Black Jaguars on old IFI controls

[Racer26]

In case anybody was wondering, last night I performed a little experiment.

Our robotic forklift cart uses a power wheelchair as its base, and the controls were replaced with a 2004 IFI system, minus radios. Its motors were 24V.

We've been running the system off two deep-cycle marine batteries in parallel providing 12V at a combined rating of approx. 150AH. Running the drive motors at 12V, the cart has worked well since 2008 when we built it, but its always been a bit slow (it moves at a walking pace). Knowing that the motors were designed with 24V in mind, and knowing that the cart already has 2 12V batteries on it, (as well as a trailer we may be developing for it to extend its range, and provide an additional 350AH or so of batteries, along with inverters and so forth to run controls and tools) and knowing that the black jags we have will happily operate at 24V, we decided to give it a try, mainly because it would improve the power and top speed of the cart, and with the field being 1/2mi from the pit in Atlanta, we thought the extra speed might be nice.

I hooked up two robot batteries in series for the purposes of this test, powered up a black jag, and fed it a PWM wire from the 2004 IFI system on the cart. Works flawlessly. (I'm not sure why we have to differentiate Jaguars from Victors in the cRIO code, but I know it has something to do with the refresh frequency (200Hz on Vics, 15kHz on Jags?)). Does this mean that while the Jag is happily running, its just not running to its full potential?

For simplicity's sake, I tested the jag at 24V with a CIM hooked up to it (yes, I realize the CIM is a 12V motor, and running it on 24V could damage it. This CIM had already been damaged to the point that we wouldn't use it on a robot.) Wow. CIM's make a TON of torque on 24V as compared to 12.

In the end, we decided NOT to make the change, as it would require considerable re-wiring of the cart to provide a separate 24V rail for the drive wheels, while maintaining the 12V system for the controls and other accessories, as well as making the integrated charging system considerably more complicated, and considering we only had 3 days until the trailer has to be packed up, we decided there wasnt enough time to make the change and test everything out properly.

[Russ Beavis]

You powered a CIM using a Jaguar with a 24V supply? That would be one heck of a startup current. I'm very happy to hear that the Jag handled that transient event.

Sounds like a great exercise in electromechanical engineering!

Russ

[Joe Ross]

The main difference is the maximum PWM pulse length. On the jaguar, it's longer then the IFI control system supports. When you give full speed on the IFI controller, the jaguar will only be at about 2/3 speed. Be sure to calibrate the Jaguar in accordance with the manual and you will be able to get the full output range.

The PWM chop frequency (120hz for Victor and 15khz for Jaguar) has no bearing on the control system. The maximum PWM update rate (approx 100hz for Victor, 200hz for Jaguar) does have an effect in that the cRIO will update the Jaguars faster then the Victors, but it's inconsequential for the IFI control system because it updates slower then the max for both.

[EricVanWyk]

The real reason that you have to select Jaguar or Victor is not because of the differences in chopping frequency (~200 vs ~15000 Hz), but because of the differences in the signal pulse width.

Victors use 1 to 2ms, whereas Jaguars use something about 50% wider. ((I can not recall the exact numbers right now)). This allows the Jaguars to get more precision from this interface.

As a result, using the wrong setting has poor results. Using a victor with jaguar settings usually results in losing low speed control (it goes to +100% or -100% too soon). Using a jaguar with victor settings usually results in losing high speed control (it never gets to +/-100%).

This is probably what saved you in your experiment. Since the IFI system assumes it is talking to a Victor, the Jaguar was interpreting full power as ~70-80%. So, you still got more more power than a 12V system, but not the full 24V power.

If you want to try it with full power, just recalibrate the jaguars as described in page 13 of 33 in the Getting Started Guide .

The Jaguar's internal fault detection and prevention will probably prevent the CIM from dying, but I'm not sure. The CIM simply isn't designed for 24V operation. You will know it is protecting its output if it turns off: command it to 0% to reset the protection. Be ready to sacrifice the CIM, and more importantly BE CAREFUL. You are playing with a fair amount of power.

[Alan Anderson]

Originally Posted by 1075guy

(I'm not sure why we have to differentiate Jaguars from Victors in the cRIO code, but I know it has something to do with the refresh frequency (200Hz on Vics, 15kHz on Jags?)).

It's actually the input range and "neutral" point of the control pulse width. The IFI control system theoretical neutral value is 1.5 milliseconds, corresponding to a value of 128. Jaguars are designed to that standard. Factory-calibrated Victors tend to have their neutral deadband centered closer to 133, which can cause asymmetric forward/reverse power unless you recalibrate everything or program things to deal with the offset neutral point. The WPI library for the NI control system takes that difference into account.

[Racer26]

Originally Posted by EricVanWyk

The Jaguar's internal fault detection and prevention will probably prevent the CIM from dying, but I'm not sure. The CIM simply isn't designed for 24V operation. You will know it is protecting its output if it turns off: command it to 0% to reset the protection. Be ready to sacrifice the CIM, and more importantly BE CAREFUL. You are playing with a fair amount of power.

I was fully prepared for it to ruin the motor. This motor had been on the 2010 Robot as our winch for lifting. It had spun itself in its mount and twisted and stretched the wires all the way back to the victor it was powered by. As a result, it couldn't handle the current any longer to perform in its needed function, so it was replaced. As a quick test though, it was an easy motor to grab.

Ultimately, If my understanding is correct, a DC motor is a DC motor, while it may be designed for a particular voltage, it will happily run on more or less any DC voltage, however the current draw and mechanical power produced will change, with the obvious caveat that running any electromechanical device outside of its intended operating characteristics is playing with fire.

[EricVanWyk]

Originally Posted by 1075guy

I was fully prepared for it to ruin the motor. This motor had been on the 2010 Robot as our winch for lifting. It had spun itself in its mount and twisted and stretched the wires all the way back to the victor it was powered by. As a result, it couldn't handle the current any longer to perform in its needed function, so it was replaced. As a quick test though, it was an easy motor to grab.

Ultimately, If my understanding is correct, a DC motor is a DC motor, while it may be designed for a particular voltage, it will happily run on more or less any DC voltage, however the current draw and mechanical power produced will change, with the obvious caveat that running any electromechanical device outside of its intended operating characteristics is playing with fire.

Your comment matches my understanding pretty well. Also, watch for maximum RPM. We've had motors rip themselves apart angry-centrifuge-style when over-volted.

My safety comment was aimed more at personal safety than motor safety.

[Racer26]

It would be interesting to characterize a CIM motor running on 24V for extended periods of time, to see the power and torque produced, as well as whether or not they'll explode, implode, melt, or otherwise be irreparably damaged. I mean, we all know they produce 343oz-in of torque (about 1.78 lb-ft) @ 12V, I wonder what the relationship is if you bump the voltage to 24V. Do you get 686oz-in? My observations of the motor I was playing with certainly seemed to suggest that the motor was generating ALOT more torque under 24V, but was in no way an accurately measured change.

I wouldn't recommend over-volting a motor without some safety precautions, and a willingness to break it rather handily, however, under-volting motors (as we are with the 24V motors on the cart) seems relatively safe, as long as you're not asking them to operate near stall, where they'll heat up more and probably melt the windings, causing shorts and all sorts of fun stuff.

I imagine a motor ripping itself apart "angry-centrifuge-style" is rather amusing to watch, if potentially dangerous and destructive.