The chronicles of Luchador 1 through 3

As I’m currently returning to Luchador for BBB Champs, figured it about time to get a build log posted. :smiley:

So MK1:

So as some of you know I started building Ants in lockdown, then after my first event The Great British Bodge Off - I was hooked. The next BBB event was beetles so naturally I had to pull together a beetle for it!

I quickly decided to make my life harder than needed by pursuing the noble art of the suplex :mechanical_arm:
And thus Luchador mk1 was born. Inspired largely by the snappy build video it was built largely from 10/6mm HDPE - cut with a jigsaw, heat bent and wood screwed together.

I decided to use 3D printed gears to allow me to iterate a ton and powered the lifter with a 20kgcm continuous rotation servo (largely because I didnt know any better). This proved to be a mistake as the lifter simply lacked the torque to do proper grab and lifts/suplexes :frowning:

I did manage this little lift on RIP though :smiley:

mk1-lift

After this lack of lifting beans I switched over to a pololu 37D gear motor - which had ample torque and could lift up to about 4kg reliably! It performed well in this config at robonerd that year and my hopes where high going into Champs.

Then I drew Endboss rnd1 and well… this happened :joy:

Amazingly despite getting blown to pieces Luchador never lost mobility during that fight or the rest of that event! - but it did rapidly devolve into some sort of eldritch abomination as I frantically threw it back together after each fight


All said Luchador (or what was left of him) ended up - somehow - making the round of 16! Knocking out Saw Loser in the process - in a fight that ended in a moment entirely created by my excellent driving and design and that had absolutely nothing to do with blind luck at all…

At the end of this event Luchador was incredibly unwell and sadly had to be humanely euthanised and dumped into a box never to be looked at again.

5 Likes

After Champs I knew I had some work to do. Clearly I needed to build more tanky to stand up to the nastier spinners at Beetleweight scale. So having hidden the shame that was the mk1 remains in a box at the back of a dark cupboard - I started the ground up redesign work on my next beetle - naturally called: Luchador mk2. Going into the design process my goals were: to make something super robust, more compact with a lower CoM, ideally 4wd and that could still do the suplexy thing.

I had yet to get access to any CNC facilities so this was another gargage build using the same jigsaw, heatbending and woodscrew methodology as mk1 - using printed paper templates of my profiles and sticking them to the HDPE and cutting round the lines etc.

The process went ok, and I was a bit surprised at just how much smaller it was IRL

I was aiming for it debut being Brawl - however covid isolation forced me to drop out! :frowning:
So instead it ended up debuting at Scouse Showdown

Electronically though it was practically identical to Luchador mk1. The same Pololu 37D lifter motor, same pair of BBB 22mm brushed drive motors, same RX and Bec etc.
I was trying a lot of new things on this bot mechanically however. Custom cast PU wheels, 4WD with belts, switched to steel gears on the lifter etc.

Most of which seemed to work and Luchador mk2 proved to be a super reliable and robust bot. It took an absolute beating and kept coming back for more! Also got a couple of nice suplexes in - including this great moment against Angel Slice - where I learnt that suplexing spun up horizontals is a dangerous endeavour :stuck_out_tongue:

After the dust settled Luchador limped away from liverpool with many scars and a 4th place trophy!

My best result to date! - so naturally I did what any sane and rationale person would do…
I retired the bot after this single event and started work on Mk3! :joy:

8 Likes

So after the great success and the single event lifespan of mk2 I designed Luchador mk3. Well actually I started designing mk3 long before even running mk2 in an event. Such was my appetite for iteration at the time :sweat_smile:

I went into this design with a bunch of goals, including:

  • stronger drive,
  • the ability to reliably drive mid-lift,
  • better geometry for dealing with verts
  • outriggers/longer forks to counter lever my lifts and avoid faceplanting

The end result was this:

The chassis construction shifted to full CNC routed - thanks to now having access to such facilities.
The core chassis is held together with barrel nuts and the wheel guards are attached primarily via 6mm aluminium stand offs.

The U-shaped chassis and use of timing belts, let me get some small wheels (35mm diameter) right up front to help with lifting and driving. It also - in combination with my fork set up - made it easier to engage with verts without them getting engagement.

Unfortunately, it couldn’t self right with the double grabber arms, so I quickly revised it back to just the single grabber like on previous iterations.

The internals at this point were made up of 2 diy rotalink 1806 conversions and the same 37mm brushed motor from previous versions. It was a right squeeze internally and my initial build had too much slack on the drive belts so I used one of the stand offs as a temp/bodged belt tensioner:

I ran it in this basic configuration for a number of events - Rapture 2022, subterranean 2022 - then in the run up to Champs 2022 I made some further tweaks…

For a long time I’d been looking for a viable replacement to the Pololu 37D weapon motor - while having tones of torque and being relatively cheap they’re also slow and very heavy. Grab and Lift is a tricky beast and the torque requirements are significantly higher than on regular lifters. I had been eyeing up some of the hefty high torque (but incredibly expensive) 4S compatible servos from the world of RC crawlers for a long time - but was reluctant due to the cost and unproven (in combat) nature of these parts. However after a cheeky bonus at work and learning that Ratfish over in the NHRL was running one of the models I had been eyeing up - I took the plunge and converted Luchador’s weapon over to a SHV800 Servo - running directly off 4S and rated to 93kgcm the thing is a monster.

I did some testing with the SHV800 servo prior to designing the needed changes, and oh boy was I impressed. It’ll easily lift and hold a 1.5kg mass on the end of a 20cm lever no problems. During these initial tests of slamming a 1.5kg weight around - I did have a brief scare when it seemed like I killed it - however on closer inspection the servo had stripped out the teeth in the aluminium servo horn! Discovering that the servo horn was a known weak link was a huge relief - having a cheap disposable point of failure could/should serve as a great way to protect that expensive servo.

When it came to implementing the new servo I decided to drop the gears and instead move to a linkage system similar to my antweight Luchador jnr. And this was the result:

Luchador mk3 had a tough draw at Champs 2022 and was drawn against End boss round 1! - a rematch of the previous year’s rnd 1 murdering of mk1!
Mk3 faired much better - lasting the full length of the match and miraculously even managed a lucky clutch win in the final seconds. Just going the full length of the match felt cathartic and really gave a sense of closure to the evolution of Luchador and it becoming a far more robust and capable machine compared to mk1!

Luchador’s next match was against Fang. I tried out a wedge anti-horizontal config for this but fang was running some serious forks either side of their spinner and I simply didn’t have an answer to that set up combination. I did my best to stay on top of Fang, but eventually they got a hit in that managed to sheer one of my motor shafts. Crab walking and with no way to win the ground game - fang eventually pitted Luchador with ease.

So whats next for Luchador? - I’m making very few changes for Champs 2023. I’m making a batch of fresh wheels and I’m going to be running slightly longer tips on the lifter in the hope that will let me use my weapon a bit more frequently. Thats it really.

9 Likes

So Luchador went to BBB Beetle Champs recently and… it went rather well!

Like 1st place well! :open_mouth:

With the major spoiler out the way lets talk about the beetle for a bit.

I’m not going to do a full play-by-play description of my fights - because frankly what is the point!? - video will do that a million times better than any waffly words could, so if you want to see my run/fights checkout my event edit on my team page here: https://fb.watch/otsZw6abmc/

Instead of describing the fights I’ll concentrate on talking about the robot itself, my intentions and insights that watching them might not give!

Overview

Luchador last competed at Champs 2022 and it remained largely unchanged going into this event- it was the same fundamental design and the same actual chassis from last year. This lack of major overhaul did allow me to focus on smaller quality of life tweaks - I revised the mounting for the forks to be easier to change, added a nice link door, made a new horizontal set up, new internal motor guards - those sort of things.

This was a strange experience for me - I’ve never really returned to an old bot before - always iterating pretty intensively and fielding a new version most events. During the event prep I lost count of the times I found myself muttering “what idiot designed this!?” :joy:

As the event approached I went in feeling reasonably confident that the bot would perform well - this version of Luchador was the best its ever been - I had no doubt about that.

I had a bunch of tough fights - some against scary spinners and some we’re just great driving matches - but the bot performed great through out. My TX less so - but we’ll get to that.

Rnd1 3 way: Luchador v Home Improvement v Brugh Deluxe


This was an interesting match up - I knew Home Improvement would be one to watch - Jon-Luke consistently builds really interesting and tough bots. I knew I would have the ground game advantage - so my tactic would be to use that and try and keep them off balance as best I could. This worked initially, however eventually Home Improvement got a hit on me while I was inverted - peeling up a corner of my baseplate which intermittently caused high centering and drive problems - the final nail in this coffin came from me reversing over a seam which caught the peeled baseplate and made it peel further enough that I just had no drive what so ever.

A loss for Luchador and into redemptions we go

Redemption: Luchador v Blast Furnace.

I’m a big fan of Blast Furnace - anyone who double weapons is a king imho. Going into this fight - the lack of destructive weapons in the arena mean I knew it would be a simple driving match - the ideal scenario for Luchador really. This proved to be true and it was great fun and thankfully Luchador came out on top

Round of 32: Luchador vs Sniper

Oh boi what can I say about this fight. I pushed this fight back twice due to what I thought was mystery RX gremlins - Luchador was behaving very strangely during arm up - ignoring stick commands, moving on its own - very strange. Later on after this fight, I found out it was my TX and got Luchador set up on a mates Zorro for the rest of the comp ( Big :heart: to Jed for this), but on the third attempt at starting the fight I decided to just send it - thankfully I regained control partway into the fight and didnt loose it again until I placed my TX on the arena post fight! - could easily have gone differently!

I was trying out my new anti-horizontal TPU booties/pontoons for this match up, a bit of a unknown and against Dylan’s very scary Sniper (side note: I love this build so much Dylan!). They performed great, tanking hits, flexing to dissipate KE and ablating nicely. This allowed me to tank those first hits and perform quick follow ups - for control and dictate the pace of the fight. Eventually getting a super solid grab and lift on Sniper and suplex-ing him OOTA!

Round of 16: Luchador V Abracagrabra


For the round of 16 I faced Becca and Kat with Abra and the mini bot. I love Abra so much - its such a solid control bot and Becca drives it spectacularly!

Another great driving match with us both getting lifts, pins and shunts on each other. Big shout out to Kat on the mini bot tho - interrupting my only solid Grab N Lift, repeatedly high centering me and generally causing me so much nuisance that I caught myself focusing on the mini bot far more than I should have been!

Eventually I managed to get Abra into the pit and marched on to the rnd of 8.

Round of 8: Luchador v Jackhammer


The fight with Jackhammer was a strong match up for Luchador. Despite only having 3mm HDPE lids, Luchador has its whole lifter protecting anything important- leaving only its link in the drive pods as a vulnerable target for hammersaws really.

They had some long forks that I knew I would need to get around the side of - which I managed to do reasonably regularly. Even got a great moment where I threw my lifter up to parry their hammer saw blow!

Eventually they got inverted and couldnt self right - on to the round of 4! - at this point I dont think I really knew wtf was going on.

Round of 4: Luchador vs EVA02


Into the round of 4 I drew EVA02. As with jackhammer, I wasnt too concerned about the weapon due to Luchador’s design. We also had similarly aggressive front end fork game - so I new this would be a real driving match and hoped that my lifter would make the difference.

Eva managed to get pretty close to my link with one of their shots - If I hadnt added the link door would they have gotten it? :thinking:

Eventually I got EVA into the pit and secured my place in the final! - wtf I was fully in shell shocked mode by this point.

Also BIG UP the builder here - second beetle, second event and they’re running this deep into the comp - crazy good!

FINALS: Luchador vs Gizmo

So in the run up to this fight I think my internal monologue was roughly:
“AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAHHH”

My goal here was to keep my front towards Ellis as much as possible and hope/trust my front end to do its job. To try and get Ellis inverted and keep him there if possible - which didnt really happen.

Toward the tail end of the fight I lost a drive side - post event tear down showed this to be the motor coming off the gearbox. I had serviced the gearboxes a couple of weeks prior to the event and neglected to Loctite them. In the few days running up to an event I usually do a full strip down and rebuild with loctite - The Great Loctitening I call it. I didnt loctite during servicing because of this - then come event week I decided to not do it because “the only thing that really needs it is the drive and of course I would have loctited during servicing” - WHOOPS :sweat_smile:

I did get a few flips and a suplex in there and held out long enough to get the perfect flip that made Ellis land on his weapon and bounce into the OOTA and take the win! - a super lucky shot/flip!

So what is next for Luchador
I do have some design tweaks to the drive I’d like to do. I would also like to revise the baseplate so it is inset away from the edges of the robot to avoid a repeat of the rnd1 peeling - but I do also have an entire spare chassis’ worth of plastics so will probably hold off on actioning those changes.

instead Luchador will get rebuilt “as-is” with the parts I already have and then take some time off while I focus on TCD and a new thing for a little bit.

As always a big thanks to Joe and Craig at BBB for hosting a great event, David for his tireless efforts running and maintaining the arena and all the volunteer crew for helping make the day happen! And another big thank you for all my opponents and fellow roboteers for attending and being a generally great community!

8 Likes

I have loved seeing Luchador evolve and it has been a big inspiration for things I’ve done with Déjà Two. Seeing its performance at Champs was great to see, and getting the grab-and-lift’s holy grail of a suplex out of the arena was awesome to boot!

1 Like

Thanks Thomas! - that means a lot coming from you, all your bots are wonderful technical marvels. :purple_heart:

1 Like

Congratulations to you on a fantastic victory! Showing the spinners a thing or two :grin:

Your driving was fantastic, and your weapon broke the spinner meta (especially in the final!)

It certainly wasn’t an easy matchup for Home Improvement early on, you turned it over multiple times (a real test of the spinner’s durability, I’m very lucky it stayed in one piece)

Good luck with the repairs and your future plans :slight_smile:

1 Like

Home Improvement’s ability to take being flipped repeatedly and the ensuing bouncing and still self right and recover without loosing a belt or weapon functionality some other was really impressive! - especially for a new bot!

And the damage you did to Stratus was really impressive too!

Big fan of all your bots :grin:

1 Like

Hey man, love this bot!

Where did you get the servo which you used?

Cheers!

It’s a Holmes Hobbies SHV800 but I think they’re discontinued now.

Not discontinued but bloody expensive!

Oh looks like they’ve actually dropped in price.

But yeah high powered servo are expensive - it’s one of the various reasons I typically do not recommend them for grab and lift.

Torque requirements for grab and lift are drastically higher than for a conventional lifter and the limited range of motion to actuate both the grab and lift portions of the weapon motion makes down gearing them externally for more torque almost impossible too.

I would generally recommend a pololus 37D gear motor and gear driven approach for a grab and lift build, especially a first attempt.

While it’s true that Luchador does none of this - being a servo and linkage driven design these days - but it is also the culmination of dozens of rounds of iteration across 3 complete redesigns.

A gearmotor however allows for unlimited motor revolutions and when combined with external gearing to operate the weapon allows for both complete freedom to adjust the speed and torque via gear ratios and predictable consistent torque output.

So I’m probably gonna edit this post a few times as I gather pics etc and it’s likely to be a bit of an essay…

Ive been a bit reluctant to write up this post - because it’s mostly bad stuff but i think for a build thread to really have value it’s important to talk about the failures as much as the wins.

So the big one is: the spenny servo (shv800) is dead. Died in Runt my 750g at the BBB pub sportsman beetle event back in May - stalling out trying to flip the massive lobster was the final nail but I’m certain it was an accumulative thing of numerous shock loads.

Upon stripping it down I discovered that the gearbox was fine, perfect even. Not a single damaged gear tooth to be seen. This surprised me as I was expecting this to be the failure point of the servo and actually had a full rebuild kit on hand. What I discovered was much worse, the motor was spinning but the pinion wasn’t - but the pinion wasn’t slipping on the shaft - it was the shaft slipping in the mag ring of the tiny brushless inrunner that powers it! - not something I could repair and in my attempts to do so I probably did more harm than good. A write off really :sob:

I went into buying the servo fully aware that despite the exorbitant cost it might well die - but had been eyeing them up so long that I had to try it - that power to weight ratio was just too appealing.

So, with this happening at BBB pub beetles - this was just a few weeks prior to rapture, where Luchador was competing :grimacing:
A quick pivot was needed and Henry Strang gave me just what I thought I needed - a 2812 paired with a 100:1 reduction gearbox

A frantic rejigging of internal layout and some machining of hdpe gears later and I had this - the last version of Luchador mk3.

Being a gear motor switching back to gears instead of a linkage set up felt like the obvious decision and chonky hdpe ones felt like the play based on others recent experiences (grab crab especially)

I machined a motor mount out of a small scrap of 20mm HDPE and then printed a surround to support the rest of the gear motor and keyed that into my rear bulkhead. I also made the motor mount loop round the lifter shaft to constrain them together - and prevent flex induced gear slipping


I used an on brand thatchers guard to keep wires out of the gears because my internals are hell :joy:

Once together I was able to start testing and quickly discovered that while powerful AF this brushless weapon lacked the torque at slower speeds to do slow controlled lifts😭 It could however, get impressive suplexes with ease (that classic sensor less brushless rpm x power curve at play) .
Not the ideal for what I want from Luchador but very acceptable for a last minute pivot.

So off to rapture it went… But it did not go smoothly :sweat_smile:

Turns out it’s easy to carefully manage a lifter and avoid stalling it out in calm testing - but less so in the frantic action of an actual fight. Come rapture and Luchador basically killed a gearr box in each fight. Mostly by being overdriven past the weapons range of motion and stalling out I beleieve.


The main offender for me was the joint between the output shaft and the stage1 carrier plate - these were press fitted and were coming loose under all the torque loads :smiling_face_with_tear:

I hadn’t really planned for a complete failure of the gear motors and so drastic alterations were needed to keep Luchador in the fray… After eventually giving up on the weapon entirely resist-ador was born.


(I had hoped it’d be invertible without the weapon uprights, but I was wrong. Despite this resist-ador got a convincing win against paradigm spin)

Shakey then lent me one of Boring bot’s old servo hats and I quickly set about scrap heap challenging together some form of lifter arm to finish out the event with.

Conclusion
So overall a pretty rocky event and iteration of Luchador. These big reduction brushless gearmotors need some real care taken to stop them just imploding themselves - and I’m now of the general opinion that 1/10th scale RC servos just arent robust enough to take the kinda torque and shock loads I want to throw at them. I finished up this event pretty bummed out tbh - and feeling like Luchador probably needed a pretty fundamental rethink moving forward…

so TLDR:
Dont spend 140 quid on a servo - it’ll die
Dont stall out a high reduction brushless gear motor - it’ll die.
Luchador mk3 has probably ran its course.

6 Likes

Just something I’d mention. Got some different 28mm boxes, different shaft (and I’d assume carrier plate?). David ran them in Heratoss at Dojo and have seemed to held up fine. Will be building something with it myself at some point, but might be an option to save machining titanium shafts every gearbox :stuck_out_tongue:

2 Likes

Luchador is using the 24mm boxes. Id prefer to stick to that form factor for now - for reasons I’ll get to in subsequent posts.

I don’t want to spoil the retrospective narrative journey of Luchador induced metal health crisis’ that’s to follow :rofl:

3 Likes

Ohhhhh, 24mm. Ignore me completely then :joy:.

Awaiting mental crisis report, looking forward to it!

1 Like

Chuck us a link when you get the chance xoxo

So, with the Holmes Hobbies servo dead and the 24mm brushless gearmotor alternativehaving self imploded repeatedly What was I gonna do from here?

I considered my options and the big factors at play for me were:

  • I didnt want to loose any of the speed that Luchador’s weapon had gained since moving away from 37mms - the quick self-rights and quick little flips had been crucial to its sucess at champs 2023 - so that ruled them out.
  • For reliable grab and lift I need a huge amount of torque - 70kg+ at a guess from my experience.
  • The positional control when it was servo driven was a real game changer - so I wanted to retain that
  • I did not want another incredibly expensive unit that I couldnt fix/repair
  • I wanted something that I felt could take the shock loads I throw at Luchador’s weapon without failing and I’m kinda convinced that doesnt exist in a 1/10th scale servo

I looked at around at what was available and found nothing that quite fitted what I wanted… so I did what any sensible roboteer would do in this situation…

97wcqc

I decided to make my own custom servo, driven by the 2812/100:1 24mm brushless gear motor with a bunch of additional reduction and an Arduino/potentiometer to monitor output position and run the PID control loop.

First issue was to fix the gearmotor and hopefully prevent it from failing in the same carrier plate/output shaft slip way again. So I got a single piece output shaft and carrier plate turned from Gr5 Ti rod. This should be much more robust than the stock one and combined with positional control eliminating the risk of overdriving should sort out the gearbox issues

With that sorted next I needed to design the bloody thing…

Using an Arduino/Pot to turn a gearmotor into a servo is something I had actually explored waaaaaay back in the days of Luchador mk1 - when I experimented with adding positional control to a Pololu 37D.
https://fb.watch/vrj3roM2ht/
I didn’t pursue those experiments back then - beyond loose prototyping stages - because I felt that adding that level of complexity to a combat robot was asking for trouble, and just adding additional potential failure points.

But while past me was wise - present me was full of hubris, so onwards to designing I went - oblivious to the calamity that would follow.

Initially I wanted to make the servo a completely self contained unit with all the mechanical parts, the arduino, pot, bec etc all contained in a single package.

My first round of design work ended up with a result that looked like this:
firstdraft
Not bad! - this has the gear motor in the foreground with a spur gear internally on its output shaft, driving the servo output via 2:1 spur reduction. On the output shaft is a repeat hub and a large printed TPU horn for a bit of shock insulation.

I felt fairly pleased with this, until… I actually tried dropping it into the luchador chassis cad:


It be a Heckin Chonker.

So after a bit more iteration and attempts at condensing it down a bit, I eventually settled on using a ring gear integrated into the output horn. This allowed me to greatly reduce the horizontal size of it. I then settled on pulling the arduino and other electronics out of the package and installing them separately to make fitting it all into a robot more feasible. After all of this it ended up looking like this:

And here is a cross section of the output ring, you can see how it keys into the HDPE shoulders that then sit in the HDPE cups either side.

The output ring and shoulders
outputring

And the motor side cup, output ring and pot side shoulder.

While still fairly large this looked far more reasonable to try and fit in a robot when I dropped it into mk3’s chassis.

So I set about making it. The gearbox support was printed, while the output/ring gear was 4mm lasered hardox, and its shoulders and the cups they sat in were cnc’d from thick HDPE.

Assembly went pretty smoothly and I was able to move onto experiments in code !


I set up a test jig and wired it all up to an Arduino mega - although my plan for the robot was to use a Arduino pro mini, they do not feature a USB port… So iterating on the mega would be much more convenient…

Tune in next episode for: “Jacks foray into code, WTF IS A PID!? And will it even work?”

3 Likes

I love that hardox ring gear design! I have tried the DIY servo thing a couple of times too - Initially I used a worm gear assembly but I wasn’t happy with it. I switched to a hobby servo but it needs to be a low profile model for packaging reasons, and there’s precisely one that runs happily on 3s, so I’ve been working on a retrofit drive board that I can drop into a cheap low profile servo along with an upgraded motor.

Also I know this is very retrospective and I’m not sure it’s still helpful, but I would recommend the Arduino Nano instead of the pro mini, they’re about the same price and size and you get USB programming with the deal.

1 Like

Worm gear would be an interesting approach.

The lack of back drivability would be good for holding position but I suspect harder on all the components.

Yeah, my posts ATM are very much retrospective. But as for the Arduino pro mini, I already had them on hand from my previous experiments and they’re proven in Sproing - which was part of my decision making when I initially purchased.

Will bear the nanos in mind for future tho :ok_hand: