Robot refuses to turn!
 

Alright its the last few days of the build season and everything is going fine. Our kicker is looking good, electronics are mounted, and our weight is right at where the CAD model said it would be.

Well almost everything is good, our robot won't turn.

We built a basic six wheel design. There are 6 8in plaction wheels, 3 to a side, all hooked up by a chaindrive, so all 6 wheels are powered. So we take a battery reading 12.5 volts and the thing flies forwards beautifully. Next we switch up the wires so that one side is going forwards and the other back so we can test our turning. It doesn't move at all. We take a look and we realize we made a design mistake, the distance between our front and rear wheels is greater then the distance between the two sets of wheels. Oops. So no big deal we think, we can fix it easily enough. We pop off our front wheels thus halving the distance between our "front" wheels (the old middle ones) and our back wheels. We then go add the power again and it still doesn't turn at all. With a bunch more testing we find out that the only way we can get our bot to turn is if all the weight is supported by one parallel set of wheels.

Our bot has a 24 to 1 gear reduction, so that translates to about 6 feet per second under normal load, and when moving forward our bot does close to that. Perhaps 4 to 5 feet per second. the distance between the one wheel and the next is 12.5 inches and the distance between one drive and the other is 24 inches. Nothing is jamming, the motors are spinning all in the correct direction when we take the weight off the wheels but no matter what we do the bot will not turn under its own power. Are we breaking some kind of cardinal rule here? I would appreciate any input.

We thing the best fix will be to replace the front and back wheels with omnis, leaving just the middle ones being plaction wheels. We hope this will solve our problem, but we do not want to throw $200 bucks at the problem unless we know it will fix it.

Final thought. We were testing this by hooking up the motors directly to the battery. I assume many people will tell me this is a bad idea, maybe for safety reasons, maybe because its hard on the motors etc. I just want to say that I am aware of the risks, but we needed to know if our robot would work at all not two days from now after we have figured out that the problem wasn't an electrical one. If we would have hooked up the motors to the jags and run everything the way it was supposed to would it have any significant improvement in the performance of our bot?

Re: Robot refuses to turn!
 

Is is a 6WD "drop center" or "flat"? What motors are you using?

If you're running flat, get 4 omnis and put pairs in diagonal corners. This may help. (Why pairs? it smooths out the ride so you're replacing wheels less often.)

I'd guess that you're experiencing the "scrub" that causes many 4WD robots to bounce around and many >4WD robots to need omnis or a dropped set of wheels to turn.

If you're running one motor per side, you might add a motor to each side.

Re: Robot refuses to turn!
 

I forgot to mention that we are running 4 CIMs. The original configuration was a flat drive, but by removing the front two gears the design should have handled like a drop.

Re: Robot refuses to turn!
 

When you say it doesn't turn, what are the symptoms? Do the motors stall? Do the wheels spin? Is your battery voltage above 10 volts or so when trying to turn?

Also, it might help if you provided a picture of your robot so we can look for potentially other problems.

Last time I ran the numbers, 6 ft/sec with 4 CIMS and roughtop (I think I used a coefficient of friction of 1.5) should be able to spin tires while drawing only about 40 amps per motor. We've also used 6wd long for a couple years, with a dropped center wheel, and not had trouble turning. So, I'm having trouble visualizing what's going wrong with your setup.

Re: Robot refuses to turn!
 

Battery was topped up at 12.5 volts.

Wheels twitch when you add power, you can easily tell they are pushing. If you lift up one side of the bot moves merrily. I am assuming that the bot is at stall amp draw because it is hooked directly to the battery an it is not moving. Also the wires heat up, so significant amps are flowing.

I can post detailed cad models in a minute.

Re: Robot refuses to turn!
 

A 6 wheel long robot WILL NOT TURN unless you drop the center wheel by 1/8" to 1/4" of an inch.

A skid steer, or tank drive, literally "drags" the front and rear wheels through a turn.

A wide orientation robot with 4 wheel drive will turn if geared around 9-10 fps. This wheel ratio of length to width is necessary to maintain if you want to be able to turn. That means when you go to 6 wheel drive you need 2 of those wheels off the floor to get the same length and width ratios that you have on the wide robot. Essentially by dropping the center wheel you'll be on the middle two wheels all the time, and then either the back or the front wheels depending on your center of gravity.

If you do not want to drop your middle wheels, then you can put a single omni in the front and in the back. If you are still having difficulty turning, you can dump the traction wheel for something with a little less side-traction, like the old skyways, a traction wheel with the tread worn smooth, or last year's ABS wheels.

Re: Robot refuses to turn!
 

What are your dimensions?

Distance between axles also your width.

24:1 between your wheels and the CIM's may be a bit light for an 8" wheel with that much traction but I would think it should be able to turn if you get the right ratio between width and center distance.

Give the dims and let's talk.

Joe J.

Re: Robot refuses to turn!
 

http://www.flickr.com/photos/47595136@N03/4364660600/
http://www.flickr.com/photos/47595136@N03/4364660576/
http://www.flickr.com/photos/47595136@N03/4364660548/
http://www.flickr.com/photos/47595136@N03/4363918531/

As you can see a simple 6 wheeler, undropped.

While we were testing we removed the wheels up by the kicker to simulate a dropped drive to see if that solved our issues. Because the majority of our weight is in the back it rested nicely on just 4 wheels. Sill didn't turn with only 12.5 inches between the back wheel and our "front wheel"

*One sec the pics are not appearing
*It's poor form but I just hyper-linked the photos, not going to sort out why the images didn't want to load.

Re: Robot refuses to turn!
 

@ Joe

Between axles 12.5 in apiece, 25 total
Width between wheels side to side, 24 in, measured from the middle of the wheel.

Those specs were in my first post, not surprised it was passed by, I have handed in shorter essays to my English teachers then that post.

* found a typo in my first post, I see why the specs were missed when vs wheel changes the meaning of a sentence a bit.

Re: Robot refuses to turn!
 

Any chance you dont have the motor polarities switched properly, so that pairs are working against each other? Or, shorting the battery?
Just a thought.

Re: Robot refuses to turn!
 

Checked polarities, we thought for sure that was the issue, but to no avail. the motors run fine when the wheels are of the ground, ie. one side (2 motors) is going forwards, one is going back (other two motors).

If by shorting the battery you me crossing leads some where so the energy doesn't get to the battery, then no we are not. Everything runs fine in Forwards/Reverse or when there is no load on the wheels.

Re: Robot refuses to turn!
 

Better go with the omnis, or grab the slicks from the KOP if they'll fit.

Re: Robot refuses to turn!
 

Actually we at one point replaced the front wheel with a slick, but nothing changed at all.

slick in the back would have been good, but axle is this odd hex shaft that is a real pain to switch out a wheel on it. looks like a project for tomorrow.

This whole thing is bugging me because one we switched our wheel base to just 12.5 in that thing should have spun like a top. I am just agonizing that I am doing something dumb here, similar in caliber attaching the motors wrong, but I just can't find anything wrong.

Btw are omnis the same diameter and width as the plaction wheels? Andymarks specs have left me feeling a little wanting.

we have these http://andymark.biz/am-0514.html
and i think these are basically the same form factor http://andymark.biz/am-0048.html, barring the .5 in difference in the hub width.

Re: Robot refuses to turn!
 

Something is wrong.

A 24inch wide 12.5 inch long wheelbase with two cims on each side geared at 24:1 (the low gear on the supershifter ratio) should be able to turn. I'm taking this dimension from your test where you removed both front wheels.

You say you're directly powering the cims on both sides from the battery, with no breakers involved.

You say you've taken a battery with 12.5 volts. That is not a charged battery. A fully charged battery will usually read between 12.8 and 13.1 Volts. A 3/4 discharged battery can still read 12.5 until it is under load, then it can drop to 7 and 8 volts.

#1 Confirm that the battery is fully charged.
#2 Confirm to us that you are running on regulation carpet WITH NO PADDING under it. Carpet with padding is death to turning.

Edit: Can you please confirm that you're running dual cims on each side, and not singles?

Re: Robot refuses to turn!
 

Did you remove the two front wheels or just the sprockets on them? Removing the sprockets doesn't do much of anything for you, the wheels still have to drag across the carpet.

If you removed the wheels, then as Tom suggested, that robot definitely seems like it should turn.

Re: Robot refuses to turn!
 

You are not getting full torque to the wheels, simple as that.

Either you are not powering both motors (all 4 really) or you are do not have the gearboxes build right or your chain is eating up the torque or something.

Traction is not the problem.

Hunt down the torque loss.

Joe J.

Re: Robot refuses to turn!
 

The whole wheel is gone, nothing is touching. The carpet has concrete beneath it.

Yes I am directly powering the cims on both sides from the battery, with no breakers involved.

Can you please confirm that you're running dual cims on each side, and not singles? ~ This is true as well.

I will check that the voltage isn't dropping too much, but if we are drawing stall current (130A?) from 4 motors the voltage will decrease a ton no matter how charged it is.


As you said, something is wrong. The output shaft spins without applying to much force, and you can easily spin the wheels by hand. You can tell there is some efficiency loss but it seams to be working more of less in proper order.

I will see if I can find a way to make it all spin better, but I am not hopeful.
I will also see if i can get a battery reading 13 volts.

Re: Robot refuses to turn!
 

Ed,
You have demonstrated a classic issue with high friction wheels in a tank turn. The motor wires heating up is a sign that should not be ignored. If you were feeding the motors through circuit breakers, they would be tripping. As others have pointed out above, teams have discovered that raising/offsetting the height of the wheels prevents the incredible side load of all wheels on the floor. This cause each of the CIM motors into near stall current draws. At 129 amps each that is a total of 516 amps. You might have also noticed that the Crio is inhibiting output for a short period of time (if it is not rebooting). When the battery voltage falls to 5.5 volts, the Crio inhibits output in anticipation of a voltage fault. This causes a somewhat jerky motion during the fault. With a 516 amp draw, the battery will not be 12 volts no matter how well it is charged. You will drop voltage across the internal resistance of the battery that may take it down 5 volts or more.
Dr. Joe will give you some insight on design and as other have pointed out, wheel choice can have an effect as well.
If your motor wires were hot, the internals of the motor were hot as well. A "nose" check around the motors will tell you if you have stressed the motors.

Re: Robot refuses to turn!
 

See if you can measure the current when the wheels are off the ground. You can also infer current by measuring the speed and backing out the motor speed then measure the motor current and doing some simple calcs to get motor current, but I prefer to know motor current directly.

Hey, I just had a brainstorm. How are you managing thrust on your wheels? Scrubbing the tires puts in a very high side load on the tires. If this side load is not managed well (for example if you let the tires rub against the frame) then that adds to the turning resistance that your motors have to overcome.

I bet that there is something going on there. Take some pictures and let us know.

Joe J.

Re: Robot refuses to turn!
 

This is whats driving me crazy, we have in effect done this by removing the front wheels.

@ Joe
Sounds like a good idea to test the motor draw.
There is about .75 in clearance on either side of the wheel that is set by a spacer that doesn't rub the wrong way against the wheels in..... wait a sec.

I have an idea that could potentially cause the bearings to seize up. I'll get back this evening.

Re: Robot refuses to turn!
 

Four sticky wheels on the floor are enough to stall the motors. If the robot runs OK with the wheels off the floor then that is like waving a red flag.

Re: Robot refuses to turn!
 

This may sound like a dumb question, but by any chance are your supershifters in high gear when doing these tests? It may sound silly, but that might alleviate your problem with turning in the mean time.

As for the AM omnis you posted, they are the same diameter.

Re: Robot refuses to turn!
 

On a team I mentored a few years back during Rack n' Roll, we had a 4 wheel tank drive with high traction wheels. It would not turn either. We overcame the issue by adding zip ties in an "X" pattern around the front wheels.

With six wheel drive, you would need this on both front and rear wheel sets. If this works for you, it won't cost much and is easily modifiable by adding more or fewer ties across the tread.

Alternatively, run one more layer of tread around the center wheels only, thus giving a "drop center" effect. Make sure to secure this second layer very well.

Re: Robot refuses to turn!
 

With just the middle and rear wheels in play the aspect ratio is 24 Wide by 12.5 between wheels, this is enough that the thing should turn, even with very grippy wheels (given 2 CIMS each side plus 24:1 to 8" wheels).

Either the torque isn't getting into the gearbox, it isn't getting to the wheels or it is getting to the wheels but something else above and beyond traction to the carpet (including scrubbing loads, aka thrust loads) is using up the torque.

He says that the motors are both wired right. The thing moves at about the right speed going straight. I suppose that going straight it will use the right amount of current.

To me that means something is binding when the robot tries to spin. The wheels could be hitting something, the frame could be racking and causing a bind, the wheel's may have bad thrust bearings or bad thrust load management, the chain may bind up when the wheels see the twisting loads, etc. ... we are looking for something extra that happens only (or most prominently) when the wheels see significant side loading due to being forced to scrub along the carpet.

I'd bet a Mt. Dew on it.

Joe J.

Re: Robot refuses to turn!
 

Hopefully I read everything correctly. You stated that you're running a 6WD bot, which is great!... but you need to consider a few things that have been mentioned already. First off, when everyone says to lower the middle wheel, you really need to do this somehow. We've gotten away with it in the past with pneumatic wheels that we could air up to raise and lower the center of the robot. Very good solution might I say.

Attachment 8705

That is a quick sketch of a lowered centered wheel. I believe we've used about a 1/8" drop which has worked great for years.

If you can remove the front wheels, and the robot maintains stability, then you have a balance problem. The entire weight of the robot should remain mostly on the center wheels, with very little weight on the outer wheels. You said that you completely removed the front wheels, which leads me to think you have way too much weight in the back. If you haven't already, you should move that roughly 12lb ballast called a battery to the far front, and if that's not enough, the other ballast (cRIO) may work out, as well as most of the other electrical components.... or if you have spare weight, just add some weight to the front.

Like Chuck said, make sure you're not in high gear... although it should be able to turn in high gear, but it would be a very easy programming fix that could shift it into low gear when the joysticks are the inverse of each other. (one is positive, one is negative).

Good luck! I hope you fix it! I know it stinks to be this late in the season when the design is coming together, just to find out that you have a major problem like this. Every team has probably been there at some point, I know ours has. Just think it through, get creative. I do recommend a pneumatic wheel if your design permits. Remember when you're turning, the outer wheels do nothing... except bind the robot.

Re: Robot refuses to turn!
 

Hrm. Still having problems? Let's confirm a ew more things:

1. Live axle or dead axle, cantalevered or not, and are you using bearings or bushings?

2. How tight are your chain runs, are they nicely alligned, are your transmissions stiffly mounted or flexing?

3. How long are the motor shafts, are the sporckets mounted as close to the tranny as possible to prevent bending moment?

4. What trannies are you using? There are a number of cim pinions that are very close to the same but will not function well interchangeable under load. Have you disassembled the trannies and confirmed, step by step, correct assembly and lubrication?

Re: Robot refuses to turn!
 

basically, it sounds like you have too much traction on the outside wheels, omni's are probably the best solution, but slick wheels would probably work

Re: Robot refuses to turn!
 

Ed,
I guess we are at that point that we need a real photo of the wheels. As Joe has suggested, there might be an issue with the axle/transmission/wheel interfaces. It is the kind of thing we can look at and have a "I should have had a V8" moment and know exactly what is wrong. For instance, an unsupported shaft can have extreme friction in turns that does not exist in straight driving.

Re: Robot refuses to turn!
 

What ever happened to this poor robot?

I have a sad image in my mind of this little guy going back and forth over the bumps and back again but never turning, never facing the goal or the tower... back... and forth... back... and forth...

Give us an update please.

Joe J.

Re: Robot refuses to turn!
 

In order to fall asleep, most of us count sheep. The Good Doctor counts robots!

Who knew? :rolleyes:

Mike

Re: Robot refuses to turn!
 

I don't see the rule about starting on top of another bot this year. Perhaps they're shifting they're strategy?

Re: Robot refuses to turn!
 

Umm, how would you accomplish this when every robot has to start in a separate zone?

Re: Robot refuses to turn!
 

Sorry it has been a busy weekend. The bot is up operating and fully functional. with a high gear of 9 to 1 she really moves. We have ball possession and a kicker going just fine. No climbing.

Re: Robot refuses to turn!
 

So it actually turns correctly now? :confused:

Re: Robot refuses to turn!
 

And if it does turn correctly now, what did you determine the problem to be?

Re: Robot refuses to turn!
 

Put omnis on the corners. The spin test we did was impressive. We assume the issue was drive train inefficiency, we did the math and the bot should have turned....

Re: Robot refuses to turn!
 

Glad you figured it out.