I have read a little of the PID stuff and i sort of get it. I am wondering a few things.
How does your team power the rotation of their arm and get it to stay in a lifted position?
Do you use the window motors or some of the smaller motors to power its rotation and lifting?
is there not a risk of burning up a motor if just barley enough voltage is applied to keep an arm lifted?
This is my second year in FIRST… Im not sure that the robot my team produced lastyear should be considdered something FIRST worthy. I really need help or pictures.
You can write a very simple piece of code that will do a decent job of keeping your arm up. Store your target potentiometer position, and set your motor output to 127, plus the target, minus the current position. (Or the other way around, depending on which direction is up.)
A window motor can lift a small arm. For bigger arms, you’ll want to look at the RS-540, the FP, or the CIMs. The Good Practices document will tell you what torque is available without additional reduction, and what your peak power is.
Some motors burn up, others don’t. FPs have a reputation for letting out magic smoke at stall, whereas CIMs have a reputation for being essentially indestructible.
I don’t know of any cases of motors burning up without being at or close to stall. I do know that the Globe gearboxes start to break if you sideload them.
Thanks, that’s reassuring! We’re aware of the side loading issue with the Globe, and I think we can support the shaft properly without too much work, after thinking about it for a week.
Make sure to set the victor your motor is hooked up on to the “brake” setting; this will make it more difficult to back drive and will make the motor better keep your current arm position.
Also, the type of system you use to move your arm can automatically help keep your arm stable. Screw drives are basically impossible to backdrive, but you’re limited to less than 180 degrees of rotation with them.
Using some of these things may make it unnecessary to keep a PID loop in programming monitoring the position of your arm.
We are working on different designs using either the globe, BB, or window motor(s). The globe looks to be the closest to what we need as far as torque capacity and rpm, but the others are there as backup if it doesn’t work.
And to pile on with another idea: A common solution to preventing back drive is to interface your motor with a worm gear, if possible. You just need to make sure whatever motor you use can withstand axial thrust. I believe some of the KOP motor transmissions utilize worm gears (i.e. the window motors)…
The window motors do use worm gears, although they are a bit weak (torque) compared to the globe and BB motors. But two window motors can work together if needed.
Or using a “big” motor, CIM as an example to do the main drive of an arm and use the window motor for its anti back drive feature in conjunction, it won’t add much to lift but locks position pretty well.
BTW globe motors back drive pretty easily with seemingly little forces on them
The plan is to have the arm motor mounted “up high”, so weight is a big concern…less weight up there means less tendency to tip. Adding a remote drive system to it also adds some weight “up high”, but not as much as putting a CIM up there.
But that is an interesting idea of combining two different types of motors to drive the arm, thanks!
BTW globe motors back drive pretty easily with seemingly little forces on them
We did an experiment with a Globe using a wrench and a scale…the Globe will back drive with about 1/5 the load we plan to apply to it, so this could be a serious problem.
We used the two motor solution a few years back with an extendable ladder like structure. The trick is to make the window motors holding direction so that you are driving the armature toward the back of the motor. I.e. hold the motor in your hand and if the output shaft is facing you and the motor part is to the right and up, the stopped load you would want turning the output shaft clockwise. We have had very bad luck with the holding load the other direction. The other thing is to match the free speeds of the two motors you are using. The published info is good. Hooking them up and counting revolutions is even better. In 2005 A Globe and the Keyang window motor was a good match. We had a coupler with 16 teeth for 25 chain machined the linked the two shafts together. It was the best working part of the arms on that robot. And it stayed where you left it with out driving power. This year you get two Globes and two Keyangs. I will attach the inventor file. We did have some issues with the machining and had to do some of it by hand.
Biff
Thank you for all of your input but my question has still not really been answered.
do you guys build your own gearboxes for the globe motors?
Do you use any special programming to make the motor stay slightly engaged to keep the arm up?
Do you use globe motors, van door, or window?
This is our team's robot from 2005. On the first stage of the arm we used two of the 6v FP motors (they messed up that year and gave a bunch of teams the wrong motor). On the second stage we used a single van door motor. We used cable and bicycle wheels to gear down the main arm about 18 to 1* and we geared the small arm down to about 9 to 1*. It had no problem holding a single 9 lb tetra 13' out. Here is a link to a picture of the robot.
As for programming we used a potentiometer on each arm and compared the value they gave us to a target position for the arm. Based on that information we passed a value to the motor to move.
If you have any questions please feel free to ask.
-Dan Karol
*numbers are a guess and may not be what we actually used.
While some of the motors may “take the heat”, its a good design practice to attempt to minimise the load on the motor when the assembly is static. Counterweights help greatly with this. Another option is gas shocks like the ones used to hold up car trunks and hatchbacks. They not only put out alot of force but provide some dampening. Worm gears are definately a good idea as well. I would highly recomend them. I should note that the gearset in the van door motor is NOT a worm gear set. It is a healical gearset which can be backdriven.
Also, if you plan on using PID, you must make every effort to minimize the amount of backlash in your system or else put the sensor directly on the motor. Chains are especially bad about backlash unless active tensioners are used.
Keep in mind that when you need high-torque motors (like in a programmed non-back drive mode), it helps to use two of them, if possible. If you wire the two motors together (+ to + - to -, then the combined + and - to the victor controlling them), now you are taking two circuits and combining them into one, distributing the power evenly. Not only will the two motors operate in perfect harmony, it will also multiply your torque and stall rating by two.
