Important note about USB-6009 input voltage limitations

[jerry w]

one more thing to be aware of.
the data sheet on this device lists the max voltage at +- 20 volts.
on the robot we have Victor controllers that pulse 12 volts to the load. a load is typically inductive. the result can by rather high voltages on the output terminals of the victor.
although the victor has suppression to avoid damage, it does not try to protect other devices on the line. it would be good practice to avoid hooking this DAQ to anything on the 12 volt side of a controller.

Jerry W

[Danny Diaz]

Originally Posted by jerry w

... the data sheet on this device lists the max voltage at +- 20 volts ... it would be good practice to avoid hooking this DAQ to anything on the 12 volt side of a controller.

Remember that the +/- 20V is only for differential mode measurements (using 2 wires with the measured voltage being the difference between the voltages on the wires) - a single-ended mode measurement (one voltage with the reference voltage being ground) can only handle up to +/- 10V. Most measurements should be made in differential mode as this mode is less prone to problems with differing ground references, with single-ended measurements only made when you're measuring a floating source, or a source with no connection to ground at the signal source.

-Danny

[DonRotolo]

Would it be a good idea for teams to install some kind of overvoltage protection scheme, at least until they understand what they are doing? I am thinking of a zener diode, but that might not reach fast enough. Any other ideas?
Don

[Al Skierkiewicz]

Dan,
In looking through the on line manual for the 6009 it is possible to condition the signal externally with clamp diodes and noise circuitry to give a good signal at the input of the 6009. I have to agree that measuring the output of the Victor feeding a motor will expose the measuring device to voltages in excess of 100 volts with extremely fast rise times particularly during direction reversals. Simple passive techniques can put these signals under control. If you are interested in data acquisition on these excursions, a simple balanced "H" pad would give the scale you need to accurately convert the high voltage pulses to something the ADQ can handle. (at the sacrifice of low voltage resolution)
However, there is not much information to be gleaned from a simple voltage monitor on the output of the speed controller. Monitoring the input current to the controller tells so much more about the motor behavior and loads. For instance, monitoring the input current on a dual Chalupa feeding a transmission will tell if there is a shared load on both motors, a mechanical problem with coupling on one motor compared to the other, the failure of one motor/controller set, changes over time in the system and accurate comparison between drives on both sides of the robot.

[Larry Lewis]

I believe you are right. A zener diode will offer the best protection in parallel with the motor. It has a higher amount of current draw, therefore the voltage cannot build up across the motor. If you search digikey you may actually find a selection of surge protection diodes. But you may be able to get by with a regular zener.

[Danny Diaz]

The hardware engineers that designed the NI USB-6009 tell me there's already a small bit of protection in the 6009. To supplement this protection, they recommend the circuit I have attached - really it's just a matter of placing 2 diodes on each input, in the orientation specified (just make sure the Vcc for the circuit is the 12V power rail of the robot, NOT of the DAQ device). This allows a surge to drain into the power rail, where you'd most likely just momentarily charge the battery with the power surge. The Victor probably already has this type of protection, so placing the DAQ device inputs closer to the Victor will help if you're just super concerned.

For the diodes, just sizing the diode for the steady-state current of the motor should be sufficient.

Realize this does not work well if you're using a switched power supply, only a battery (thanks to the impedence of the battery!).

-Danny

[DonRotolo]

Originally Posted by Danny Diaz

For the diodes, just sizing the diode for the steady-state current of the motor should be sufficient.

Yow, for my CIMs that might be 20 to 30 Amps! I think I'll take a chance and use 5 Amp diodes, it is unlikely that the microsecond surge has enough energy to cause damage. OK, maybe I'll actually TRY it first...

Thanks for the nice circuit Danny, much less expensive than zeners. Also thank the HW guys for us.

Don

[Al Skierkiewicz]

Dan,
The circuit Danny has included is standard protection for both inputs and outputs. Any voltage that exceeds the power supplyrail by more than 0.6 volts (the diode forward voltage) will be connected to the power supply rail. However, the diodes normally used, 1N4000 series, are usually too slow to allow really short and big spikes from getting through. There are several MOV devices that can really eat up the spikes (and are non polarized) available from Digikey and others. I recommend both solutions. You don't want to destroy the ADQ.