Turning problems, we need suggestions PLZ

[Oumonkey]

Hi, My robot is having turning problems now. Our programming seems to doing what it is to but our robot won't turn. We lifted it up and made sure it was working right and the motors are going in the right direction, but when we put it on the ground the motors will work forward or backwoods but when we try and turn it it just stands still and the motors struggle to move. I think it is the extra traction we put on the wheels to move smother, so if someone can give us ideas to check it would really help us a lot. If you need some more specific details I might not be able to check them for a few days.

[mechanicalbrain]

YOU AGIAN okay this i can help. its one of several problems one STOP TESTING if its making noises and not moving you have alot of power and you dont know what its doing it should be okay but still.... okay so check you speed controllers and tell me if you are using a gear box and your wheal layout in relation to how their being powered (chains, which wheals are powered, ect...) chances are its an electrical or mechanical problem.

[David Guzman]

First thing u should check is gear boxes, if it works when you lift it up then maybe there is something lose that is bending once you apply pressure to it (put it on the ground)

If thats not what the problem is then check for too much friction on the ground, you said you wouldnt be able to get more specs but can you tell us what type of drive it is.

2wd? 4wd? 6wd? are you using kit gear boxes?

Dave

[mechanicalbrain]

if you have two wheels powered by chain with one gearbox as i suspect you do (we had similar problems) make sure that one your motors turn the right way when you turn or you will break your gear box. and move your robot off any carpet try on a smooth floor. FIRST's double wheel design is really bad for turning.

[JVN]

Originally Posted by mechanicalbrain

FIRST's double wheel design is really bad for turning.

Aaaaaactually...
It turns fine, when used correctly. I'm curious what exactly you're talking about? Perhaps you could give us some clarification or justification?

Back to the problem at hand...

You mention adding traction to the wheels; this leads me to hypothesize that the added scrub friction during turning has now overcome your turning moment.

The symptoms you describe, also make me believe that you may have a low battery. (Low battery = less motor torque = less turning moment). When was it last charged? (I'm just speculating here...)

If the battery is charged after all and the problem persists; cover the tread on 2 of the wheels (rear or front) with duct tape and see if that helps. If so, your problem is as I described above.

There are many solutions for this ranging from quick and dirty to elegant and expensive.

Good Luck,
John

[mechanicalbrain]

Originally Posted by JVN

Aaaaaactually...
It turns fine, when used correctly. I'm curious what exactly you're talking about? Perhaps some clarification or justification?

Back to the problem at hand...

You mention adding traction to the wheels; this leads me to hypothesize that the added scrub friction during turning has now overcome your turning moment.

The symptoms you describe, also make me believe that you may have a low battery. (Low battery = less motor torque = less turning moment). When was it last charged? (I'm just speculating here...)

If the battery is charged after all and the problem persists; cover the tread on 2 of the wheels (rear or front) with duct tape and see if that helps. If so, your problem is as I described above.

There are many solutions for this ranging from quick and dirty to elegant and expensive.

Good Luck,
John

if you have two wheels on each side connected to a single gear box and your robot is rectangular the wheals will fight each other when turning. its part of the reason car wheels swivle when turning.

[Ben Lauer]

I also am not sure of your problem, but I would check out this white paper....
http://www.chiefdelphi.com/forums/pa...le&paperid=222

It relates to turning in a "tank" style train, and depending on the setup of your wheels how hard it is for the robot to turn.

[JVN]

Originally Posted by Ben Lauer

I also am not sure of your problem, but I would check out this white paper....
http://www.chiefdelphi.com/forums/pa...le&paperid=222

It relates to turning in a "tank" style train, and depending on the setup of your wheels how hard it is for the robot to turn.

Yes... I was just about to reach for it... but Ben is quicker on the draw!

The "Hibner Turning Whitepaper" is a GREAT reference for all those who are trying to understand the physics behind skid-steer turning.

Once again, Chris Hibner we salute you.

-John

[mechanicalbrain]

or you could do what we do now and use omni wheels and turn on a quarter (yes a quarter becuase we didnt have a dime). you can also drive sideways or even diagionally. were still trying to get it rotating while moving in a straight line.

[Ryan M.]

Originally Posted by Ben Lauer

...how hard it is for the robot to turn.

I'll just go off of this, because a lot of the other posts have been de-grammaritized, de-capitalized, and de-spellingatized to the point of unreadability.

I'd make sure that you're sending full power to the Victors in your code.

You can check if it's your code by setting up a quick pwmXX = 255 and comment out your normal speed code. If your robot can't move with that, it's definitely mechanical. On the other hand, if it does move, you know it's the code.

P.S. I'm not saying that your post was unreadable, Ben. I actually understood it.
P.P.S. It's bad to use made up words in posts which critique spelling.

[Oumonkey]

Ok, thanks for the link guys. I'll do that things to the code to make sure its not it
But to make sure what ryan M. is talking about you want pwmXX + pwmXX = Limit_Mix(2000 +p1_y - p1_x + 127) ,well I don't remeber if that limit mix line is exact but whatever, to pwmXX + pwmXX = 255? I just want to make sure my programmins skills aren't to good But the wheels fighting against each other was something brought up but I guess we will see.

[spears312]

Orignally posted by: mechanicalbrain
if you have two wheels on each side connected to a single gear box and your robot is rectangular the wheals will fight each other when turning. its part of the reason car wheels swivle when turning.

Actually, we had a 4 wheel drive train with two wheels on each side, both pairs attached to a AM gear box, and we never had any problems with our turning.

Anyway, Oumonkey, hopefully you will be able to work it out through code. You mentioned something about already lifting up the robot and trying to turn it and the wheels seemed to move in the turning fashion (one side forward, the other backward for tank drive like motion), did they not? If so, and it still doesn't seem to work, try and see if you can measure the speed of the wheels on each side of the robot. Make sure they are going at the speeds you want them to. If not, that may be your problem, possibly caused by a lack of power to the slower motors, a chain problem, a motor problem, or even still be a programing problem.
But, thats just my $.02

[sanddrag]

Originally Posted by spears312

Actually, we had a 4 wheel drive train with two wheels on each side, both pairs attached to a AM gear box, and we never had any problems with our turning.

Just because you bought an AM gearbox doesn't mean you can turn well and just because someone uses a kit gearbox doesn't mean it can't turn well. I guarantee you I can build a robot with an AM gearbox that will not turn at all and I guarantee you that I can build a robot with a kit geabrox that turns like a dream. There is a lot more to it than just the gearbox and number of wheels (final gearing, battery charge, current consumption, voltage drop, wheelbase, track width, ground pressure, tread material, coefficient of friction, interaction between tread and carpet, tread and surface deformation under pressure, etc etc etc.) Such general comments should not be made and no one should rush right out and buy an AM gearbox (or any other gearbox) because they think it will make their robot fast, or because it will make their robot powerful, or it will make their robot turn better. That is simply not the case.

A photograph (top view preferable) would give us the pretty much the whole story on mechanically related problems. Once we get that, we might be able to rush right out and say "yep, that's why you can't turn." Also, if there is anyway you can get a video of the attempt at turning, that would be awesome. Picture is good but sound and picture is much better.

[Oumonkey]

I might not be able to get a pic soon. But the more I think/talk about it I think we are getting to much traction. and if all fails we should be able to put it back to 2 wheel drive pretty simply right? Well I talk with Brian, the electronics guy on our team and he said he will get on here as soon as he can to answer the question I can't. Hehe I brained my hurt

[sanddrag]

Originally Posted by Oumonkey

I might not be able to get a pic soon. But the more I think/talk about it I think we are getting to much traction. and if all fails we should be able to put it back to 2 wheel drive pretty simply right?

Again, there is a heck of a lot more behind the way a robot turns than the number of wheels or the number of wheels driving etc. I cannot say "putting it back to 2 wheel drive" will fix your problem without seeing your setup . It is even probable that it will hurt performance even more, who knows?

Let's start with this:
1. How many wheels does the robot have?
2. Where are the wheels positioned?
3. What kind of wheels are they, and what exactly is the tread surface material?
4. How many of the wheels are being driven and which ones?
5. How many motors (and which ones) are driving the wheels?
6. What top speed (in feet per second) is the robot geared for? - if you don't know this, what gearbox is it using and what size sprockets are you using and what diameter wheels are you using?
7. Are you sure the battery was charged and if you are sure, how/why are you sure?

I know you probably answered some of these things before but let's get all the facts down in one place and then we can start to analyze this. Trying to help you we are kind of in a tough place here because we can't diagnose the problem with something if we don't really know what exactly that "something" is.

Also, I would recommend getting a few pictures of your robot to keep for yourself. You built it, you should be proud of it, you should have some pictures of it. (If there's anything I've learned in FIRST, it is do not put your eggs (in this case pictures) all in one basket). Robot pcitures are good for hanging on walls, setting as background images, browsing through when bored, taking up unused hard drive space, and posting on CD!