I have been helping my team out with the programming and electronics for the 2004 First competition. Initially, we tested our programming code with last year’s bot while the 2004 robot was being built. We programmed some autonomous code and it worked great! However, after completing the build of the 2004 robot and executing the same exact code on the new bot, we got extremely different results.
Both robots used the 2004 innovation first controller and drill motors for the drive train. However, the old robot used Victor 883’s and the new robot used 884’s. Thus, we swapped out the 884’s and put the 883’s into our 2004 bot and it worked fine!
I read on these forums that the 883’s cannot be used in this year’s competition, is this true? I couldn’t locate that rule anywhere on the First site. Also, we hooked up the 884’s up to an oscilloscope and determined that the Victor 884’s were only changing the output voltage from 127(0V) to 180 (12V). So in our code, it doesn’t matter if we output 190, 210, 220… All of these values cause the robot to drive forward at full speed!
If anyone has any advice, it would greatly be appreciated. Our robot is pretty hard to control in this situation, so it would greatly improve our bot!
A. Yes, it is against the rules to use Victor 883’s instead of 884’s. Check the inspection page.
B. Put in the 884. Press the calibrate button. Wait a while. Then use the code to send a 254 signal and 0 signal to the PWM output. Let go of the calibrate button. Do this for each of the 884’s.
If you calibrated the Victors several times and still get these results, you may have a bad Victor. Is this happening with all 4 of your speed controllers?
what is your criteria for full on looking at a scope trace? does it go to a flat line +12V?
did you have the motors connected at the time?
to calibrate you need to push the button until the light blinks, then continue to HOLD it while you output 254, then 127, then 0, then 127
then release the cal button
it is possible the victor is messed up - they have not given those out for 2 years now, so the old ones might have been through some real abuse that you dont know about.
We had tried 3 different 884’s and they seemed to each act a little differently. One of our 884’s was found to be bad by another team member. We calibrated the 884’s and 883’s with the joysticks a few times and it didn’t seem to make a difference… We used the same method of calibration for the 884’s and 883’s. Hold the button down…full forward, neutral, full reverse, neutral…release the button.
We looked at both the DC level and the AC signal of the motors. We probed from the negative (black lead) to the positive (red lead) of the motor using a 1X probe. The motors were connected and we were changing the speed in user mode with the joysticks. The output voltages were flat lines with some occasional harmonics in them for all speeds…
I think your instrumetation is off - the victors work by pulsing the 12V to the motors to contol the speed - when you are going slow it pulses it on briefly, at half speed the pulse is on half the time, and off half the time - and at full speed the victor outputs a solid 12V
but you need to have your scope set up so you can see the pulses - if the X axis (time base) is too fast you will only be seeing part of the pulse, maybe the 12V part, maybe the gnd part, or maybe some back emf from the motors.
try it with the motors disconnected, and look for the pulses - adjust your sweep rate until you can see 2 or 3 - then as you adjust the pwm commands to the victor you will see the pulses changing in width.
you might also need to play with the scope trigger - make sure its on the same channel as the probe, and make sure the trigger level is around 6 volts.
there is one other difference - the older victors have a wider dead band, so using closed loop feedback (PID loops) is thrown off a little by the 20 or 30 pwm values around 127 that output nothing.
We will have to try using the scope again with the motors disconnected… However, besides calibrating the Victors, what else could be done to make them work correctly? We calibrated the 883’s the same way, and we could control them fine (127 through 254), but the 884’s only took 127-180 every time.
For your information, Team 177 has never needed to calibrate a victor 883 or 884 since they came out quite a few years ago. We use them right out of the box, trim the joysticks for 127 counts in “neutral” and live with any small loss of performance due to the joysticks not giving a full 254 or 0 at the end of travel.
You may be opening a can of worms by doing calibrations.
Originally, the problem with our 884’s was a slight tap of the joystick when the default code was loaded would result in our robot flying forward out of control. Therefore, we tried calibrating the victors again, however, this didn’t change our controls at all! Anyone have some troublehooting advice?
Supposedly the victors are pre-calibrated. However, sometimes they aren’t,and therefore you will get very strange behavior from your victors.
Try this to troubleshoot - print the pwm values to see if something is wrong. If you don’t see a problem, try outputing PWM values from the code and slowly incrementing it. See how the motors and speed controllers behave.
While I highly doubt that a proper calibration would ever cause these problems, the factory calibration is stored on the victor permently. You can revert to it fairly easily.
To reset the Victor to it’s factory calibration, and start from scratch, you need to power it with the cal button down. In detail:
Power down the bot
Press and hold the cal button down.
Power up the bot
When the indicator light on the Vic flashes green, release the cal button.
You now have the factory calibration, which depending on the sticks you use can give you close to full range. I am not sure about this years sticks, ours were way off and peaked early. Use an older model if you have it, I trust them much more.
I’ll just note here that the 883 and 884 are very much the same speed controller. The diffrences are in the deadband around 127 (4% on the '4 and 10% on the '3, IIRC), and a 40 amp rating on the 884 vs the 883’s 30 amp. Otherwise, they are pretty much the same. I don’t really understand why they would give you such funky results. The only thing I can think of is that your calibrations are off. When you return the stick to center and release the cal button, you have to be extreamly careful the stick comes directly to center and stays there. This years sticks are realy bad about this.
This one has me pretty stumped. I would be amazed if you got several bad victors all at once. Heck, I would be surprised if you got one- Innovation First makes some high quality stuff.
Perhaps all the victors were victum of some freak electrical storm?
check the joysticks you are using and make sure that the center wheels next to the sticks on the base are in the center position. If you can not tell if they are centered, hook them up to your oi, and connect that to a computer via the dashboard port. Launch the IFI dashboard program and see what the values you are getting for center on the joysticks, maybe what you think is 127 may really be a really different value. Keep us updated on your tests and if you can fix the victors… Good Luck.