Crumbs! (And Friends) - Team Manuva's Antweight Build Log

Printable View

18th April 2020, 19:27
Mr Manuva27

10 Attachment(s)

Crumbs! (And Friends) - Team Manuva's Antweight Build Log
Hello!

Building an antweight was always on the agenda for me, but it was put on the backburner due to upgrading Bread Dead Redemption between events. I've found myself with a bit of time on my hands and a 3D printer, so it's time to start on my new antweight!

Introducing.... Crumbs!

I've titled the build log "(and Friends)" because most events allow multiple robots and I will probably attempt more ambitious antweight projects in the near future to improve my CAD skills, so any other robots will be lumped in here too.

Specs

So Crumbs consists of the following:
- 3D printed chassis (ESun PLA+)
- 2x N20 motors
- 1.8A dual motor ESC from Ebay.
- Flysky FS-RX2A Pro micro receiver
- 2S LiPo
The Build

I played around with a few ideas for Crumbs, but my limited CAD skill let me down. I decided to do a test print of an antweight chassis downloaded from Thingiverse before I settled in to a designing session, just to see how everything worked and slotted together. Having a 3D model on your screen is one thing, but I learned a lot by just having a physical version to compare.

Once I had deleted a few attempts, I fired up Sketchup and decided to make a simple 100x100x20mm wedge.

Attachment 9439

Quite a hefty wedge at the front, all sides angled to deflect spinners. Even managed to get some motor mounts and lid supports built in. I was very proud of myself!

Attachment 9440

Test print number 1. Printed at 50% infill to allow the wedge to have enough structure, but also hopefully flex a bit under impact.

Attachment 9441

Small banana for scale.

I experimented with a few different designs for wheels, but eventually settled on 24mm lego wheels. I found a hub adaptor for the N20 shaft to fit the lego tyres on Thingiverse. Works like a charm....

...downside is, it doesn't fit in the chassis. Back to Sketchup!

Due to the size limitation of UK antweights having to fit in to a 100mm-ish cube, I could go up to 142mm in any dimension, providing that I don't expand the other 2. Having a preference for wide wheelbase robots, I upped the width to 130mm and tried again.

Attachment 9445

Attachment 9442Attachment 9443

Perfect!

Attachment 9444

Size comparison with the previous chassis.

Next was to design a lid. I wanted this to include a small block above the motor mounts to press the motors in place. The lid is 3mm, dropped in to the chassis as the posts I included in the chassis design are 17mm.

Attachment 9446

Result! The lid just about fits, but with it being such a tight fit, I don't think it would stand up to impact too well if the chassis took a knock. Perhaps I should take it back 1mm or 2 in future revisions.

Attachment 9447Attachment 9448

And that's where we are at so far! I still need to add in a hole for the power switch and wire everything up, but at least everything fits in place. Currently weighs ~121g, so still some room left. It's a rough first draft but I have a functional antweight!

The final plan is to make it look like a mini toaster, so I can maintain my team branding. With the extra weight I may add some stainless steel to the front and top, as I have some left over from the holes I cut in Bread Dead. The final robot will also be printed in white PLA+, with toaster slots either drawn or designed in to the lid.
18th April 2020, 20:05
Ocracoke

Looks good, well done!

Quote:

Due to the size limitation of UK antweights having to fit in to a 100mm-ish cube, I could go up to 142mm in any dimension, providing that I don't expand the other 2. Having a preference for wide wheelbase robots, I upped the width to 130mm and tried again.

The official limit is 101.6mm or 4 inches. ;) I tend to aim for 100mm cubed for mine to ensure it fits.
19th April 2020, 19:27
McMullet

Great stuff! How do you find the PLA+? I want to play with the 3D printer at the Hackspace at some point and I'm curious as to how PLA+ stands up vs normal PLA.

Maybe this one could be modelled as a pat of butter instead? Would be a great mini bot for BDR!
19th April 2020, 20:24
Mr Manuva27

Quote:

The official limit is 101.6mm or 4 inches. ;) I tend to aim for 100mm cubed for mine to ensure it fits.

Thanks! I knew someone was going to bring it up haha! I've been informed that it doesn't have to go flat in the cube though, so will just wedge it in nose first :D.

Quote:

Great stuff! How do you find the PLA+? I want to play with the 3D printer at the Hackspace at some point and I'm curious as to how PLA+ stands up vs normal PLA.

I like it! It's supposed to be 10x stronger than regular PLA, but that depends on your manufacturer as they all have different blends and additives. I stick to ESun who are supposed to be pretty good. Benefits are that it doesn't snap like regular PLA, but flexes more like ABS. Good for impact resistance.

From what I can tell it's relatively untested in combat robots, most people use ABS by default. I can't say for definite until this one dies a violent death in battle. At the very least it is a happy medium between PLA and ABS, as ABS can be quite awkward and toxic to print.

Quote:

Maybe this one could be modelled as a pat of butter instead? Would be a great mini bot for BDR!

Now that's an idea! I've committed to making this one a mini-toaster, but a butter dish sounds like a plan for another one! Although I'm not sure about using them as BDR minibots. I don't think they would survive too well in FW. Even then, my future plans for BDR will likely be using up the whole weight limit. The new welder will be put to good use.
19th April 2020, 23:08
Ocracoke

Quote:

Thanks! I knew someone was going to bring it up haha! I've been informed that it doesn't have to go flat in the cube though, so will just wedge it in nose first :D.

Hahahaa! Yeah, it is just a case of "will the robot fit in any orientation".
26th April 2020, 14:42
Mr Manuva27

1 Attachment(s)

So a slight mishap in that I ordered the wrong reciever..... oops. After some discussion with people on Discord, it was established that it wasn't compatible with what I was trying to do. The main reason I ordered that one was because it was UK based and wasn't shipped from China. Disappointed, I decided to bite the bullet and ordered 3 Flysky FS2A receivers for delivery anywhere from late May to mid-June.

The next day, I get a message on Discord informing me that Bristol Bot Builders had put the exact same Rx up on their store an hour ago. Typical! I couldn't believe it!

I ordered one from BBB while I wait for the Chinese ones to arrive. With a surplus of receivers, I'll just have to build more ants!

After some rather awkward soldering, I'm delighted to announce that Crumbs is funtional!

https://www.youtube.com/watch?v=0RvPVgMnszs

The mixing is a bit wonky, but that will just be a case of switching the motor wires around.

I'll take to fine tuning the CAD design soon for a final print. I still need to add a hole for the power switch and some walls around the wheels to prevent wires getting tangled up in them.

-----------

Anyway, I said I'll have to build more ants, so I'll share my next plan with you. I never want to set out to build a "boring" robot, and Crumbs is certainly heading towards that way. It's a first attempt at CAD, so I suppose it is ok, but I want something more extravagant for the next one.

So, I'm introducing plans for "The Walking Bread". I wanted to do this at Beetle level, but Antweights are AFAIK the only weight class remaining that has a weight bonus for shufflers, with the exception of a few competitions.

I'm basing the design from Mike Franklin's series 6 robot Anarchy, the most successful shuffler in Robot Wars history and one of my favourite robots of the original series. This could be achieved with my 3D printer and would give me 225g, allowing me to focus on adding a powerful weapon like a brushless spinner.

Anarchy uses 2 banks of 4 cam gears driven by chains to move the legs, which may be a bit much for an ant. Looking at the mechanism for Scuttle, the Technogames predecessor to Anarchy, I think it is run from one motor, with a set of smaller gears driving the bottom bank of cams. As far as I can see, the top bank is unpowered.

https://www.youtube.com/watch?v=PMlfEEvoHx8

So at the moment I'm in two minds. I either copy Scuttle and end up with many small gears (See below draft), or I utilise a pulley system along the bottom bank with 3D printer timing belts. The pulleys would be easier to design and print, but may be heavier and less durable.

Attachment 9449

It's definitely going to be a challenge. I'm still working all the details out, but I've posted it now. I have to follow through with it.

If you have any ideas of feedback, please let me know! :)
27th April 2020, 09:25
Ocracoke

Looks good, well done so far. Receivers I find can be a bit fiddly but then again, I tend to stick with DSMX. Lemons I find work best with the OrangeRX transmitter I have. I think the gear system on a walkker could work but it would be awfully exposed to weaponry, you may need to layer it like Anarchy. Bonus points for kudos though.
29th April 2020, 11:07
RogueTwoRobots

2 Attachment(s)

First of all, 10/10 for naming it The Walking Bread :p

You're right with Scuttle, only one motor driving each side. The motor pinion drives the middle small gear which in turn drives the bottom row which is then geared to the top. I've attached a couple of pictures of it I took at RoboNerd 2018; there's very little inside it! While timing pulleys may be the easier option, there's more chance of a belt slipping or similar, which would knock the walking gait right out of sync and you could end up with legs that move but don't propel you anywhere. With properly meshed gears, once you've set up all the legs in their positions, they won't go out of sync. Just for reference, Scuttle's legs are set 90 degrees apart in the gearing so only has one foot on the ground per side at any one time, whereas Anarchy has 180 degree spacing so has two feet on the ground per side.

Attachment 9450 Attachment 9451
29th April 2020, 18:36
McMullet

Quote:

Originally Posted by Mr Manuva27

Now that's an idea! I've committed to making this one a mini-toaster, but a butter dish sounds like a plan for another one! Although I'm not sure about using them as BDR minibots. I don't think they would survive too well in FW. Even then, my future plans for BDR will likely be using up the whole weight limit. The new welder will be put to good use.

I have this vision of a whole family of toast-related machines now... But good call on using up the weight. Good luck with the welding!

Quote:

Originally Posted by RogueTwoRobots

First of all, 10/10 for naming it The Walking Bread :p

Agreed... I am struggling to get past the greatness of the pun, but that looks like an awesome build. Should be fun!
11th May 2020, 20:32
Mr Manuva27

8 Attachment(s)
Quote:

I've attached a couple of pictures of it I took at RoboNerd 2018

Jamie, you absolute hero! I was struggling quite a bit with that, that helps out massively. Thank you very much for the info!

-----------

Crumbs

Thank you for all the kind words, everybody. I'm afraid I haven't started the shuffler yet, but I have been focusing on making Crumbs a bit more interesting than just a wedge. I wasn't happy with how cramped it was inside the bot and wanted to make space for a servo so I could incorporate a lifter. I wasn't too happy about it, but I tried moving the wheels to the outside.

Attachment 9526 Attachment 9527

This new design incorporated new individual motor mounts, plus mounting posts in the back panel and lid for a power switch. Print time!

Attachment 9528Attachment 9529

Ok, so:

Positives
- Motor mounts fit perfectly
Negatives
- Wheel gaps are too tight and exposed
- Not much extra space
- Print quality is awful.
- I absolutely hate it.
After a few days of swearing at SketchUp, I came to the conclusion that I had lost all sight of what was good and bad within the project, so I scrapped it and started again.

The following result was an improvement:

Attachment 9530

Finally, some steps in the right direction! Can we guess what the next steps are?

...

...

..If you guessed "Scrap it again and start learning a different CAD software", you would be absolutely right!

Yes, I downloaded Fusion360. I initially used SketchUp from it being the only free software I knew of, but when I was told that Fusion was free for hobbyists, I signed up immediately. It took a day or two to get used to, but over the course of the weekend I ended up with the below:

Attachment 9531

Much better. This design includes enough space for all the electronics, plus a servo. With the layout of the design, the servo gear can be used to drive a lifter, axe or grabber, depending on which opponent I am fighting.

Attachment 9532

Above: first test print. Below: Weigh in with all electronics fitted.

Attachment 9533

Still not quite there yet, it still needs a lid and a weapon. The chassis itself needs a few tweaks too, like lowering the walls of the motor mounts and including posts for screws, power switch, etc. With only 22g remaining, I may be a bit pushed for weight, but hopefully I can work around that.

All in all, very pleased with how it is going!
11th May 2020, 22:40
Ocracoke

It is looking mighty neat in there, well done! For posts, couldn't you fill in four corners of the print a little and use something like a M2 self tapper to hold a lid down with? To save weight, might be worth shortening the wiring a bit?
16th May 2020, 19:20
McMullet
Quote:
Originally Posted by Mr Manuva27

Positives

Motor mounts fit perfectly

Negatives

Wheel gaps are too tight and exposed
Not much extra space
Print quality is awful.
I absolutely hate it.
That last point made me chuckle. :mrgreen:

The new Crumbs looks really nice, very impressive. Good plan on using F360 too, I was also a bit incredulous that they let us use it for free!
26th May 2020, 19:24
Mr Manuva27

2 Attachment(s)

Bit late, but Crumbs is done.... kind of. This whole build has given me a new respect for those who have antweights as their "main class", because this has been so awkward and fiddly to build. Probably a large portion of that is due to my design choices, but still.

https://www.youtube.com/watch?v=8eZQLy3IQ6M

Lifter gears have been put in place and the lifter works. The lifting arm itself is just some leftover 1mm stainless steel I have from the toaster on Bread Dead. It doesn't lift much, just kind of waggles. I think the pivot is too far forward to get proper leverage. Looks cute though!

Had some issues with the gears slipping, but that was solved with a larger lifter axle and printing the servo gear 1mm bigger.

Printed some top panels, then it was time for a test drive!

https://www.youtube.com/watch?v=L2QNrtAasL0

So Crumbs is fully functional. It weighs in at 142g, so may use the final few to redesign the top panels so they print better.

Attachment 9573 Attachment 9574

One final development is that I have managed to get my hands on one of BBB's beta sale of the new Malenki-Nano Rx/ESC, which can drive 3 channels and replace both the Rx and ESC all on one 2.6g board. I'll fit this to Crumbs eventually (key word being eventually. I have a big project to be working on soon. More on that later....;)), but this will save a lot of weight that could be used for reinforcing the front with more stainless and changing the lifter servo out for an N20 for extra lifting power.

Link to the Malenki-Nano specs if anyone is interested, they are planning more stock soon.
https://shop.bristolbotbuilders.com/product/malenki/
27th May 2020, 01:08
Ocracoke

Looks awesome, well done! I do like the look of the lifter mechanism, even if it isn't the most effective. I guess the only way to test it is to try and lift 150g+. I know Ikari can lift that much with a similar forward pivot, just awkward to get underneath opponents due to the shape.

Quote:

This whole build has given me a new respect for those who have antweights as their "main class", because this has been so awkward and fiddly to build. Probably a large portion of that is due to my design choices, but still.

Indeed, I struggle to build Ants for relatively similar reasons. Takes a few goes with a Dremel and a steady hand to get it right :P
31st May 2020, 10:16
McMullet

The flipper looks great! Agreed on the fiddliness of building ants. I think the layout will get easier after a few goes but I'm also struggling with the tiny solder pads.

I saw the nano ESCs on the BBB shop, it'll be good to see your test run. For our ants I've come up with a halfway house using one of these taped to the standard dual-channel ESC. Saves the weight of the wires and connectors at least!
19th December 2020, 16:00
Mr Manuva27

So BotFest was due to happen a few months ago, but was unfortunately cancelled due to lockdown. I had redesigned Crumbs! to improve the lifter as the servo driven one didn't have enough power.

I always wanted to build a 4-bar lifter but kept it under wraps until the event. When it was cancelled, I lost a lot of motivation to work on it and it sat in my Fusion360 files for a few months, until last weekend!

https://www.youtube.com/watch?v=YTISj8Yy7Kk

The arm is now powered by a 100rpm N20 rather than a servo, plus a 300mAh 2S battery.

It needs reprinting at a thinner layer thickness before it is truly combat ready, plus superglueing a stainless steel beak (leftover from BDR) to the front arm. Regardless, I'm super happy with it.

Hopefully events get back to normal sooner rather than later so I can take it for a proper test run!
19th December 2020, 23:30
Ocracoke

Yeah, I hope I get to see this soon, love the arm action! I've been cleaning out the shed here and found the BotFest arena in there, sitting there ready to go.
20th December 2020, 14:19
McMullet

Looking great Andy, I'm also working on an antweight 4-bar lifter (very much languishing in work-in-progress CAD though).

Do you have a way to stop the motor when the lifter gets to the end of its travel? I was planning on using a servo to stop it damaging itself, but it doesn't seem like a great power-to-weight ratio whereas that is super fast, looks more like a flipper than a lifter.
20th December 2020, 21:47
Mr Manuva27

Quote:

Yeah, I hope I get to see this soon, love the arm action! I've been cleaning out the shed here and found the BotFest arena in there, sitting there ready to go.

Thanks! Really pleased with how the arm movement came out. I still need to adjust the horns at the back so it can self right, but aside from that it's perfect. Fingers crossed we get the all clear to fight again soon.

Quote:

Do you have a way to stop the motor when the lifter gets to the end of its travel?

I do! It's called the floor ;). Because my transmitter has both sticks on spring return, just a quick burst of the motor is enough to get full extension. Don't think it's enough to burn the motor out, but proof of the pudding shall be in the eating.

Quote:

I was planning on using a servo to stop it damaging itself, but it doesn't seem like a great power-to-weight ratio whereas that is super fast, looks more like a flipper than a lifter.

Servo should be fine. It's worth noting that a 4-bar arm requires different force depending which end is powered. The "crank" (rear arm) takes less force to lift than the "rocker" arm (front). If you're worrying about a servo not having enough power, you'll be better off linking it to the back arm. It should be sufficient. I'm only using an N20 because I'm using a Malenki Nano ESC, which works better for motors than servos.

If you're making a 4-bar, I recommend the AskAaron blog and geometry calculator/excel sheet for it. Took lots of trial and error, but was worth it in the end.

http://runamok.tech/AskAaron/lifter_FAQ.html
23rd December 2020, 17:28
McMullet

I've spent FAR too long playing with the spreadsheet. Happy with that part, just messing around with the infinite design space of 3D printing now... :mrgreen:

Quote:

Originally Posted by Mr Manuva27

I do! It's called the floor ;). Because my transmitter has both sticks on spring return, just a quick burst of the motor is enough to get full extension. Don't think it's enough to burn the motor out, but proof of the pudding shall be in the eating.

That's a good point. I was assuming I'd just stupidly hold the stick down with the motor stalled but in reality I'd probably be able to let go after a second or so... N20s look like a better power to weight than a servo.