pic: Testbench with PDB backpower



This was probably a bad idea but all the batteries were at an outreach event and we needed to get Beta Task 2 done.

Why did you choose to back-power the PDP instead of just plugging the power source into the input like a normal battery? Weren’t you afraid of the possibility of breaking the many expensive pieces of electronics by using them “off label”?

Mostly because I didn’t have connectors to hook up the power source to the PDP input terminals. I wasn’t afraid of breaking it because previous experience has taught me that all of the 12v and ground rails were connected directly to each other. The only thing I was worried about breaking was the power supply. (which I may have done something to with the power surge from a CIM stopping)

But it worked well for the test!

They cannot all be connected directly to each other: each circuit has its own breaker and current meter. Running power backwards through the breaker would be of no consequence (as your power supply is probably not capable of enough current for long enough to trip a 40A breaker), but running current backwards through a current meter designed for polarized current certainly could.

I recommend verifying that you check on circuit you used as the “input”, both to verify that it still works, and whether its current meter still functions properly before you build a robot with that PDP and have incredible fun troubleshooting the problem.

(Besides, you wouldn’t have needed any more hardware than you have in the picture - just loosen the input 6AWG feeds into the PDP a bit, slide in your smaller supply wires underneath, and tighten.)

Is there a way to check the current meter without having to put much in the code? I assume it’d be easy to display it in SmartDashboard or in the web config.

Also, maybe I should just make a straight-up test program since I highly suspect we’ve damaged a few current meters already due to an unrelated thing. From the manual:

Warning: Inductive loads (motor, compressor) must have a power management device(motor controller, PCM, spike) between itself and the PDP. No inductive loads (motor,compressor) may be directly connected to the PDP channels as this can damage current sense circuitry.

We have probably done this with a prototyping setup involving a dipole switch connected as an H-bridge to a motor. It might be safe though, if the issue is only referring to the reverse voltage that a motor generates when slowing down.

No code necessary. Go to the roboRIO web interface, pull up the PDB, and I think you click test. I’m going from memory here. You’ll see the estimated current value for each circuit. Change the current going to that circuit and see what changes.

You can also see it in the log viewer in the driver station: that’s incredibly useful for debugging stalling as well.

I thought about this problem too, but then I remembered that any type of back current from the robot moving also generates this current, so hopefully the PDB is designed to handle running current backwards.

The problem of course being that microcontroller reading the voltage drop across the shunt resistor will have negative voltage at it’s inputs. However, I bet it goes through an amplifier before going to the microcontroller, and those often handle negative voltage just fine.

Back on topic though. We have a similar test board using a 12V, 30A power supply that is wired into the battery input. It’s been working great for testing anything from robot code, working on vision processing, or just trying new things.