Creating PWM signal for victor

[neg0riz0r]

Hello,

I am trying to figure out a way to get a possibility to quicktest all our Victor 884s and 883s without using the cRIO or any other FIRST hardware.
So I've built a simple timer circuit which gives me a
PWM input rate T = 22 ms
and a PWM puls rate of 1.1 - 2.3 ms (controllable with a potentiometer).
I monitored that signal with a storage oscilloscope to double check this data. The values are correct.
According to that information, it should work fine:
http://www.vexrobotics.com/products/...ontroller.html


But, with the Victor 883 it doesn’t work at all. It blinks orange, which means no PWM signal.
And with the Victor 884 it only works for reverse and stop. If I come to the puls rate > 1.5 ms - which should be forward - it starts blinking orange very fast (faster then the slow 'no PWM signal' blink)

Do you have any idea what I am doing wrong?

neg0riz0r

[NOV8R]

You need to boost the pwm signal to work with the victors. A few years back we built a control box using an airplane r/c transmitter and receiver to test drive our bot. We bought special pwm cables that had a built in amplifier to make it work. I can't remember who we bought the cables from. In the jaguars the pwm input is opto isolated so you don't need as high a signal strength. Hope this helps.

[Mark McLeod]

Here's our PWM driver: http://www.team358.org/files/electri...MGenerator.pdf

The circuit diagram is at the end if you want to compare circuits to see what differs.

[neg0riz0r]

@NOV8R: Yes, I realized with the oscilloscope that the signal is not stable when the motor is turning. So some kind of current amplification was my idea, too. But I don't know how to do that :/

@Mark McLeod:
I tried that before. Had similar problems. Signal not stable. Does this really work for you with victor 883 and 884 ?

[Mark McLeod]

One of these circuits was used last night to run an 883 driving a window motor on a test frame.

I suppose your 883 blinking yellow light could be caused by a missed/bad contact at the PWM connection, since those are easy to miss inside that cover.

Do you have access to any old IFI robot controllers you could borrow for testing your Victors?

[neg0riz0r]

No I don't have access to those.
There is no contact problem at connecting. Thats for sure.
With the Victor 884: One directions works, the other not (here we have also the blinking thing)

[Al Skierkiewicz]

Originally Posted by neg0riz0r

So I've built a simple timer circuit which gives me a
PWM input rate T = 22 ms
and a PWM puls rate of 1.1 - 2.3 ms (controllable with a potentiometer).
I monitored that signal with a storage oscilloscope to double check this data. The values are correct.
According to that information, it should work fine:
http://www.vexrobotics.com/products/...ontroller.html

neg0riz0r

It may be that you are controlling the wrong parameter. The pulse repetition rate is 16ms or about 60Hz always. You need to vary the pulse width to control the Victor. Please take a close look at Mark's reference on page 2. The width of the pulse changes from 1ms wide to 2ms wide. At 1ms wide the Victor is in full reverse, at 1.5ms the Victor is at neutral and at 2ms the Victor is at full forward. At 1.5ms the signal will remain low for another 14.5ms. You are able to control the Victor in reverse because your rep rate is close enough for the Victor to decode useful data but not as you advance the pulse to 2ms.

[neg0riz0r]

I thought something like that, but in the pdf I posted in my first post it says:

PWM Input Rate: 15 - 30 ms

According to that my T=22 ms are ok. Hm. Anyway, I will look for another capacitor in series, that brings down the T a little bit. I will tell you if it worked.

[sparrowkc]

The "filter" in the middle of some pwm cables just inverts the signal twice, forcing it to a valid logic level. I wouldn't guess that it can supply any more current than your timer chip can, but if your signal isn't clean it might help.

[neg0riz0r]

OK now I have exactly 16 ms.
I was wrong with my statement concerning the Victor 883 before. I really must have done something wrong while connecting it. Sorry at this point. Because now the 883 runs perfect with the circuit. All directions. Full Speed, Stop..everything is fine.
And it is not important if T = 16ms or 22ms. Both works.

But the Victor 884 still has the same problem. Reverse works to Stop works, but the area above 1.5 ms (which is forward) doesn't work at all. And only in this area the signal becomes unstable, too. Maybe it is a problem with the victor itself. But reverse works, so I just assumed that it is fine...
I will have to get another 884 to check that.

[Mark McLeod]

I have several Victors that only work in one direction, so that's quite possible.

[neg0riz0r]

Oh? Thanks for this information. Thats really helpful. Does your start blinking orange, too, if you apply the signal for turning in the not-working direction?

[Mark McLeod]

I don't remember if any of them blink. I think that on some the light just goes out, but I haven't tested them in the broken direction in a while. They get labeled and put on a shelf for later repair if necessary.

You just need a secondary way of testing them that confirms the problem before you can be sure there is a problem. That's why I suggested an old IFI controller.

They're still useful if you have a one-way device, such as, a flywheel or shooter mechanism.

[neg0riz0r]

ok circuit works with different victor 884. This one is apparently somehow damaged. Thanks for your help.

[semiferger]

The Victor 884 has a deadband with respect to the PWM signal, which is approximately 117 to 137 . Any PWM signal within the deadband results in no output.The PWM deadband accommodates joysticks that do not return to the exact same centre. The PWM deadband effect can be overridden by Robot Controller code if needed.