Thursday, 11 September 2014


Title lies. Only one beetleweight really.

Some of you may remember eggbeater. It is a comically large beetleweight with an equally obese weapon. 500g of hurty titanium. Going ZOMG MASSIVE WEAPON is abhorrently newbie, but in fairness I was incredibly newbie when I gave the dimensions when it was made about a year ago. It featured a much nicer rebuild in February. Still a long way away from good but it slightly functioned and it actually was able to drive.

I broke the previous version down into what sucked and what didn't. The overall concept was fairly solid, the execution, material and basically everything else was rubbish. Stronger, smaller and more sensible was the goal for the new version.

While hanging out with Ellis we went to his material stockist and I accidentally a half meter length of 80*10mm aluminium flat bar. It seemed like a good idea at the time okay? I started this robot by lopping off a hunk, carving it up and bumming holes in it.

After drilling the pilot hole out to 12mm (largest drill I currently own) I stuck the boring head in and began trying to make bearing pockets. Normally my holes are painfully oversize by about 2 thou or so.

Not this time though. The fit was great. Still getting used to the boring head but both the bearings plopped into place nicely.

Finished bulkhead. Came out very nicely. I decided to quit while I was ahead and finished up for the day. The next day I began by doing the marking out same as the first one, then more holes + boring and doing all the curves and milling work. It came out much better the second time. I changed the cutter from a fairly blunt 3/8" to a sharpened 10mm. Finish was worlds better. Lesson learned.

A little tool doink happened on this part sadly, and the finish was ropey until I changed cutters. But you can see the difference in the last picture. Left hand side was with the new cutter.

The bulkheads then got slotted to fit the 5mm aluminium flat bar that made up some of the frame. Bit overkill I spouse, but that's about the thinnest that can hold a semi decent thread and with the beater spinning to self destructing RPM I would like a substantial metal plate between it and the lipos.

Also at this point I had a Hobbyking order come. Spare brushless ESC's in case of death by cheapness/abuse/hella current draw as well as some small servos and a couple of these.

Daww well isn't this a little cutie. 3800kv and 4s makes this a total beast. Fast motor + high reduction seems to be my choice for weapons. Hope it works okay without exploding itself or detonating my escs. Only bad points so far about this motor are the stepped shaft (booooooo) and the fact it has M2 and M3 mounts. Y u no all M3??

Ellis was a proper lad and machined my dual groove pulleys for eggbeater. They have a 4mm rounded groove instead of a V or a flat which means they grip like buggery and hopefully won't need extreme tensioning.

Worth every penny. All dimensions correct to .02 at least.

To engage on the beater, the boss of the larger pulley was to have a 17mm wide slot milled in it in order to engage with the 17mm flat portion of the beater. This was so much more reliable than setscrews or anything else I could really do.

Machinists/anybody halfway competent avert your eyes. Dodgy machine set-ups are my speciality.

 It worked though, so I'm not complaining. It was a pressfit on the beater in the end. Not so extreme it can never come off but it would need a fair amount of persuasion in order to do so.

Putting the set screw into the smaller pinion pulley (NFI what the correct term is) Admire Ellis's amazing grooves while you're here. The screw is large in comparison to the shaft to aid in making setscrews not suck. Should be enough to work at this level. They never entirely unsuck though.

Around this point the belting material arrived for the drive and weapon system. I quickly and I mean real quickly jammed together some belts and dug out the dual ESC for the drive. I had some massive problems with it to start with. One motor would just jitter like hell and cut everything out. At first I though it could be because I was using a DX3R transmitter for the first time over my DX5e. After a bit of head scratching I isolated one of the motors as being at fault. The brushes were totally cattled causing the jittering and were shorting out causing the emergency shut downs. Changed motor and everything was fine!

Ignore the bobbing. It just does that on the paving slabs. Flat surfaces it is fine on. I will try and magnets on the back anyway for more weight on the motors.

I ran it hard after that. Just to see what would break and if it could survive itself. I'm not overly sure on time but I had this on in the background and it took me the first two and a half songs to get it trimmed and mixed okay, so maybe 30 mins or so?

What happened:

Shit broke. Clumsily-welded-with-hope and a lighter belts failed. As expected truth told. Will make a proper welding jig for them out of nylon or HDPE or something me thinks. The other one held great I will be keeping that as a spare or cutting and rejoining properly.

Also everything got hella hella hot. Like Emma Watson levels of hotness. Within about 5 minutes the sabertooth was too hot to touch (but somehow kept functioning fine) and the drive motors were baking. I initially put it down to stupid robot design, the motors having to lug a huge weight around with a mechanical disadvantage. It all became clear when the belts came off. The motors were turning without the belt, but the tension was so high with the belts on it was stalling the motors out. I'd tried to trim them a little bit to get the robot to go straighter. No wonder everything was toasty with an almost constant stall. Sorted this then all was well and robot was happy again.

Start of a new day and I had a chance to have a closer look at the parts that came the day before (I was rushing out the door to work when the post came) Silversteel shafting, bolts of various flavours, and a sheet of carbon fibre. You can see Fakdisc relaxing there too with its shiny new bolts fitted.

Popped my motor apart to get shot of the shitshaft and replace it with a custom length of silversteel. This eliminates the stepped down to 3.175mm nonsense and it stops the MASSIVE STRESS RISER  circlip grove from being a problem

Rough cut to length then I faced off the ends. It was less of a bitch to turn than thought/remember. I had some work to do in silversteel back in college and I remember it being a huge pain in the hole. It went totally fine now.

Booped the motor back together with the new shaft. I was expecting it to be a bearing fit out of the box but NOPE chuck testa. Probably because shonky bearings in the motor. Big, small, cheap, expensive, happy, sad I have yet to have a brushless with bearings I would qualify as *good* But yeah. Took the shaft down to size so its a nice fit. Span the motor and it ran beautifully. Sweet.

Began marking out/cutting the carbon fibre. I had planned to do this with an angle grinder but my neighbours were having a barbecue (slightly in the rain) and spraying people who don't complain about late night/very early morning drilling are people you don't want to spray in the face with toxic CF and epoxy dust.  Did it by hand with a hacksaw with a fine toothed blade. Went well but a tad wonky. Nothing some sanding can't fix. Because of the wonk I made this piece the base plate. I just Pretend-O'd it on as the lid to see how it looked.

Tapping the mounting holes (M3) into the bulkheads.

Hello scrap of 3mm aluminium. Your destiny is to become a motor mount. Prepare to be milled and Swiss cheesed full of holes.

Excuse the horrible screws. I didn't have any hex countersunk M3 in the right length. I didn't bother with the set of M2 screws because I flat out didn't have any. With loctite these should be enough.

Mounted to the bulkhead with some lovely standoffs from Ellis. Pulley is also attached. Still all spins good. The motor goes like a mad bugger and pumps out a tonne of air. Hopefully that will help with some of the heat in the robot.

Made the rear armour. No pictures of the process as it wasn't that interesting (not like that has stopped me before) but it was just booping 8 holes with the hole-booper and a bit of tapping. The piece is 6mm HDPE.

 Internally its pretty spacious. Like, not just room to swing a few cats but the owners of said cats and their significant others too. No excuses for not having neater wiring now.

A bit of wiring later and I had stuff ready to go for a spin test this time. Round  one:

Result: Disappointing IMO. Naffy belt joins failed again (one did, the other stayed solid) but it span on one belt which is good. Dat flap tho. One thing I stupidly failed to account for was the heat caused by friction in the system to soften the already softer belts causing them to stretch. Like.. really bad, it went from Tyrion Lannister to Michael Jordan.

Bad points:

-belts got warm (unsurprisingly) so stretched. This is why its flapping around in the video
-I'm loosing a several thousand RPM somewhere, probably because of belt tension but very annoying all the same. Hope I can fix that.
-One belt snapped on spin down of the first time. But as the second one was fine throughout all the testing can put it down to dodgy joints.
-30a ESC hella hot so switched to the HK own brand £7 lumptroller. Heavier but it didn't miss a beat.

Good points:

-Nothing exploded
-It span with no tension and one belt
-Motor makes noise to scare opponents
-£7 ESC actually worked.
-Batteries were not hot to the touch

About dat motor noise. The motor is srsly over revving

Belting saw a change for the better and all parts of the weapon system got a bit more attention lavished on it and it was time for round 2.

It sounds better. It feels a lot nicer. I think physics has something to say about the theoretical max speed and air resistance combined with the motor power means I highly doubt it will ever get there.

It then got left alone for a few days because I got asked to do some extra days at work.

Not the neatest wiring job ever, far from it but it works and everything does what it's supposed to.

Cut out the lid from more 2mm carbon fibre and sanded down the edges. A slot for the belts was also cut out.

I had some problems with bearings moving outwards on hits or after extended running. Dave Moulds suggested making some washers that bolt to the beater itself and hold the bearings captive. I got a sheet of .08" brass out.

Many bloody hours later.

Horrorshow. Should be grand. One side is a much better fit, the problem only occurred on one side so for reasons of weight (and that this part was a pain in the arse to machine) I may end up with just one.

Wheel armour. Well "armour" it's 2mm hdpe. It will be bouncy and flexy and stop instant belt death but I'm not going to kid myself into thinking it will be any good. It changed from the CAD because weight reasons. Still. It's mostly accurate.

This is kind of it before the competition. I'm fully expecting it to get totally rekd if it has to fight anything more deadly than a small paper cube. Event rundown coming soon.

No comments:

Post a Comment