Lara's Lego H-015 Sentry Droid

[kresty]

Got 3/4 of the motors replaced.  Tried pretty hard to get the wheel off the shaft of the 4th, but it just wasn't going to happen, so I guess it gets to stay that way.  If it falls apart in normal use, we'll replace it's mount too.

It runs *way* better than I remember

New Motor Mounts on Lego H-015
This works WAY better - one still using the old mount, however we couldn't get the wheel off the hub, so I guess it's OK.

So: Fixed in time to show off to the First team we're meeting this evening.

[kresty]

Oops.

Was working on replacing the corner bits with the better fit corner bits (probably won't do that 'cause they didn't print in time anyway, but still...)

Oops, forgot to take bearing out of 3D model before sending to printer.

[kresty]

I'm going to have to find/post video of our outing, it was cool...  He did go haywire a bit which resulted in him knocking apart his antenna (he ran under an overhang that was too low).

I'm using the Pololu Simple Motor Controllers.  After his crazy attack, I checked the motors and only 2 were running (1st and last on the daisy chain).  Cycling power made it all happy again.

The Pololu SMC's have a safety mode, I'm wondering if they got into that mode.  And maybe why those two.  We'd let a kid drive him, so it might have been running a little more than we would, maybe they overheated?  

I might modify the code a little to query the controllers frequently and see if they've gone into safe mode.  If so, then shut down all four motors to prevent this again...

Presumably if it wasn't a kid unused to the controls we would've caught on and stopped sooner too...

Anyway, I'm reprinting some antennae.

[kresty]

Got 3/4 of the motors replaced.  Tried pretty hard to get the wheel off the shaft of the 4th, but it just wasn't going to happen, so I guess it gets to stay that way.  If it falls apart in normal use, we'll replace it's mount too.

New Motor Mounts on Lego H-015

Well, that's annoying, the 4th one broke pretty early on today and started dragging.  I pulled the whole motor and mount off, shifted the battery over so the center of gravity was away from the missing wheel, and then ran it on 3 motors, which it did surprisingly well.  It's "straight line" paths were diagonally front-left and rear-right, but it wasn't that difficult to adjust to mentally.

Originally I'd planned on bringing in the parts to fix it correctly, but we've reconsidered as that'd mean 30 min or so that it couldn't be on display while we tried to fix it, so we'll just run it on 3 wheels tomorrow too.

Here's a clip of it running on only 3 wheels...

[kresty]

Well, the 3rd one was what I'd hoped. The set screws on the couplers managed to vanish on the wheel side. Guess I should've tightened them better! At least that's an easy fix. One other wheel had a couple loose, but they hadn't fallen out yet.

Still out a 4th wheel as I'd put the parts in a "safe place." Then the workspace got cleaned. Oops.

I might adjust the accel/decel curves for the motor controllers, I fear they may be starting/stopping too fast, helping stress the joints like where the set screws fell out. I think I may need to inspect them after an hour of run time or something to see if this is going to continue to happen.

[kresty]

Installed the 4th wheel on the Sentry, one chore down.  Took a pix in the middle of swapping out the old mount for the new mount.  Printed at a different time, so it's a different color.

Working on getting new wheel mount installed
Finally getting the 4th wheel to behave.

Seems to drive better now

[Toddo]

If it helps, we're using:
* 4 Pololu 37D 30:1 motors,
* 4 VEX 4" omniwheels,
* 4 Pololu 18v7 Simple motor controllers,
* a bad 12V LiFePO4 battery,
* an Arduino (happens to be Leonardo), and
* 5 Pololu addressable LEDs https://www.pololu.com/product/2535 (program as WS2811)

Code is posted http://www.rebeldroids.net/forum/showthr...98#pid3198

Are you using motors with encoders, or just straight positive/negative wires?

[kresty]

Are you using motors with encoders, or just straight positive/negative wires?

We're only using DC motors with +/-.

Another option would be steppers. You're welcome to stop by sometime if you want a better look @ him.

[kresty]

Wondering how to switch to 18V. Likely switch out the 12V Pololu 37D for 24V ones, maybe geared a little less. He's got crazy torque anyway.

Broke two of the couplers at Emerald City Comic Con 2023 in Seattle. I think the wheels are suddenly stopping too quickly. Like when we hit the edge of the wood we set out for the toolbox droid.

Kresty's Droids at ECCC 2023
Lego BB-8, D-O, H-O15 Sentry, MCK-Y, Toolbox, L3-G0

[kresty]

Considering battery issues, my Shorai aren't designed for deep cycle and don't have the Ah they'd advertised. We've used them in all the droids, so were limping along. (They are too easy to over-discharge, are only 7Ah @ 12V, and the chargers are iffy). So, I'm considering moving to Ryobi 18V as I already have tons of them. And it comes in a 9Ah, so that'd be about twice the Wh! (160ish vs 80ish).

The Pololu 18v7 motor controllers are 18v, so that should be OK. (L3-G0's are 24V, and he's run 24V before, so that should be OK if I move the droids to 18V). Side note: I was double checking that 18V batteries would be OK, since they're way higher than 18V when full. They'll be fine, but I noticed the prices changed. I got my last ones on sale which was maybe 1/2 off list at the time, and now they're 4x what I paid!

The Pololu 37D 30:1 motors, however, are 12V. They have a 24V version with comparable specs at 24V (at half the amps), but if I run them at 18V, the 30:1 will be slower than our current motor. We have plenty of torque now, so I can probably go for a lower gear ratio. Excessive torque may have even killed our couplers. The question is probably whether to go for the 19:1 or 10:1. I'm not quite sure how to do the math.

19:1 has 530RPM (no load) vs the current 330RPM. 10:1 has 1000RPM. But we'll be running below 24V and the speed seems to drop in about a linear fashion. Torque would drop from 14kg*cm to 8.5 or 4.9 - but we seem to have tons of excess torque right now.

I guess 75% of 530RPM would be about 400RPM, which is still faster than our current motors (and he already spins way too fast, though straight line isn't terrible.) 75% of 1000RPM would be 750, which seems pretty insane for our current use (we could always limit it though programmatically). The torque for the 10:1 would be a lot less at 3.8, and even 6.3 is a lot less than our currently spec'd 15. I guess I'm going to have to do math somehow.