, I'm all ears. With the extra weight I was able to do a few things:
All that being done, Re-Roll still came in under weight. I already have a few ideas for a next generation design. But for now I'm super proud of how it turned out. I never expect to win an event, but I'm really excited for the potential of this robot. Plus, I think it looks pretty. Finally, a photo of Team Conn Bots at Robogames today. Already through safety. Ready to rack up some wins tomorrow.
It's been almost half a year since my last post here. It's been a busy 6 months. Work is really busy for me right now. So I focused more on my professional robots than my personal robots over that 6 month. But I still got a lot done on my combat robots, I just haven't had much to show for it until now. The main project I've been focusing on recently is Drop Test, my 1lb Antweight spring flipper. If you scroll down far enough you can find the post where I was hoping to have Drop Test ready for Smashbotz 2021. That did not happen, for a lot of reasons I'll try to cover quickly in this post. What Was Wrong with Drop TestDrop Test 1.0 was a really fruitful proof of concept design for this spring flipper mechanism I developed. I spent a lot of time re-designing the robot for what I was calling Drop Test 2.0. I even made an entire post about the upgrades that showed how I overcame the short comings in the original design and left it off where I was going to build that design and implement it in time for Smashbotz 2021. Obviously that didn't go according to plan The first thing that was wrong with the original design was the concept of the 3D printed unibody. This was a concept I had from the original version of Drop Test that I wanted to improve on for the second version. The original frame was pretty weak because it was so light weight given how heavy the flipper mechanism was. I knew I could do it better for the second version. And I did, sort of. The problem was, at the time, I did not have my own 3D printer. So I was paying online services to do it for me. This got very expensive very quickly when I would print large unibody frames. Attempt to use them. Realize they wouldn't work for one reason or another, then have to wait a couple weeks to do the whole cycle again. Very frustrating experience overall. I have since purchased my own 3D printer, and created an enclosure and dry box for it to print nylon at home. This meant I could iterate on my own quickly and cheaply. Unfortunately, this meant I also had to deal with my own design issues. The original design used a lot of support and took forever to print, leading to some big failures. Between the difficulty in manufacturing and the difficulty in assembly, I moved away from the unibody in favor of 4, smaller, more printable, frame rails. One of the other main issues with the initial redesign I made was belt driving the flipper trigger mechanism. In an effort to make the weapon assembly lighter, I switched out the gears in favor of belts. In theory, these timing belts should be able to handle similar forces to the gears I was using, while being lighter weight for the same level or reduction. In reality, due to the scale, materials, and methods of manufacturing I was using, the belts would skip long before they provided the level of torque that the motor was capable of. The frame flexed to allow the less than perfect 3D printed teeth on the trigger mechanism skip, meaning the whole mechanism could never re load itself. This was something I brought up in my last post as something that would have to be fixed. I went though methods of tightening the belt more, but it was never successful to to the degree I wanted. I could make it work, but I'd have to give up a lot of flipping power. I ended up going back to the original gearing system, which meant I'd have to account for the extra weight somehow. I ended up modifying the geometry of the flipper mechanism to be a steeper angle. This allowed for the whole robot to shrink a little, giving me back every gram of weight I could. Finally, in my pursuit of an ultra high powered spring reloading mechanism, I had this grand idea to use a Pololu Baby Orangutan motor controller, potentiometer, and high powered gear motor to make my own high powered, multi turn servo motor that was integrated into the frame of the robot. This almost worked, but not really. It came down to two issues: fitting everything in as compactly as possible, and my programming skills. Unlike real servo companies, I don't have access to injection molding to make really tiny parts really precisely for making a servo motor on this scale. The components I had to use we're bulky and disjointed that they still took up a lot of space in the robot, making the whole thing larger. On top of the weight issue as a result, my basic at best programming skills meant that the feedback loop for the servo control was lacking and resulted in inconsistent reloading, especially given the belt skipping issue I was having at the time. That's when I had my big breakthrough find. This 5 Turn, High torque servo was exactly what I was trying to build in an off the shelf product. This servo was brand new and I had to have it. It's designed to go in Servo City's servo gearboxes which is almost exactly what I'm building here. It allows for servo control of the trigger over a large enough range of motion to be effective after the additional gear reduction. Best of all, it weighed the same as the custom setup I was making, but in a smaller package. These 3 major upgrades finally allowed me to make something that was functional and within weight for the first time for this version of the build. After way too many iterations, I finally had something close enough to complete to share. There's still a few small details I need to work out before I call this design event ready, but I'm so happy with how this turned out. It looks so much better than the original, both in design and performance. Really excited to get this thing in the arena. It looks so clean and cute though I almost don't want to fight it...Almost It had been almost a year since I'd last been to a competition. A long year, for a lot of reasons. November Necrosis the year before was a successful event, but not ideal. It was mid-November, and I'd heard it usually never rains in southern California. But, it was cold, it was rainy, it was outside. I'd been far to coddled going to Motorama exclusively for a couple years of my recent combat robotics career. I had a relatively successful tournament, but it could have gone better. Although I didn't finish Drop Test in time to attend this event. I remembered that I had prepared much better for November Necrosis the year prior and had over prepared for the event. I was planning on bringing both Hercules, my ant-weight lifter, and Drop Test, my brand new ant-weight spring flipper. Due to the large number of entries to the event, I could only bring one robot per weight class. This meant I only brought Drop Test, but Hercules was ready to go. I'd stashed it on a shelf in my garage. I was so close to selling it. With the number of new people getting involved in the hobby, prices for pre-made combat robots, especially kits, are in high demand. If I learned anything from selling El Tigre, they sell fast too. But I never got around to it. Partially because I never prioritized it over building the robots for the competition, and partially because deep down I didn't want to. Hercules was my first real successful robot. It had gone through so many iterations, and had been so successful by being so durable, I couldn't let it go. I was actually pretty excited to drive it. I had no real expectations for it. But it was such an established design at this point, it practically couldn't do poorly in my opinion. It had fought in Motorama 2019, and done okay. But I didn't love the opponents that I fought, and lost twice by decision. Hercules came out essentially unscathed, but left something to be desired. I wanted to see it reach its potential. Plus it was just so pretty. HerculesMy first fight with Hercules was against a new bot with a first time builder from a local robotics program called Dispatch. I could tell he was anxious about out our fight, so I tried to convince him his robot would at least be okay by the end of the fight because Hercules doesn't do any real damage. In all honesty he had nothing to worry about. He made a great robot, especially for a first try. He immediately flipped me over in the fight, rendering Hercules less than effective. His vertical spinner broke, but that actually benefitted him as Dispatch could control the entire rest of the fight, winning handedly by judges decision. Dispatch went on to win 3rd in ant-weights out of a pretty large field. Very impressive The second fight was against Chopper, one of two identical ant-weights. Only distinguishable by there color and driver. I competed many times against a pair of identical horizontal bar spinners at Motorama, both driven by the son of a father-son team. This pairing was extra special because the son was even younger, the dad drove his own robot, and they drove the same robots with heavier spinners as a multi-bot beetle-weight robot. I love it. Chopper was the sons robot, and done well, but simply out matched by Hercules. He didn't get any significant hits in and the design, while innovative, is rather unstable when flipped. Leading to Hercules controlling the whole fight. Hercules's third fight was against The Snip. I was not excited for this matchup, as Hercules historically dis poorly against vertical spinner. The Snip did very well at November Necrosis the year before as well. But Hercules was working pretty well. The snip used a single piece bent wedge which was an interesting choice as I saw its builder sizing up Hercules prior. Hercules was able to flip The Snip around, the same way it did Chopper. I thought he had me a few times, but the screws on the front wedge prevented him from really hitting the lifter on Hercules. He had planned to remove the wedge with the screws for a different setup, but was unable to remove those exact screws. That's how it goes sometimes. Details count. Fourth fight. Hercules continued to do what it had been doing all day, control spinners and get that judges decision. Radio Active was a solid little egg-beater, but it wasn't driving that well and I was able to not even get a scratch until they are on top of me and are able to clip one of the wheels. Because of a manufacturing error with the frame, the shaft extends to a place for a bearing, with no bearing. It was at this time I realized I should have done something about that. I need to make some weight on Hercules, but I plan to print bearing holders for it to actually support the long motor shaft. With one wheel down it was a boring rest of match as Radio Active had drive issues as well. Hercules wins for controlling the first half of the fight. As you can tell from the videos, it was getting late. I wasn't super invested in how well Hercules did, I just wanted to go home. But it was doing well and I wouldn't quit without getting properly destroyed. I was somewhat glad, yet disappointed to see this next and final matchup. It was two half pound titanium wedges. If it were just one, I'd be set. But two meant they could split up. Hercules has a low ground clearance, and has no way to damage them. So I was going to be out flanked. But at least I would come out unharmed. The fight went as expected handing the victory to Wedge Wedge, leaving Hercules 4-2 and fifth overall due to a forfeit. I'm very pleased with how Hercules performed, especially with essentially no prep work on my part. But I feel it could be better. It barely took any damage this event, again. So it's essentially ready to fight. That's a big part of why this design has lived on for so long, it can go 2-3 events without needing any real repairs. But that also means I'm not forced to iterate on it often. I have been trying to think of ways to improve it, and came across this old picture from this post: A design I drew up but never made. I actually like this. I'm not sure why I never made it, but I think with a little more thought, I could make this really effective. The concept is to do forks to get under wedges and some vertical spinners. But I'd want to tweak a few things. Namely make the external arms that pivot out of UHMW, make the inner ones longer, protect the servo arm and attachment point. I'll draw these up eventually and almost definitely make it. Should be relatively easy using Send Cut Send for the flat parts and 3D printing the more complex parts. ButcherThe biggest reason I had for coming to this event in the first place was Butcher. I'd put a lot of work into Butcher to make it work. I wrote about it here, and I was ready. It was the best it had ever looked. My dad put in a lot of work to make the frame for me, but all of those little details combined with some electronics upgrades made everything so much better. Butcher's first fight was against Washed Up. A drum spinner from a local robotics club that was attending the event. Great bot for a high school aged student to have designed and built. But it was clearly un tested as these things go. It takes a lot of time and resources to prepare long enough before an event to test everything out safely. The new, battle tested Butcher took it out with one enormous hit on the corner. Winner by knockout. Whomper was the opponent I feared most out of the hobby-weight field. Matt Vasquez, builder of the Battlebot Whiplash, created a horizontal spinner that Chris Rose would definitely call something Tombstone related. It was also in attendance at November Necrosis last year, but had a few unlucky matchups for a losing record. I was not excited for this fight as this was an experienced builder with a deadly weapon and something to prove. Whomper had the advantage in reach, but Butcher had armored wheels. The wheel armor eventually gave out after repeated blows letting Whomper smach the hub on one of the drive wheels. Butcher was left turning in circles, but Whomper in it's battered state became unresponsive. Giving the battered Butcher the win by knockout. There ended up being something so wrong with Whomper they had to call in another robot to push it into the wall as it could not stop it's weapon on its own. It was a victory, but it felt a little too lucky for comfort. Shredder XL was a vertical spinner that gave me an opportunity to try out my new tri-bar knife style blade. Theoretically it would make it harder to hit off axis when going head-to-head with a vertical spinner. I also just wanted an excuse to use the new blade as I needed to take the pulley off the single tooth blade anyway as it was destroyed in the Whomper fight. I never got to prove my theory as Shredder XL threw its belt immediately and ended up tapping out to give Butcher the win. Butchers fourth fight was against Target Practice. At 3-0 Butcher was still in the winners bracket and this fight was for a spot in the finals. Target practice was a stout lifter with a similar mentality and strategy to Hercules. I knew I'd have to cause some real damage if I wanted to win, because he wasn't going to go down easy. The new tri-bar design caused some real damage to both Target Practice and the arena. Causing the event organizers to pause the match to tri and fix the arena that Butcher just broke. Despite my protests that I should have won for breaking the arena, they decided not to encourage that behavior and continued the fight. Butcher ended up causing enough damage with proper aggression to win the fight. You can see in the picture below that the front corner of Target Practice was torn off and one wheel was busted, leaving Target Practice with some work to do and moving Butcher on to the finals. Again, you can tell by the change in lighting how late this fight was. I had earned the easiest spot in the finals and had to wait for the loser's bracket to shake out a champion to meet me. I sat around, watching Whomper slug it's way back through the loser's bracket, smashing its opponents in the process. It was clear we would meet again in the finals, which I was not looking forward to as I had only barely won the last match. I knew my only chance of winning was to hit Whompers exposed wheels, but with how squirrely the new drive was, it wasn't going to be easy. I did manage to get one hit on Whomper's wheel, and take a decent chunk out of the back, but it wasn't enough. About a minute in Whomper smashed the 3D printed weapon pulley leaving me no choice but to try and push him around to also break his weapon. This lead to Butcher taking a few more big hits for the rest of the fight. Eventually Whomper won the easy judges decision. As the winner of the winners bracket, I technically had a second chance at the title, but Butcher was in such bad shape I knew I couldn't get it back to a place that I'd be competitive against Whomper. It was so late and I was already really happy with how Butcher had done I decided to forfeit the last match meaning Butcher ended up in second place with a record of 4-2. Overall I was very happy with all of the new upgrades to Butcher. It was immensely more durable and reliable in the box. And that was really my intention with Butcher from the beginning. I wanted to make a horizontal spinner that was capable of dealing out big hits, but keeping on going. That hasn't been the case most of the time as it has been plagued with reliability issues, especially in the weapon. But I made great strides towards that goal with these upgrades. There were two main issues that I wanted to work on for the future: drive controllability and weapon pulley robustness. I think the first one is relatively easy. The new brushless drive is really powerful, but quite jerky. You can see me over-turn a lot in the videos from this event. I think this should be relatively easy to tone down with the proper settings in the SimonK firmware on the drive ESCs combined with some tweaks from my transmitter. The second one is a little more difficult but still doable. This version was 3D printed out of nylon for ease of manufacturability and cost. But I think I just have to go to a real machined aluminum pulley like I did on the weapon motor to make it work. With both of these pulleys cracking this event, if the aluminum version lasts any longer it will pay for itself anyway. I'd also like to try and shield the belt slightly to avoid instances where horizontals like Whomper can easily smash the weapon pulley.
Additionally I'd like to try and rework the armor somewhat. The drive wheel shaft support blocks I made, did little to nothing. The wheels didn't really need to be supported, but I was hoping it would provide some stability and armor during big hits. Instead it shattered on impact and threw my expensive bearings on the floor. I'll likely get rid of these and take that weight to add some thicker armor around the wheels. Maybe even a plow/wedge on the back of some sort to slow down Whomper should we meet again. Overall I had a great event and am really excited that events are coming back slowly. I'll continue to work on these bots when I have time. Hoping to get Drop Test up and running first. Then focus on these upgrades for Butcher. I'll continue to post updates here when I can as well. Until next time. After a long hiatus of both building and attending events, I've been ramping up to attend the Smashbotz 2021 event in LA on October 16th. My goal for this event was to bring two bots: Butcher and Drop Test. Butcher because it's still mostly in tact and only required a few minor upgrades to get it going again. And Drop Test because I feel like it's really close to something special as an ant-weight spring flipper, it just needs the proper work put into it to make it happen. Butcher 3.0Butcher's performance had been less than stellar recently. It had two core issues:
First, for the weapon system I had two parts that I changed out to make this issue hopefully disappear. First the weapon motor pulley in previous versions had a plane bore that was held on to the can by some small set screws and eventually just epoxied on there when the set screws failed. The core issue is that shock loads would get transfered back to the weapon motor from the disc when it hit something. The weapon motor in Butcher is big enough (4250 brushless outrunner) that it needs a secure attachment to make sure this joint doesn't fail. To combat this, I had a custom aluminum pulley CNC machined to perfectly fit the motor. You can see in the picutre below that there is not only a hole pattern on the top face of the pulley, to interface with the matching bolt pattern on the top face of the motor, but there is also a set of 5 bosses that are machined on the inside of that top face. These bosses slide into the cutouts in the top of the motor with a slip fit. These are designed to contact the machined body of the motor before the pulley can rotate enough to put this force on the tiny M2 screws in the face of the motor. The screws should be enough to hold the pulley onto the motor, and the machined bosses can take the shear loads from the sudden impacts meaning hopefully this pulley will never come apart from the motor. A less critical but way cooler change I made for this version of the weapon system as well is the new disc I designed. Again this is detailed in the design post linked above, but instead of the large single tooth disc that I have been running, I created a 3 tooth disc of roughly the same weight that looks like a set of spinning butcher's knives as an ode to the name of the robot. This disc not only looks way cooler, it will be nice to have as a secondary option instead of the single tooth disc. In theory this will be able to deliver smaller hits while continuing to spin as well as spin up faster given the smaller MOI. Both of these traits will be really useful in fights against some of the tank like wedge robots that are commonly found in the hobby weight class. The other system that got a major upgrade was the drive train. Like I said above, the issues I was having mostly came down to the fact that I had picked inrunners for the old drive train because I didn't want to have to deal with outrunners and protecting wires from them. But after multiple events of the motors overheating and torching themselves, twitchy drive due to the high input RPM, and multiple occasions of the pinions on the 3mm shafts coming loose, I decidided an upgrade was necessary and long overdue. The drivetrain upgrades started with the motors. I switched to the Turnigy Aerodrive SK3 3536 1050kV motors. These are lower kV, higher power, higher torque motors that should help with the drive issues I'd been having. They also have a 5mm output shaft that should help keep the pinion gears more securely attached. The one downside to these motors however is the fact that as outrunners, the can of the motor is spinning at a very high rpm and can wear down anything touching it, including key electrical wires that would cause you to lose a fight. To protect these cans I developed some 3D printed wire guards that bolted onto the frame that surrounded the motors entirely. I wanted to continue supporting the rear of these motors like I had in previous versions of this design as I had had the problem of the motors coming loose of the gearbox in the past and wanted to make this drivetrain bulletproof. Because these are outrunners now, that meant adding bearings into the motor guards that slipped onto the nub on the end of the brushless motors to support them and avoid this issue. I took a similar approach to the output shaft of the gearboxes by adding a bearing into the wheel guards on the other side of the wheel. I haven't done this in the past due to weight, but with some of the recent upgrades, I made weight for some of these more experimental design elements. I'm interested to see how it works out. I should prevent the gear boxes from seeing big shock loads from being thrown around the arena, but might over-constrain the output shaft if it gets warped and might cause a different failure in practice. I'm really hopeful this version of Butcher will live up to its name. It was designed to be a big hitting bot that itself could take a few good hits, but recently it's had small issues that lead to big performance issues. We'll see how these upgrades hold up to the abuse of the competition at Smashbotz. Drop Test 2.1Part of the reason this post has been such a long time coming was because of Drop Test. Being by far the most complex combat robot I've designed to date, I expected there to be some growing pains and setbacks. But there's been even more than I expected. The core problem is that flippers in general are hard to build. They typically end up being relatively fragile because you have to fit so much in the same weight limit than other designs like spinners and wedges. The flipper mechanism takes up so much of the weight you don't have weight left over for other things like armor. In order to cut down the overall weight to make the design in the limit, I made some design compromises that came back to bite me since the previous update on this bot. I won't go into too much detail here, but one of the key issues was that in order to cut down the total weight of fasteners in the bot, I tried to make the frame all a single 3D printed piece. This ended up with me spending lots of money outsourcing the printing of the frame, only to have issues with is since it wasn't optimized for 3D printing and having to scrap the entire frame multiple times due to small errors. I have since bought my own 3D printer and re designed Drop Test to be more printable. You can see a PLA mockup of the new design below that shows the 4 independent frame rails that are much more printable, even without supports, and make assembly and maintenance way easier. An issue with these changes though was weight. Like I said, I had designed the unibody to eliminate fasteners, but this design, though it had it's own advantages, had about twice the number of fasteners. This meant both the screws that hold everything together and the brass heat set inserts in the frame. And at such a small weight limit, that added up quickly. So weight had to be cut elsewhere. I had originally planned on using the PCB from a large servo I disassembled to drive the motor for the flipper system, but at 11g and fairly large, it definitely wasn't ideal for the re-design. I found a smaller servo board from a slightly smaller servo, but it was slightly under the current rating of the motor I was trying to use, and I didn't love the fact that I had to buy the whole servo only to waste most of it to steal this board for my own use. That's when I can across the Pololu Baby Orangutan, a "robot controller" that integrates the same microcontroller as an Arduino Nano into a smaller package with built in H Bridges for motor control. At 1.5g without the headers, this seemed perfect for my application. I could program it so serve as a servo board with a trigger input from the receiver to be able to run the firing and reset sequences without me having to worry about firing and resetting while I'm driving. I forged ahead with this as my selection, completely disregarding the fact that I'd never used this system before and would have to learn to interface with it and source/create code for the servo feedback system to drive the trigger mechanism to my desired positions and timing sequences. As you might be able to tell from my bitter retrospective tone, this didn't turn out that well for me. This project throughout its lifetime has been the most complex combat robot I've made to date. And as such, requires much more focused attention to detail than any of my other projects. Given that combat robotics is just a fun hobby for me, I don't always put the same effort into details as I would for my job. I also don't spend nearly as much time on it as I would for more professional projects, maybe a few hours a week. All of this meant two things:
All that being said, I did make some good progress on this design, it just didn't come to completion. The biggest win in my opinion is the fact that I got the motor driver working as intended. I was unable to get a good video of it as the belt that connects the motor to the trigger was too loose and couldn't handle the loads required. As such, the video below shows the auto firing of the mechanism by way of my exciting the triggering pin on the controller by hand to simulate the PWM input from my transmitter. You can hear the belt skipping in the video, but it does work. This isn't the update on Drop Test I was hoping to give, and it likely wasn't the one you were hoping to read. But I promise there will be more updates in the future with better news. I won't give up on this, at this point, it's personal.
In the past, a lot of my designs have been created out of a desire to do something different. Hercules was designed to be a tank to counter the high power ant weight spinners of the time. Butcher was designed to be a high power horizontal spinner to counter the Single-Tooth-Disk vertical spinner meta that is becoming more popular in all weight classes, but especially the hobby-weight class at the time. And my most recent bot, Drop Test, was to prove that with some additional work, you can make wedges less boring and still effective. One of the few designs I've made over the years that was designed to be competitive was Barrel Roll. Barrell Roll was a vertical drum spinner, 2WD wedge bot. Very inline with the current meta of winning robots. Unfortunately, my design and manufacturing skills weren't up to my lofty aspirations. Barrell Roll was plagued with problems: off balance weapon, RF interference issues, underpowered weapon motors, durability issues, etc. Barrel Roll never won a fight, even through 2 iterations. It was a frustrating experience to say the least. I eventually gave up on the design for my other beetle-weight design, El Tigre. I've since built and sold El Tigre and haven't had a beetle-weight since. Recently, I've had the itch to design another beetle-weight, and given the fierce competition out there in the beetle-weight class, I want it to be competitive. To show not only how far I've come as a designer and engineer, but to show how far combat robotics has come in the past 5 years, I wanted to re-design Barrel Roll. I wanted to keep the same overall layout and design, but with some serious upgrades that should help solve lots of the issues with the original design. Looking at the last version of BR, I knew I wanted to keep the wedge shape, the twin vertical disks, the wide, and the slender frame. Lots of the issues that the original Barrell Roll had came from the electronics reliability and the material choices. Both of which I'll get into later. I wanted to start from scratch for this design to help eliminate any lingering design issues from the previous version. My design process as of lately always starts with a layout sketch in Solidworks. This allows me to easily and quickly get a feel for how big things are in relation to one another. Later on, I use this sketch to actually make some of the parts for a "top down" assembly design where the single sketch controls the dimensions of many parts of the assembly in order to easily tweak the design later on. There's a lot goin on in this picture, but I'll highlight a couple of things. You can see the overall design is still the same: big wheels, big vertical spinner, front wedge. You can see the 'rabbit ears' on the top that prevent the disk from hitting the ground when upside down and where the ground would be in that case. You can also see the layout for wedgelets as an alternate wedge layout for fighting other wedged bots. I didn't take any pictures during the design process, but I'll highlight some of the design features here. In the first picture you can see how the layout sketch directly correlates to the design. The overall design ended up being a little narrower and longer front to back than I originally wanted. The original design had lots of design compromises to make that happen, so although this doesn't "look" exactly how I wanted, but has much better engineering design principles that pushed it to look like this. Weapon SystemFirst, the part that people care about the most, the weapon system. The twin single tooth discs are 3/16" AR500 steel with a 3.25" effective diameter. They're separated by a mirrored pair of 3D printed spaces that integrate a timing belt pulley profile into them. Each of these parts is wide enough for a single 3mm timing belt from Fingertech so the entire weapon assembly has space for dual belts. The assembly is held together by 4 3/16" shoulder bolts. There are 3D printed end caps that hold the bolt heads and the nuts, offset so they can be the same part on each side. Instead of bearings there are 2 high load bushings that run on a 3/8" steel shaft, similar to my setup for Butcher. The weapon motor is a Propdrive 2836 brushless outrunner with another custom 3D printed timing pulley that fits around the can to save space. There is another 3D printed part that both supports the output shaft of the weapon motor as well as shielding the wires that run between the two halves of the bot. And finally a 3D printed cover to protect the belts from attacks on the top of the bot. This is a lot of printed parts to solve these issues, but I'm hopeful that when printed in the right material they'll be able to standup to the abuse of the beetle-weight class. Hopefully by the time I get around to building this design, I'll have my own 3D printer setup to print nylon parts to keep overall costs down by not outsourcing these parts. ElectronicsA lot of the problems that the original Barrel Roll had were in the cheap electronics I was using. I replaced the cheap chinese drive motors I was using with the Rectified robotics drive motors and wheels. They seem super solid and I'm excited for these big chunky wheels to help this design. The motors are mountd with some Servo City 22mm mounts to keep the width of the robot down and help support the gearbox instead of face mounting the motors. Drive ESCs are still the Fingertech Tiny ESCs as I've yet to find a better ESC for this weight class. The biggest challenge I had in this design was finding a proper battery to fit in here. There are limited options for batteries in this capacity range with even more limited size options. I ended up settling on a pair of Turnigy Nanotech 850mAh 2S LiPo batteries in series for the equivalent of a 4S 850mAh battery. Weapon ESC is a Turnigy Multistar BLHeli 32 41A ESC. It's pretty crazy to me how small these ESCs have become over the years with the advances of drone technology. Also in the picture above is the Flysky 4CH receiver and Fingertech power switch, both staples of all of my insect weight robots. One of the big problems with the original design was that I didn't put enough consideration into the wire routing. This resulted in the wires occasionally being clipped by the weapon and the bot killing itself. For this design, there's a specially sized channel at the back of the bot that keeps the wires separate from all of the spinny bits in the center. This should also allow for easier maintenance by removing the rear frame piece as easy repair has become more and more important to me over the years. There never seems to be enough time at events when it counts. StructureThe main structure of the bot is 3/8" UHMW. They're 2D profiles with some end tapped holes. These should be easy to CNC route and hand drill the end tapped holes. The front wedge and rear plate are 1/16" Titanium that will probably be from Send Cut Send. I just need to figure out a way to bend the wedge the way I want it. Top and bottom plates are 2mm carbon fiber from CNC Madness. The bunny ears are 1/8" Polycarb from SCS as well. These also act as end caps for the weapon shaft to help stiffen the frame even further. All of this should be a lot sturdier and more durable than the old all aluminum body. Forks EditionAs previously mentioned, in the competitive nature of this design, I've designed an alternative configuration for the bot to face other wedges/vertical spinners. I've never been a huge fan of alternate configurations, but if you want to win, you've gotta play the game. If everyone else is doing it and you're not, you're putting yourself as a serious disadvantage. This configuration keeps most of the bot the same, but changes the front wedge to something super minimal to make weight for 4 hinged forks. These forks are 1/8" titanium riding on 3/16" shoulder bolts for their hinging action. The shape of the wedgelets along with the small wedge keep them limited to about +/-15 deg of travel which is hopefully enough to compensate for floor inconsistencies while preventing the bot from high centering itself. Both desidgns come in at 2.90 lbs leaving me some wiggle room to make weight with the variability of materials and things that aren't included in CAD like wires and connectors. The center of mass is just inside the wheel diameter meaning the bot should be very control-able while maintaining enough weight on the front wedge to still be effective.
Overall I'm very happy with the design. I think it showcases a lot of the things that I've learned over the past 5 years on the design side, as well as a lot of the advances in the field of combat robotics including smaller electronics, combat robot specific parts, and readily available online manufacturing resources. I hope to build this up sometime in the next few months or so after I finish rebuilding my other two active bots, Drop Test and Butcher. I'm definitely excited to get back to being competitive in combat robotics. With that I'll leave you with this lame attempt at a render of my design so we can all pretend it's already real and not just some files on my computer. After getting thoroughly destroyed, twice, at November Necrosis, I wanted to spend some time improving Drop Test with the free time I had over the holidays. As a reminder from that post here are the main conclusions I had after that event:
With these goals in mind, I started my redesign. I finally had access to Solidworks again, so I was planning to move the old assembly from OnShape to Solidworks, but with so many things changing, it was easier just to start from scratch with the learnings I had in mind. As with the first version of Drop Test, I started the design with a general layout sketch. The overall flipper design worked well in its first competition, but needed some tweaks to help improve. First, I wanted to shorten the overall robot to make it more compact so that I could increase the durability of the 3D printed frame. This meant increasing the angle of the wedge slightly to compensate. This meant moving things around slightly to maintain the same overall working principle. As you can see in the picture below, as a result of this re-shuffling of components, the trigger mechanism (below left, right side) is fully contained within the bot instead of sticking out of the rear of the bot like it did in the first version. This shortening also meant that the leaf spring was shorter and had a greater tip deflection when the flipper was fully loaded. Without going into the details too much, this should create a higher flipping force than the original design as well. Finally, I realized that the original layout of the trigger mechanism was highly inefficient mechanically. The main issue was that the trigger had a roller on it that acted as a pseudo cam follower to ride along the backside of the flipper arm to load the mechanism. I realized that if I switched the follower to the flipper arm side it would be more efficient and closer to a traditional snail cam design. Essentially, if you imagine the pair of rotating arms as gears instead, with the roller on the trigger, the driving gear had a fixed, large diameter that would result in a "gearing down" scenario which produces more speed and less torque. Instead what I wanted was a "gearing up" scenario where the roller, now on the flipper arm, contacts the trigger arm near the pivot point, such that there is a much smaller torque arm, similar to a small driving gear. This change should allow me to use a lower gear reduction on the flipper motor to allow for a quicker reset time overall. For more information on the math behind all this go check out the original post. With the layout sketch complete, I started in on the actual parts design. I didn't take any picutres during the CAD process, so I will split up this section into focused reviews of system upgrades to the robot. Wedge DesignThe idea behind the original wedge design on the original bot was to help protect the delicate flipper mechanism from horizontal spinners. The dual angled face of the wedges would theoretically deflect any incoming blows from spinners. This idea worked (until the 3D printed body that the wedges attached to snapped in half), but provided additional issues. Mostly it made it very difficult for me to control y opponents and line up the flipper as they would usually just slide off of the side of the bot when I tried to push them up on the wedge. For the redesign, I opted to make things a little more simple and have a flat wedge face that spanned the whole width of the bot. Having this wide flat surface should help to control my opponents much more easily. To handle horizontal attacks, I angled the outer edges of the wedge, similar to Duck. This likely won't be as resistant to large horizontals, but should still provide some deflection to prevent the entire wedge being ripped off with minimal material. The real upgrade here is more optimally using the inner volume of the bot by changing the wedges to allow for a stronger attachment method for the wedges. Following the Battlebots upgrade theme, I wanted to have a small tongue on the end of the flipper wedge to allow me to more easily get under bots to flip them. I decided to do this by bending a flange on the end of the flipper wedge in the middle of the bot. I've tried to avoid metal bending up until this point in my designs as I don't really have the proper tools for it. This being 1mm titanium, I'm fairly confident I can bend it by hand using the tools I have well enough. To help with this process, I added two little notches in the wedge where the bend line is to be. This should help the material naturally bend along the line that I intend to bend it on. Finally, learning from my first design, I decided to forgo any countersinking on the wedges, because countersinking titanium with hand tools takes forever and only provides minimal benefit. Drive TrainAnother reason that I was having trouble controlling my opponents was the fact that the wheels were all the way at the rear of the robot. Due to the height of the robot, I was worried about the bot getting stuck on its back and not being able to self right. This meant that the wheels had to be on the rear corner of the robot to keep them in constant contact with the ground to help right myself from any orientation. With the wheels at the rear, and the robot being so long due to the flipper assembly, it was very difficult to drive and control opponents. I knew I was likely going to shift things around with this version of the robot to make it more compact, but that meant moving the COG further forward, making it even harder to drive. To prevent this issue I decided to make this version of Drop Test 4WD. When I started this process, I realized, looking back at my previous bots, that I had never made a 4WD robot before. Whether it was due to the style of robots I had chosen, or my design style in general, I never really had weight or room for 4 wheels in my robots up until now. I knew I was going to be tight on weight again, especially considering I was trying to make the robot more durable, so 4 motors to drive 4 wheels was out of the question. I might revisit the idea in the future though as the N20 gearmotors are about half the weight of my current drive motors, although less durable. My solution for this was belt drive. The rear wheels would be connected directly to the drive motors, while the front wheels would be spinning on a dead shaft. The two sets of wheels would be connected by an O ring acting as a round belt. This required me to design my own hubs that incorporated round belt pulleys into them. Given that I was already 3D printing lots of parts for this project, it seemed easy enough to design and print my own hubs specifically for the task. Both hubs would use the Fingertech foam tires that I used in the previous version. The rear hub is based off of the original Dale's Hubs that were sold by RobotMarketplace years ago, where the main hub mounted to the motor shaft through a setscrew and an additional plate screwed into the end to secure the wheel to the hub. The front hub would ride on a shoulder bolt that would act as a dead shaft for the hub to spin on. For this hub, since I can't screw on an additional plate to secure the wheel, I intend to simply glue the wheel to the hub and treat them both as disposable since they are cheap enough. Both hubs include a pulley feature to hold a 3/32" O ring that is stretched 15% past nominal to tension the belt to the point that it will drive the front wheels. I'm not entirely confident this system will work as intended, but for a first go at a custom belt driven wheel solution I think it's good enough to build to test out. Flipper UpgradesWhile the first version of Drop Test achieved my goals of successfully flipping a bot in combat, it had it's flaws. Due to my underestimation of the power required to load the intended flipping mechanism, and weight constraints, I had to sacrifice flipping power to make everything work. As a result, the flips that it was able to dish out were more of quick lifts than launching my opponent like I wanted. I also switched from a servo motor to a regular brushed DC gearmotor to be able to have more power for the same weight. This gave me more flipping power, but at the cost of not knowing the absolute position of the trigger mechanism. This meant I had multiple misfires in the arena leading me to defensively driving around while I reset the flipper, instead of attacking and controlling my opponent. To fix the flipping force issue, as I mentioned before, I decreased the length of the spring while increasing the deflection. Looking at the equation below we can see that this should increase the force at the end of the spring, leading to a greater flipping power. The other spring related upgrade that was made was to remove the mounting holes from the spring and instead, hold the spring at it's edges with screws. This is because spring steel is really hard and very difficult to cut holes in without the proper equipment. This removed the need for the spring mounting plate as well, freeing up some weight for the rest of the upgrades. To address the issue of the trigger positioning, I revisited the original design log for this robot. The original idea was to use a servo to power the flipper such that, when using a self centering stick on a transmitter, one direction would be to fire the flipper, and the other would be to reload the flipper, with the self centering position being the loaded position. I was unable to find a servo with enough power for my needs that wouldn't weigh so much that it would be detrimental to the rest of the robot. At the time, I really liked the idea of using the servo controller to control the output after the gear reduction, similar to how servo gearboxes work. It was very difficult to design though, without the parts on hand to measure and integrate into the design. And with all of the other difficulties I had with the initial design, that idea got pushed. Now, having a programmable servo on hand that I could take apart to use for my needs, and having the motor setup already established, it would be much easier to incorporate. In this version of the bot, I wanted to change the trigger drivetrain from metal gears to timing belts to save weight and allow me to re-organize the internals of the robot. As you can see below, I had already designed the trigger to be 3D printed with an integrated timing pulley. This allowed me lots of flexibility in the design and positioning of my servo feedback system. Because the trigger shaft was now a dead shaft, the easiest way for me to connect the potentiometer to the trigger was through a pair of equal sized gears. One integrated into the trigger, and one attached to the potentiometer. This was the servo control board that I salvaged from a servo would know the direct position of the output arm as intended. General UpgradesThe electronics for the robot stayed almost identical to the previous version, with the exception of the servo board replacing the TinyESC that was driving the trigger motor. The main upgrades to the body were in packaging. I nestled the flipper motor underneath the flipper arm, allowing the whole robot to be significantly narrower. This freed up some weight to thicken up the walls of the bot for better durability. This also meant I had to move the wire run between the two sides of the bot to the rear as the front was very crowded now. I also had to get creative with the mounting and clearance for the weapon motor to allow for the frame to be printed as one piece and still be easy to install. Finally, the robot came in at a very nice weight in CAD of 442g with the weight roughly centered over the wheels. This should allow for good mobility and leave enough weight for things like wires and latex for the wheels that aren't accounted for in CAD. It was much easier to make weight this time around with all of the lessons learned from the initial design. I took advantage of the long, holiday weekend to make some very much needed upgrades to the Butcher. Winning only 1 fight at November Necrosis and having both losses be a result of poor design choices rather than actual in fight damage pushed me to give the bot a well deserved redesign. From my previous post, there were a few things I wanted to keep in the design. I've been loving how well the layered, through bolted UHMW has been working for this bot. Still relatively cheap and easy to make, tanks big hits, and is light enough for the large body needed for a horizontal spinner. I also really liked the weight distribution in its previous event. It was much more drivable than its previous iterations. There were still a lot of down sides though, mainly related to reliability. The drive system itself needed some work to prevent the brushless motors from braking themselves, either by impact, or by loosening from the pinion gear. The weapon system also needed some work. Before it's last competition, I had worked on increasing the reliability of the weapon stack itself, but that meant that there was more stress on the weapon motor, leading to the pulley shearing off the motor, leaving the weapon useless. Drive Train UpgradesThe old drive system was sort of hobbled along through the iterations of the bot. The initial design had brushed DC motors that were heavy and underpowered, though reliable. I tried to use some drop in replacements that were smaller, 1100 kV inrunners because I needed to save weight because the old weapon disc was 4 lbs and I didn't want to deal with outrunners. Because I was remaking almost every part of the bot, I decided to re-approach the drive system. I decided to go with the well established SK3 3536 outrunner paired to a Banebots P61 16:1 gearbox. This should be slightly slower than my previous drive combination, but have more low end torque and be more driveable in general. With this new motor being an outrunner, I knew I needed to protect the spinning can from the wires on the inside of the bot. I took that need and combined it with my desire to better support the rear of the motor to prevent the relatively small face mounting screws from ripping themselves out during impacts. I created a 3D printed bearing block that supports the end of the motor while simultaneously shielding the spinning can of the motor from the rest of the electronics. I decided to keep the 4" Banebots compliant wheels as they seem to work really well in this design and give me a little extra ground clearance that the bot definitely needs. With some left over weight from the re-design, I added another 3D printed bearing block to help support the end of the drive wheel shaft to make the system even more bulletproof. Weapon UpgradesThe main problem crippling the weapon system up until this point has been the attachment of the weapon motor to the drive pulley for the weapon. The original design had it just press fit on, but because I was young and didn't know how to machine things to tolerance, the plastic began to slip and fail over time, loosening the fit, causing weapon issues. I later moved to a custom aluminum pulley that had small set screws in there to help keep it attached to the can of the motor. Those did basically nothing besides chew up my belts though. The core problem is that when the weapon hits something, that change in velocity in the form of a big impulse gets transferred back through the weapon belt to slow down the motor. With a properly tensioned belt, there is only a small amount of slip when this happens, but some of the energy is transferred to the drive pulley, to the weapon motor through whatever is joining the two. This is roughly equivalent to using the weapon pulley as a hubmotor and hitting things with the pulley. The joint between the pulley and the motor has to be able to endure these shock loads. To better understand what I was trying to do, I spent some quality time with my calipers to create an accurate model of my weapon motor. The motor pictures is actually the Turnigy Aerodrive 4250-500 when the motor I currently have is the 4240 version. Assuming it's the same thing, just 10 mm longer, I decided to design around the larger motor as I had some extra weight and space for the larger motor and it was over 1.5 times more powerful than the motor I've had in the bot up until now. Looking at the motor, there isn't a lot for a large diameter pulley to grab onto. Traditionally, these motors are designed for airplanes, so the propeller of the plane would be mounted to the motor shaft through a prop adapter that clamps onto the 5mm shaft. This won't work in this application as the motor itself is almost the entire height of the robot. The can is smooth and relatively hard. Only the top of the motor bell has any features that could be used to lock a pulley onto the motor. There are 3 tiny M2 screws around where the rotor attached to the main shaft. Again these are designed for ultralight propellors to attach to the motor, not the shock loads of combat robotics. My decision was to design a custom part to be machined out of aluminum if it isn't crazy expensive to make. The part would slip over the end of the motor, locking into place with positive features that matched the recesses in the end of the rotor, and held in place with the M2 screws. This way the shock loads should transfer through the bosses on the underside of the pulley to the motor can, leaving the M2 scews to only handle tensile loads trying pop the pulley off of the motor. In theory at least. I plan to have this test printed at least once before I try and place an order for a real machined part. In the picture above, you can see the bottom view of the robot, with the bottom frame piece hidden. This shows hoe the weapon motor is mounted as well as the new pulley. As part of these changes, I also moved the weapon motor back to be able to better place the tension idlers where they wont get hit by the blade, but have enough travel to properly tension the belt. With the new weapon motor upgrades, the weapon only has a theoretical top speed of around 4,500 RPM, which is relatively low compared to some spinners in the field. The hope is to have a much faster spin up time as most arenas for 12 lb bots are relatively small given the speeds at which these things move. This will hopefully let me land more consistent hits and stay spinning the whole match. Top speed is less important for this right now as the weapon is still over 3 lbs and almost 10" effective diameter, should still hit hard, even at lower speeds. I've never successfully made it to "top speed" with this bot anyway as the weapon system has been plagued with issues it's whole career. Given this lower top speed I wanted to try and design a disk with more teeth for when I fight wedged opponents and am not going for the knockout hit, but rather lots of highly controlled hits that don't send me flying everywhere. I had this concept in my head to make a disc look like knives as the bot is named "The Butcher" after all. I'm actually really happy with how the design came out actually. Looks really sturdy, uses the same pulley interface as my disk blade, has both positive rake and tooth relief, and comes in almost the exact same weight as the other disk. Even if it isn't as effective as I'd want it to be in combat I think it looks really cool. Might have to get some stickers made up of this as a logo to throw on the top of the bot. General UpgradesIn addition to these two major upgrades, there were a lot of small things that I fixed that don't really need their own section. Mainly fixing screw and nut clearances to make it easier to assemble. I got rid of a few of the shoulder bolts and frame pieces in front of the drive motors, opting instead to use the drive motors as from pieces to mount the top and bottom plates to. Like I mentioned in the intro, I wanted to keep the weight distribution the same and looking at the picture below, you can see where that ended up. I usually try to keep my COG just inside the front edge of the drive wheels for a two wheeled bot. That way the bot won't tip up when accelerating quickly, but is still controllable. Finally, here's a nice glamour shots of the CAD as a whole:
After a very long hiatus from robot combat due to moving across the country and events being cancelled doing to COVID, I was finally able to compete at my first event of 2020 and it felt great. I didn't win any prizes, but I had a blast, learned a lot, and have lots of ideas for improvements for my bots in the future. The ButcherThe Butcher's last event went pretty poorly. It went 0-2 at Motorama with two tough losses that I could have easily won. A big part of why I lost was the reliability of the weapon. Between moving and having limited access to tools, I didn't have a lot of time or resources to upgrade The Butcher, but I wanted to try and do something to help the weapon be more reliable. To do this I did a few things: First, cut down the weight of the weapon. This would decrease spin up time and leave a little more weight for a more robust pulley setup. Second, I wanted to make a custom pulley that had deeper grooves. The off the shelf parts I had been using had such shallow grooves that the belt would fall out almost every match. I decided to make my own pulley, complete with deeper grooves and a custom spline to transfer the power to the weapon disk. Finally, as part of this pulley upgrade I wanted to add another bushing to the weapon assembly. This would help to keep the disk from tilting on big hits which would potentially clip the weapon belt. I designed the pulley in a way that I would have the weight to make it out of aluminum, but for both cost and weight reasons, I wanted to experiment with 3D printing it. I had the custom pulley printed out of NylonX from 3DHubs. The new weapon disk was cut by SendCutSend out of AR500. Both parts came out great and fit together perfectly. The new weapon assembly was around 1lb lighter, with more, larger bushings. I had planned to use the extra weight to upgrade the drive motors, but I ran out of time before the event. I did get larger wheels to increase the ground clearance of the bot to prevent it from getting high centered. At over 1lb underweight, even with wheel guards, and no time left, The Butcher was ready for the event, all it needed was a fresh coat of paint. The Butcher vs Minor ThreatMinor Threat and Butcher have had multiple fights in the past, and they are always explosive. Being the first fight of the event, I was hesitant to get too much damage that I couldn't continue on. I wanted to fight strategically. I knew the drive was working pretty well, I just needed to wait for my opening and avoid a weapon on weapon hit. I got in a few good shots, and in Minor threats tumbling around, it got in a good hit on the top of the Butcher. When it finally stopped tumbling, I got a good hit on the wheel and Minor Threat tapped out. Nice solid win to start the event. Result: Win by KO The Butcher vs QuicksilverWith very little repairs needed to the Butcher, I was ready for the next fight, Quicksilver. Quicksilver is a very good wedge that has lots of magnets for increased downforce. I knew I would have to fight smartly. I tried to maneuver around to get the weapon spun up to avoid the box rush. The first impact dislodged the pinion gear on the right drive motor so I had only one motor for drive the whole time. I was also having trouble spinning up the weapon for the whole match, only to find out afterwords that the motor pulley had come loose from the can of the motor that it was pressed/glued onto. With those two things I had no shot of winning and was pushed around for 3 minutes straight. Result: Loss by judges decision The Butcher vs MaverickAfter the fight with Quicksilver, I tried to fix the issues I was having with lots of 5 minute epoxy. I wasn't super confident in it, but hopeful that I would struggle through the next fight well enough to pull out a win. The epoxy held for a while, but the large impacts from the drum of Maverick knocked the weapon motor pulley loose. I was able to stop Mavericks drum, but with the loose motor pulley, I was never able to land a solid hit. Maverick was a better shape to push me around. Result: Loss by Judge's decision The Butcher ended up 1-2 on the weekend, but I was still happy with how it performed. The main goal of the Butcher was always to make big impacts. I was successful in my first fight, but the reliability still isn't to where it needs to be to consistently compete. I have some improvements in mind, that include rebuilding most of it. Things That Worked:
Drop TestComing in to the event, I didn't have any expectations for Drop Test other than wanting to flip something at least once. I knew the bot looked good, but I was worried about it's durability, especially with the fleet of big hitting bots at this event. Drop Test vs Death by Single Cut (DBSC)The first opponent I had was a terrifying overhead bar spinner that was on a huge winning streak. Not a great matchup for Drop Test as the bot wasn't the most durable, and the height of the bar was right on the edge of the titanium wedge and the 1/16" UHMW top covers. My plan was to box rush him and never let him spin up and try and flip him over and disable him. That went well for a little while until he started running around and spun up to full speed. From there I didn't have much of a chance. It was more like death by 3 or 4 cuts, but after a few big hits I was missing a wheel, the flipper motor ESC had been disconnected from the motor and the frame had been split in multiple pieces. Result: Loss by KO Drop Test vs Blue WaffleAfter the fight with DBSC, I had to basically completely rebuild the entire bot. As a result, the flipper motor wasn't functional by the time I rushed it into the box for its next fight. The video above isn't the best (it's a much better view of my dog than the fight), but there wasn't much that was missed. Blue Waffle had also suffered a lot of damage in its first fight and wasn't able to spin up. With no weapons, Blue waffle was eventually high centered and counted out. Result: Win by KO Drop Test vs Shredder 249With some extra time after it's second fight, I was able to get drop test back to working condition for its third fight. It was mostly a classic pushing match between what is essentially two wedges. But Drop Test finally got in a few flips. I was very happy that it worked. Definitely could have been better, and I could have practiced with it more. But it was a good showing overall. Result: Win by Judges decision Drop Test vs Propaganda MachineI don't have a video for this fight, but I think the picture says it all. The big horizontal spinner took the entire front off of Drop Test. Didn't get any flips in before tapping out after only a few hits.
Result: Loss by KO This being Drop Test's second loss and the second frame that was split in half it was out of the competition. Overall I was still very happy with how it performed even if the number of pieces it was broken into was greater than the number of flips it successfully perfomed. Things That Worked:
Things That Didn't Work:
All in all, for a new bot I was very pleased with a record of 2-2 on the weekend. And honestly, getting destroyed isn't the worst thing as it forces me to reconsider a lot of the design choices to better optimize the bot for the future. I plan to update this page soon with the planned upgrades to both these bots. After a brief hiatus from combat robot activity for academic reasons, I wanted to get back to what this blog was originally about, combat robotics. This post will not only be a combat robot related post, but it will also serve as a record of my typical design process. This post to showcases the entire design process from concept to competition. So buckle in, this is going to be a long post (or skip to the end if you just want to see the completed project. I've wanted to make a spring powered flipper robot ever since I have been involved in combat robotics. I have always been fascinated with the obscure and different ideas in combat robots that could still be competitive. Spring powered flippers have always been in that niche for me. But the few people that have tried to make spring powered flippers have always been at best mildly successful. I'm making this post to show how much design work, iterations, and refinements go into something like this. And that's not to brag, but rather show that it takes a lot of work, but in the end it's worth it to make something cool. I began my research years ago on youtube, looking for previous ideas to build my robot off of. I found the following two videos on youtube that later lead to this Ask Aaron post. Both videos were very early concepts of a design like this, not finished projects. And both were just old enough to already be antiquated in the fast paced, ever evolving tech world of combat robotics. I loved the second video I found, It seemed perfect. It was an antweight robot that had a spring flipper. It was exactly what I wanted and it seemed to be powerful, although an unfinished iteration. Using the help from Mark on the previously linked Ask Aaron post I was off. I began drawing and making designs using my rudimentary Solidworks skills at the time and tried to make something that would work. The design started with scouring through McMaster Carr to find an appropriate spring. It's been so long since I made this model I couldn't tell you what part it was. But I knew it had to have a couple pounds of force at full extension in order to have a good flip. This was because there was such a short distance over which the flipper could impart energy to the other robot in order to flip it. The problem with this is that looking at the equations on the Ask Aaron post I knew whatever motor setup I chose to spin the central cam to actuate the flipper arm would need to provide some serious torque. Using the second video I linked as a baseline, I decided to go with a continuous rotation servo motor with an additional set of gearing to provide enough torque to properly spin the cam that actuated the flipper arm. The problem with this setup is that it was extremely heavy. With 2 large steel springs, a big servo, and some gearing, there was just not enough weight to effectively build a robot around the setup in order to actually fit within the 1lb weight limit. Because I could never get the robot under weight, the idea was filed away along with the many other failed ideas I had in the depths of my hard drive, to possibly be resurrected one day. Since then I had built another antweight, Hercules, and built many different versions of the robot, filling my desire for a lifting antweight robot. But the design always ended up leaving a little to be desired, as it won most of its fights by out-driving its opponents or surviving the multiple attacks. It is an interesting bot, but I knew I could build something bettern and more exciting. Again through internet research I came across Dale's Homemade robots and his robot Dead Air. This page is a great read for anyone interested one building a spring flipper of any level. Dale does some amazing things in his robots, including some pretty intense machining as well as building his own custom control boards for his robots. This page was incredible and inspired me to return to my dreams of building a spring powered flipper. I knew this was going to be a challenge. Dale in all his wisdom had struggled to meet weight for his spring powered flipper. I knew I would have to find some alternative solution that would help me to decrease my weight overall without the use of custom electronic boards that can monitor the current consumption of the motor to find the spot to hold the arm before firing. This did give me the idea to use a system to monitor the output though. In my original design, I had used a continuous rotation servo to turn the cam, completely defeating the point of using a servo. If I could find a way to monitor the cam angle, I could do the firing with the flip of a switch. I went down a long rabbit hole of searching through encoders and microcontrollers that would accomplish this continuous rotational feedback. But in the end, I decided the solution would be too fragile, too bulky, or above my knowledge at that point in the design process. Looking back now I could definitely accomplish this, and this likely would have been the route I chose had I not come up with an alternative solution. I came across another legend in creative combat robot designs, Peter Waller. He had designed a UK ant (150g) flipper using rubberbands as the power source. However, the use of rubber bands was not what interested me about this design. The flipper worked through a very interesting mechanism that only had the servos turn 90deg at most. This process is detailed in another Ask Aaron post here. This setup separated the loading and firing operations into two different motions from the same servo. I loved this idea, but I would have to figure out a way to scale it up into a 1lb class as rubber bands and 3D printed trigger mechanisms likely wouldn't work at that scale. Initial Flipper DesignBecause this design centered around the flipper mechanism, I knew I would have to design that piece first and design the rest of the robot around it. The design I had in my head was sort of a combination of all of the previous ideas I had seen to build the best possible flipper mechanism. I had tried to design something around a snail cam, but in order to get the range of travel I would want, the mechanism would take up way too much of the overall weight of the robot as the size of the servo and overall mechanism increases with the range of travel of the flipper arm. Instead I wanted to use a single, non constant rotation servo to both cock and fire the flipper mechanism, similar to Pete's design. I also loved the use of a leaf spring as a source of potential energy in Dale's design. Many of the early designs I created used a coil spring similar to the original Janna video I found. Many of the springs that would provide the proper level of energy for a flipper over the small travel of the mechanism would be very heavy as they needed to be very stiff. A leaf spring could provide the same energy storage in a much more compact and light-weight package. Finally, I liked the form factor of the original Janna robot I found that started this. I wanted to make a two wheeled wedge style robot with a central flipping mechanism in the middle that would have a servo on one side to actuate the mechanism. These three bots would lead my design path for this bot. My go to first step in a design process is to draw things out on quartile paper. Especially for things on this scale it is really useful to draw things at a 1:1 scale to get a feel for how big the components will be in the end. My idea behind the flipper mechanism is that the flipper arm itself would pivot at around 2/3 of the length such that the distance traveled by the cocking mechanism would be roughly half of that of the tip of the arm itself. The servo would have an arm attached to it that would be centered when the flipper is in the cocked position. This would allow for using a self centering channel on my receiver to activate it. When the servo arm moves backwards the cocked flipper arm is released and the flipper fires. When the servo arm moves forward, it pushes the flipper arm out of the way to be able to move to the underside of the arm. A rubber band would pull the flipper arm back down so the flipper arm would snap over the servo arm once the servo arm gets low enough. As the servo arm returns to the neutral position, it catches the back end of the flipper arm to cock it back into place. You can see a drawing of the mechanism idea below. From here I went on to make a CAD version of the sketch to be able to lock down the actual dimensions and verify the interactions between the moving parts. You can see the initial sketch below: Here you can get a better idea of how the two arms interact. The large triangle is a rough depiction of the body. There are two sketches of each arm: the flipper arm in the cocked and fired position and the servo arm in the cocked and completely forward position. You can see from the arcs drawn in that there should be enough clearance for the flipper arm to pass back over the servo arm so that the servo arm can catch and reset the flipper arm to the down position. It was at this time I realized that the positioning of the pivot for the flipper arm was almost exactly where I would want to put the drive motors for this robot. In order to fit the servo in place I planned on using a set of spur gears to transfer power from the servo to the trigger arm. I had considered this is some of the earlier designs as well in order to increase the output torque of the servo to actuate the original snail cam. I knew that Servo City sold servo gears, so that part was solved. The issue was the range of motion of the servo. By increasing the output torque of the servo, I would also limit the range of motion of the servo proportionally. To specify the servo and the gears I needed, I first had to decide on how much flipper power I wanted. This is a little tricky to directly figure out given the geometry, but I can do a couple of simplifications and over spec the servo from these to be sure that it has the required torque. First I needed to decide on a spring to determine the force required to bend the spring. Going off of Dale's robot, his leaf spring flipper had around 70 oz-in of torque when fully bent. This seemed to work pretty well so I figured I would design around this number and make later adjustments if necessary. In order to design the spring for this specific application I used a combination of statics of a cantilever beam deflection and comparing it to the maximum bending stress to the yield stress of the spring steel. The same thing can be accomplished using Dale's online calculators of beam deflection and flipper power. This gave me a starting point of a leaf spring with the following dimensions: From here I could figure out the power needed from the servo to actuate the flipper. Using the balance of torques around the main flipper arm pivot and the length of trigger arm I found the required torque of the servo to be around 96 oz-in of torque. This is a very conservative estimate using the maximum spring force and the full length of the moment arms. This is likely higher than the actual torque required for the trigger actuation, but is a good starting point to use. Using the HS225-MG servo as a baseline, the same servo I use in Hercules, I began to calculate the gear ratio I needed. The servo's stall torque at 6V is 66.6 oz-in. This means I would need a gear ratio of at least 1.44:1. Looking at the servo city gears I chose a 32T spur and 20T servo gear to create a 1.6:1 ratio for an output torque of 106.6 oz-in on the trigger arm. This servo has an out of the box max travel of 191 deg which, when put through the gear ratio, is reduced to 119 deg. This means a travel of only 59 deg per side which is less than the 83 deg required for the actuator to work. So I would have to find another servo. Looking on Servo City I found the HS-7245MH Servo which is a digital programmable servo. This servo has a max input voltage of 7.4V meaning I could run it directly off of a 2S LiPo battery. It has a significantly higher stall torque of 88.9 oz-in meaning the gear ratio can be reduced to at least 1.08. I switched to a 24T spur and 20T servo gear for a ratio of 1.2:1. This servo only has a 180 deg range meaning without modification the servo would have a 150 deg travel which is only 75 deg each side. This is still too small for flipper to actuate correctly. The difference with this servo is that it is programmable. one of the features that can be programmed is the neutral point of the travel of the servo. This means that I could make one side of the travel of the servo higher than the other side. I could split is so the trigger would travel more than the required 83 deg in one direction and still have enough travel on the other side of the neutral point to be able to fire the flipper. The drawings of the servo with gears and the trigger arm travel are pictured below: Component SelectionWith the flipper geometry sorted out, the next step in my typical design process is component selection. Generally, since I have competed in this weight class before, I roughly know what components I want to use. My typical strategy is just go with what I know works and import the models into CAD and design around that. If the weight is too big, start doing weight reduction until everything works. While this is a valid strategy that has lead me to design a lot of successful bots in the past, this project was going to be a little more challenging. I knew from my past versions I was going to have issues meeting the 1lb weight requirement. To help try to prevent this eventuality I decided to be a little more meticulous in my design process than usual and have a spreadsheet that would help aid me in the weight calculations and component selections. This picture represents an early version of the full BOM for the robot. I have seperated up the components based on their sub system. I have included the Riobotz combat robot tutorial recommended weight distributions for a balanced robot. My designs tend to be a little more durable than most so I tend to put more weight in the armor and structure than in the other subsystems. I imported a general set of components that I have used before (basically exactly what is in Hercules). I also added the selected servo and some initial guesses as to how much some of the other components will weigh. Overall, I was very happy with where this sat. As the design progresses, I planned update this spreadsheet with weights and prices to create a comprehensive BOM. But for now I was happy enough to continue on. I used a lot of tested components in this design. I know they will work well together out of experience. The problem is that me just showing you readers this doesn't necessarily help to design your own robot. I have already broken down how I selected the weapon components. But I wanted to quickly go over how I would select the drive system and battery. In general, this weight class is pretty limited on reliable drive motors. In my experience I have found the Fingertech silver sparks to be the most durable and versatile option for the price. There are other options that are tougher such as maxon motors, or cheaper options that can be found on eBay or other online retailers. But I found these motors to be very solid with decent performance for their size, weight, and cost. That being said, these motors are available in a wide range of gear ratios. You need to find the right combination of motor, gearbox, wheel size, and battery voltage to provide you with the right speed and pushing power. This can be a tough task. You need to take into account a lot of variables including operating voltage, motor constants, wheel diameter, stall current etc. There is a lot of math that goes into the selection of these motors. Fortunately for combat roboticists, Ask Aaron has developed a tool that helps to do all of this for you and actually provided some of the equations behind the tool if you click the "Gear Ratio Tips" button. This tool has a lot of the necessary values that are required to do these drivetrain calculations built in. It gives the user the key outputs based on the set up of your system to allow you to fine tune the speed of your robot and ensure the motors are running under stall. It also provides an estimate of how much current consumption these motors will use during a fight, however I have found this to be a very conservative estimate and generally at least double it when calculating battery capacity. Finally, there is an acceleration calculator that uses simple physics to tell you how quickly you will reach top speed and how fast you will cross the arena. I generally shoot for around two seconds. Below is the output of these tools for my selected setup: Initial LayoutWith all of the purchased components selected, it was time to start actual design process. Again, for this I generally start with a 1:1 sketch on graph paper of the layout of the robot as a whole. I knew the rough dimensions of the flipper module, so the design would have to be built around that. I knew the general layout and shape in my head, but I always try and draw it on paper first to get a general idea of how big things need to be. I started with a top view of the robot. This view allows me to make sure all of the internal components have enough space to fit inside of the robot. You can see most of the components that I listed in the BOM in this view. There seems to be plenty of space given how long the flipper unit is. The overall dimensions of this layout from this drawing are 7" x 6", which a little big for this weight class, but not too much bigger than my other antweight. I will just have to be careful to not make it too big for weight and durability purposes. Obviously this layout and dimensions aren't final, but help to provide me a good starting point. For example The flipper unit will likely be a little wider than it is here to give a little more clearance for the arm side to side. I will also likely move as many of the internal components towards the back as possible to help with weight distribution. This will also allow me to route the wires between the two sides easier as there is a gap between the servo and the flipper arm where I could make a wire channel. Just a few design features to note on this. For the drivetrain, the wheels stick out the back side of the robot to allow the robot to drive when on it rear to hopefully help self righting. I had also planned to print a few removable wheel guards to help protect the wheels and allow me to reconfigure the armor to help fit the opponent. There are some nasty horizontal spinners in this weight class that I will have to be prepared for. Another design consideration to help protect from spinners is the front wedge. I plan to have a set of sloped titanium plates in the front to help deflect hits from spinners. The plan right now is to have them angled both towards the front and the sides to act as a wedge and deflect hits from other robots. This might change in the future if I have more weight to put into the wedge to help better protect the rest of the body. CAD DesignWith the basic size and shape locked in, I began to create a 3D CAD model of the robot. This process had many different design challenges that I will detail in this section of the post. The 3D printed frame of the robot was going to be rather complex in order to take advantage of the 3D printing process as much as possible. I started with a general layout from the initial hand sketches. I then began adding internal components starting with the drive components. When all of the internal components were modeled and fit into the body of the robot I could begin designing the rest of the manufactured components of this design. The wedges, top plates, and motor mounts are all attached to the frame using heat set inset inserts for thermoplastics. All of the metal components were initially designed to be 2mm titanium to save on manufacturing costs. This was later changed for weight savings. The next major design step was to model and verify the custom leaf spring. This meant redoing all of the calculations several times to optimize the setup. Finally, the fasteners were all added and the BOM updated. Several design changes were made to make weight as can be seen in the album below: Body Design The main frame piece of the robot is designed to be a single piece printed from reinforced nylon. Because nylon doesn't hold threads well, all of the places where components are bolted to the frame are designed in a way to have heat-set threaded inserts. There are around 30 inserts in total as there are a lot of screws holding everything together. The outer walls of the body are .1" thick and the central walls where the flipper mechanism is are 3/16" thick. The threaded inserts need .310" of material to be fully inserted. In order to save weight, most of the threaded insert locations are simply circular extrusions to provide enough material to properly hold the inserts. The weapon servo and drive motors each have their own specific mounting points in the body. The servo has a platform that it sits on to space it properly relative to the rest of the flipper components. A custom 3D printed harness holds the servo in place. The drive motors are face mounted using the fingertech mounting plates. In order to support the back end of the motors, each motor has its own platform to sit on. The idea is to ziptie the drive motors to these platforms as a form of shock mounting in as small and light weight of a design as possible. Flipper Design The flipper has always been the biggest design challenge of the build. I did a lot of initial calculations to ensure that the final design would require as little tweaking as possible to fully integrate the flipper into the final design. Looking at the cross section above you can see that the overall design is relatively the same, with a few key changes. First, the servo gear was moved up off the base. This was done to ensure the gear had clearance from the base plate as well as re-position the servo relative to the other internal components. One of the major changes to the overall layout is that the spring base is angled slightly. This was done to optimize the spring geometry in order to keep the maximum stress during bending under the yield stress of the material. This required the calculations for the spring deflection, flipping, and servo power to be re-done and verified again. I struggled with this part as there is little readily available data on the actual yield stress of hardened spring steel as it varied greatly depending on heat treatment. The spring values shown above are again based on the values that Dale used as those provided a good flip without bending the leaf spring. I found the bending stress of his setup to be 120 ksi so I set that as my bending limit. Ideally the spring would provide a high force over a high range of motion to provide the best flip. The issue is the greater the spring force, the greater the bending stress. In order to compensate and lower the bending stress, the deflection has to be lowered. This optimization problem results in the ideal leaf spring being a long, wide plate in bending. This would be very heavy though. I ended up with the above geometry to meet the specifications of Dale's flipper as a starting point that could be adjusted in later versions. To verify that the same servo can still be used, the same analysis had to be done with the final setup. Using the same basic calculations as before: the spring force is 8 lbf at the flipper; the force on the other end of the flipper arms would be around 10.5 lbf; the trigger arm is 1.25" long meaning the torque required on the trigger arm is 13.1 in-lb (210 oz-in); with a 1.33 gear reduction that would mean the servo would have to produce 157 oz-in of torque. This is significantly higher than the 89 oz-in stall torque of the servo. However, this is an overestimate of the required torque as it assumes the maximum force perpendicular to the moment arm of the trigger. The trigger is designed in a way that there is no torque on the arm to maintain the loaded position when the spring force is at its highest. A more careful analysis of the forces involved would give an accurate representation of the maximum torque required to load the flipper. The first step I had to take was to update the initial flipper sketch. This would allow me to more easily see the interactions between the parts all at once. I assumed based off of the fact that the initial theoretical calculations showed a big difference between the servo torque and the torque required, I'd likely want to increase the gear ratio. With the rest of the components set I knew I'd have space to make the servo gear smaller, which also had the added benefit of decreasing the weight overall. Switching to a 20T servo gear and a 32T spur gear created a 1.6:1 reduction. Using the final flipper sketch below, I verified that with the increased reduction, the servo would still have enough travel. As you can see, the trigger has clearance on both the firing and reset sides even with the higher gear reduction. With the geometry settled it was time for the more challenging task of dynamic analysis. I debated how much of this to include on this post as it is a blog and most people don't come here to look at algebraic equations. I have included a photo of my calculations as well as a screen shot of the excel spreadsheet and output graph below for reference purposes. I won't go into too much detail on the actual mathematics and physics, but it is there if you wish to look at it. I ended up including it on here to show first how difficult of a problem designing a spring flipper is, but more importantly to help anyone in the future who uses this as a guide to make their own flipper robot. You can see from the graph above that the peak torque required of the trigger arm is around 127oz-in. With the 1.6:1 gear reduction, that would mean that the servo would require at least 79oz-in of torque to fully reload the mechanism. This is less than the stall torque of 89 oz-in of the servo selected, but very close to stall. This means the servo would likely struggle to quickly reload the mechanism ,especially with any frictional losses in the system. However, I was already starting to push the limit of what I could do geometrically to lower this torque while still operating within the usable output range of the servo. So i decided this was a good starting point, and that if the servo stalled out during testing, I could easily remedy this by trimming down the spring slightly to decrease the flipper force and thus the load on the trigger. Weight Reduction The final phase of the design process was weight reduction. As I mentioned earlier, this has always been a major struggle of this design. All of the previous versions of this design were significantly overweight. Finding the proper weight balance is difficult for flippers at any weight class because you have to store enough energy to launch your opponent without a spinning component. This doesn't leave a lot of weight for armor. Typically flippers have rather thin armor to make weight. The initial BOM I did showed that I would likely have an easier time making weight with this design than designs of the past, however this was still a major task. I had used roughly the same internal components on Hercules which made weight just fine. The major difference is that this design has a lot more metal in it as parts of the from wedge and flipper arm. Having a large steel spring didn't help either. One of the main questions that came up during this process was how to estimate the weight of the 3D printed body. Having the CAD model is helpful as I know the volume of the part, but when 3D printed, the part won't be solid. There are typically multiple wall layers printed around the perimeter and a set percentage of infill inside of that. In order to estimate this without actually printing the part, I created a formula to estimate 25% of the weight as solid to represent the walls of the print, and the remaining 75% based off of the infill percentage that could be adjusted to cut weight. The initial design pictured below weighed in around 1.25 lbs with fasteners. I knew I would have to take some drastic measures to make weight, but it was definitely doable. First, I cut the front wedges almost in half to take out a large amount of the titanium. I thinned out all of the titanium from 2mm to 1/16". I went through and thinned out a lot of the walls of the body to cut down on the weight. In addition I cut out a lot of the unnecessary material from under the wedges, motors, and servo. All of this was still not enough. I swapped the motors from silver sparks to Kitbots antweight motors that saved around a half an ounce total. I also swapped out the battery for a smaller capacity. I changed the material of the trigger and weapon shaft to aluminum as well. When the weight was close enough and a lot of the other weight reduction methods had been implemented, I started cutting down the percentage of infill on the body until the total weight was under 16oz. The final design and BOM with weights can be seen below. Construction and Initial TestingBecause I spent so much time meticulously planning this build instead of rushing through it like I have for some of my other builds, the ordering of parts went relatively smoothly as I had an updated BOM. Because this was a 3D printed unibody design I wanted to start this build with a test print of the body and make sure that all of the electronics and other purchased parts mated properly to the body. Because this was mostly for the form of the bot and not for durability I got a body printed by 3DHubs out of PLA relatively cheaply. The print came out great. Just like I drew it up in CAD. I'm always still impressed with how intricate 3D prints can be. I started by heat setting the threaded inserts into the body to mount the motors with. This all fit fine so the next major step was the electronics. Electronics Because this robot was split into two electronics bays by the part in the middle where the flipper goes, I had planned on running wires between the two halved through a little cutout in the central walls. You can see the wires running across in the last image in the album above. Because of this, I ended up adding JST connectors to the power lines of the ESCs so that the wires running across wouldn't have to be soldered together inside of the bot, especially since this would be a 3D printed chasis. I learned that lesson from all of my electronics failures with Barrel Roll. My original plan was to have the drive motors run on 3s LiPo and have the servo for the weapon motor run on 2s LiPo that I tapped from the balance plug of the battery. As a note, this is not recommended for the reason that this could potentially make the cells of the battery unbalanced from one another leading to decreased battery life or potential catastrophic battery failure. That being said, I knew the risks of the setup I was attempting to make and still planned on doing so because the servo has a relatively low current consumption in comparison to the drive motors, so the cells should not become too unbalanced. When I charge my batteries, especially the small ones, I balance charge them so that should fix the batteries between matches. If during testing they became so unbalanced that I was worried about them, I would resort to a backup plan of running the whole robot on 2s. You can see above my planned electronics setup. I initially created a wiring harness like the diagram on the left, but this only worked on paper. The concept makes sense looking at the diagram, but when I went to plug everything in the drive ESCs were still receiving power even though the switch was off. I found out that by unplugging the servo, the ESCs lost power. This lead me to conclude the ESC was getting power through the servo. I assumed it was because I had wired the servo to the first two cells so the "ground" for the servo was not at the same voltage as the ground for the ESCs. I went ahead and remade the wiring harness to look like the diagram on the right to follow my hypothesis only to find the same issue was present. After asking some fellow builders and not getting any clear responses on how to solve this issue, I decided to try and add a diode to the servo's power line to prevent the power from running back through the servo to power the ESCs. This worked to keep the ESCs off when the the switch was open. But caused issues with the motors during my test driving. The motors would occasionally keep running after I had stopped moving the stick on the controller. This is not good for both safety reasons and the fact that this would make the bot significantly harder to control in battle. At this point I had a couple of options to fix this issue. I could have added a second power switch to control the power to the servo to isolate the circuits and likely fix most of my issues. But because I was already so close on weight, I decided to go with my backup plan to run the entire system off of 2S LiPo. This would both fix my electrical issues and lose some weight from the robot. I would have a lower top speed but in drive testing this seemed to be just fine. Above you can see the final power harness for the robot. The positive lead of the battery goes through the switch to power all of the components. You can see all of the connectors here that connect to the battery, ESCs, and servo from top to bottom. I put mating JST connectors on the ESC as mentioned before for easy disassembly. The signal lead of the servo was separated in order to go to the receiver along with the servo leads of the ESCs. With everything wired I took a little drive test around my kitchen that seemed to go well With the electronics sorted, I could move on to other things. During assembly I noticed a few minor things that needed to be added to the print to make assembly easier. I also realized how big 6-32 screws looked on this robot. I want to keep them for the front wedge as there are only 3 holding the whole thing on. But I planned to replace the screws for the motor mounts and top cover with 4-40s to save weight. With the updated BOM below it was time to order the metal parts and test fit those and begin the flipper testing. Full Assembly After doing the test fit of the electronics and doing the final revisions on the design, I ordered the manufactured parts. The main 3D printed parts came from 3D Hubs and the metal pieces came from Send Cut Send. Both excellent resources for quick manufactured parts. 3D hubs does a little bit of everything, but I use them for their 3D printing services, specifically their carbon fiber reinforced nylon for most of my 3D printing services. Send Cut Send does instant quoting of laser cut pieces with a wide variety of material options. Most of these pieces were aluminum or titanium, but I also got some new AR500 weapon discs for Butcher that I will update on later. I've included some pictures of the pieces that were made below: I started by making the sub-assemblies for the bot. I installed heat set inserts into the body and made sure the servo fit in its designed location. I assembled the flipper arm and the flipper trigger seen below: I found my first design oversight when I tried to assemble the flipper. I didn't leave enough clearance between the top of the flipper arm and the boss that holds the screws for the plate that holds the leaf spring. If the arm is not perfectly centered on the bot, the arm contacts these and is unable to lower completely. See below: This is definitely something that needed to be addressed so this doesn't happen in combat. I planned to add some washers to the flipper shaft to minimize the possible horizontal movement of the arm along the shaft. But this still didn't leave enough spacing in my opinion. For testing purposes I trimmed the arms to add some extra clearance, but this would require a redesign to work properly. After cutting the tips off the arms (see below) and adding washers to the sides of the weapon arm shaft, the arm goes up and down much more smoothly. There's still not a lot of clearance there, but if there's enough force on the arm to bend the shorter end of the arm to the point where it would interfere, there's likely going to be clearance issues on the other end as well so I'm not too worried about it. I made a couple of leaf springs for the flipper out of 1095 spring steel. This stuff is no joke to machine. I bought a real sheet metal hole punch to cut the holes for the screws which just barely does the job and sounds like a gunshot every time it punches a hole in the spring steel. I'll likely look into alternative ways to retain the spring in the future so this is easier to do. But for now, I only punched two holes in there to save some weight and make it easier to get hole alignment to the spring mount. With the springs cut I could do some testing with the flipper. I still don't have the proper end tapped shafts I need to properly retain the trigger system, and I had been having trouble programming my servo as intended, so the testing had to be done by hand. In the video below I'm wearing safety glasses and only putting my hands where I can't get my fingers pinched. This flipper mechanism isn't quite as dangerous as a spinner, but safety is still a concern. I started my testing with just one leaf spring and it was rather lackluster. It somewhat flipped, but only barely better than a servo would. I did two things to help this: one I added another spring on top of the current one to increase the spring force. This definitely helped, not quite doubling the flip, but a definite improvement. I'm not sure if I'll have the weight for this on the final version, but I'm going to leave it for now. The second thing that I did was add spacers under the front of the spring mount to increase the relative angle between the base of the spring and the arm, increasing the deflection and the force at the bottom of the stroke. This helped pretty significantly, even with one spring. The initial math that I did says that this should plastically deform the spring so that they are bent downward, but there doesn't seem to be any deformation so I plan to keep this until the next version of the body where I can change the angle. Here is a video of the flipper in action. As you can see, the arm doesn't go down all the way. I would have to either shim or reprint the central spacer to the flipper arm so that the interaction with the trigger fully lowers the arm in the loaded position. This would end up being a part of the flipper arm redesign. After working my way through the servo programming software, I was able to do a proper test of the flipper mechanism. This is the servo programmer that I used to adjust the end points of my servo. It worked after I figured out the confusing GUI and very specific plug in process. This is an option, but if you're reading this and considering re-programming your servos, I would recommend looking for a better servo programmer if there is one out there. The video of the test below shoes the mechanism working mostly as intended, but there were some issues. If you notice there is only one spring and no spacers on the spring mounting plate. This is because the servo was to weak to handle any more spring force and still operate as intended. This lead to a very lackluster "flip" that didn't do anything but lift the robot quickly, not flipping it over as intended. This isn't in the video, but trust me you didn't miss anything. After seeing this, I decided that I would have to go back and redesign the flipper mechanism to have enough power to flip something the way it was intended. This meant redesigning most of the robot. This wasn't really a concern though as the event that I was intending to attend with this bot was cancelled for COVID-19 so I have plenty of time to redesign and refine this bot so it worked for whatever event it eventually attends. Additionally there were a lot of small changes that I wanted to implement into the design that I found out in the construction of this version. Drop Test 1.1Since I've already run through the details of this design process earlier in this post, I'll run through it quickly here and highlight the major improvements that I plan to implement in the next version. This starts with increasing the power of the flipper. From the video above you can see the geometry works really well. The only problem was that the arm didn't go quite low enough to clear the fixed wedge. This is a simple fix by shimming the flipper arm spacer a little when I reprint the center piece. The main problem was the spring power. Doing the tests it seems like 2 springs provided a really good flip. So I used that as a baseline. Having 2 springs would be heavy, but I wanted to try and have that option If I could to maximize flipping power. I also noticed that the spring was no where near plastically deforming as my initial math would suggest. So I decided to increase the angle between the spring mount and the flipper arm when loaded. This would further increase the flipping power, potentially to the point where I could still get a really good flip with only 1 spring. These two screen shots show the updated geometry outputs from the two calculators on Dale's website. By changing the angle, I could increase the deflection of the spring, and thus the force of the spring from 4.8 lbf last time to 11 lbf this time, over double the previous version. This meant that I could theoretically get 2x the flipping power with only one spring, or 4x power with two springs if I have the weight. The second photo shows the results of the new dual spring setup, showing the max flipper height would be around 1.5 feet in the air which is really impressive at this scale. With this as the lofty goal I was trying to achieve I would need to first find a motor capable of providing enough torque to load the flipper mechanism. The servo I bought was the highest torque servo I could find in that weight range. If I wanted to increase flipper power and stay using a servo, I would have to do massive weight reduction on the bot that is already very weight minimized. As a compromise, I decided to move from a servo to a planetary gear motor. This would allow for a significant increase in trigger torque for around the same weight, at the loss of positional knowledge. Instead of having the servo respond to the position of the stick, I can just control the gear motor speed and set it to not move when the trigger is loaded. This might be tricky to do manually, but as a back up plan I would harvest the board inside of the servo I bought and mount the potentiometer on the trigger shaft, creating a pseudo servo of my own. This is a much more complex solution that I'd want to avoid if possible. The servo that is in the previous version, has a stall torque of 88 oz-in. Using that as my baseline, if I wanted to increase the force up to 4.6 times as much as the servo version, I'd need a motor with a stall torque of at least 400+ oz-in of torque. Servo city has mini spur gear motors that are perfect for this. I wasn't sure on how fast I wanted this to turn, so I bought a couple different ratios that would all fit the torque needed. Trying to fit the gearmotor in the bot, it was clear I'd still need some geartrain between the motor and the trigger shaft to offset the two for packaging. Using the gears available on ServoCity, the lightest gear train I could get to fit had a 2:1 ratio. This means the gearmotor torque requirement would be cut in half and I would have a large range of speeds to select from as most of the series of motors would provide enough torque and all weigh roughly the same amount. The next step was making sure everything still fit in the 1lb weight limit. Below is the updated weight spreadsheet. The biggest change is that in order to make weight for the bigger gear motor, the wedges had to be thinned out to 1mm from 1/16" (1.6 mm) before. This is a little thinner than I'd want it to be but with how cheap the spares are, they could be considered more expendable and be replaced after a few battles. Everything is basically the same size and layout. I don't have enough weight for two springs, but the single spring should be plenty of flipping power. It will be close to the weight limit, but I should be able to make it work. There were some other minor improvements that I made to the bot. In my testing I realized that I needed to put a hard stop to limit the travel of the flipper arm so that it doesn't continue to swing until it wedges itself against the base plate. To limit this, I added some holes to glue a thin piece of piano wire in the bot across the flipper assembly to stop the arm from over swinging. The other part of this is that the flipper arm needs to have some spring return to push it back down after the trigger travels underneath it to grab the underside. The central spacer pushes the leaf spring back up as the trigger rotates down. This will spring load the arm down to limit the amount of time the arm is in an upright position and vulnerable. That mostly covers the general design work for 1.1. There were some additional weight reductions added, but nothing worth addressing. Below is a picture of the final 1.1 CAD v1.1 BuildIn order to keep this post shorter, I'm going to only lightly go into the build log for v1.1 of Drop Test. Most of the build is the same as the first version, with a few minor changes. I've included a quick gallery below of the build including a few closeups of some of the changes Knowing that I could get withing the 16oz weight limit, I could revisit some compromises I made in the design process to see if I could make weight for some of these small improvements. These included extra top armor and a bigger battery. The top armor I had was 1/32" 3D printed NylonX, which is paper thin and really only serves the purpose of holding the electronics in and doesn't really provide protection from anything. I had some 1/16" UHMW that I had Lying around that would both be slightly improved top armor and be semi translucent so the lights on the ESCs would serve as the power indicator lights. These would be heavier though and I would need to make weight for them somewhere as I was at the limit already. 4 grams is a small price to pay to have something reasonable for top armor. It's not going to stop any direct hits, but the risk of spilling the electronics everywhere after a hit has decreased significantly. Those 4 grams had to come from somewhere though. The first place I started with was the weapon gears. These gears are pretty hefty, being made of brass, and likely overkill for their given application. Together they weigh in at 32g which is 7% of the robot's total weight which is a lot given that the top armor weighs only 13g all together. There was already some weight reduction done to the gears by the manufacturer, but I figured I could drill a couple of holes in the big one at least and get back a few grams. Drilling holes into the big gear saved 2g from the total weight, which is less than I'd hoped for but still something. I had a radical backup plan of switching the gears to plastic instead of brass which would have saved 26g which is almost a full ounce of weight. But after looking at the gears, I decided against it as I'd be concerned about their durability in combat. I'd have to find the weight elsewhere. I definitely plan to revisit this in the future, probably replace the gears with a timing belt. Now that the driving gear doesn't have to connect to a servo, I can same some of that weight and give move space for the motors by swapping the gears for a timing belt. The other item on my list to add was a bigger battery. I had originally wanted to use a 3s 11.1V battery for the speed in the drive motors, but the servo I had was only rated to 2s and I didn't really have weight for a BEC to reduce the voltage to a reasonable level. So I had planned to use a 2s battery for everything. But now that the servo is out, the new gearmotor can handle up to 18V. So I can run the whole bot off 3s without a problem. I had received a new TinyESC from Fingertech and installed it, but the 65RPM motor I had in there just didn't seem to have enough power to reliably fire the flipper. As such, I ordered some of the 33RPM version that theoretically have double the torque. Running these motors on 3s is enough to flip really consistently, but the motor itself doesn't seem to have significantly more torque than the 65RPM version. Being that I need to find a few final grams to shave off, I need to test if the 65 RPM motor would be enough on 3s as the 33 RPM motor is slightly heavier. This would also allow the weapon to reload more quickly which would be very helpful. I have a sampling of all of the small batteries that fingertech sells in 2s and 3s for this project to test. Ideally I could use a smaller 3s battery that has lower capacity than the 2s battery I was planning to use as the higher voltage would mean the robot would use less amperage as the power consumption is based off the combination of voltage and amperage. I will have to test the bot out to see if the 3s 180 mAh pack I plan to use is enough to get the bot through a full 3 min fight. With all of these small changes the bot comes in around 6g over weight. I think I can pretty easily make up this weight. I still need to grind down the wedges. I think I can also make up a decent amount of weight by tidying up the wires and trimming a few other things. I will call it good enough for now and make the changes after I trim down the metal pieces if necessary. That being said I am going to call this project done as far as this blog goes. None of the changes I make from here on out will change any of the looks or function. Just minor adjustments to make weight. For now, until I can find an event to go to, here's some photos and a test video. |
AuthorMy name is Michael Connerton. I'm a mechatronics engineer at a robotic automation startup company. This blog serves as a record of my exploits into the world of engineering, robotics and especially combat robotics. Archives
April 2023
Categories |