Reported electrical problem for you to think about!

[Al Skierkiewicz]

Everyone,
A recent electrical problem crossed my path. I thought you should know what took place and see if it has a familiar sound to a problem you have this year or had last year. The report and cause and effect are quite interesting. Here is the report...


"I thought I would fill you in on an issue we encountered last night on our proto robot. We are testing various portable vacuum cleaners for ball control. We strip the blower and motor from the vacuum and mount the motor using a hose clamp to the frame. Our Crio is mounted on lexan strips above the frame to insulate it. We have a kicker that is able to kick across the entire length of the field so it is pretty violent when released. Surgical tubing is the stored energy, stretched by pneumatics. Everything was fine until we added a different blower/motor. Then about every third kick, the radio would cut out. Further investigation showed that the radio and Crio would lose LAN connection. Everything on the robot seemed normal, power stayed up, no other lights flickered. We monitored the 24 volt output of the PD and the 12 volt output for the radio and saw no fluctuations. We even took to beating on components with a hammer to simulate a kick with no effect. Our proto is put together with 80/20 and someone noticed that during a kick there is sufficient movement in the frame and the lexan strips holding the Crio, that the Crio would come in contact with part of the drive module. In and of itself this shouldn't have caused a problem. Tracing wiring and checking for shorts to frame in other components, we found one of the vacuums has a motor (ed. non-KOP) lead connected to the case internally. When the kick occurred and the Crio made contact briefly, the circuit was closed through the frame from motor to Crio. Should this effect occur with other robots, the cause may be intermittent contact of the Crio frame and another sneak path in wiring. As the motor is downstream from a speed controller, the normal ohm meter test for frame to power supply leads showed no ohmic contact. In other robot systems, this could be a cause of failure when the robot is running in one direction but not the other. It may only manifest itself when the motor lead that is contacting the frame is controlled to be outputting +12 volts."

I suspect that the current flow from motor to frame to Crio was able to raise the common lead (Crio case) to a sufficient level to disrupt the Crio or it's communication electronics. The report does not say that the Crio was rebooting and the time that would elapse between failure and return of control was about what would be expected for a loss of power to the wireless adapter not the interval for a full Crio boot. Please note, there was no change in the primary or secondary power supplies and no flickering lights on the PD, Crio or sidecars. Only the loss of LAN was noted on the wireless bridge. If this sounds familiar, check for motor wiring that might be contacting the frame. As reported, this could occur in only one direction of the suspect motor. In this case the vacuum was only running in one direction for obvious reasons.

[Jon Stratis]

Interesting problem, thanks for posting it. Fortunately for all of us, we don't have that problem with the KoP motors - but it's very, very easy to have an accidental short to the frame somewhere (like a pinched wire) that could cause this same behavior. This only goes to show the importance of checking (early and often!) that your frame is electrically isolated from both +12V and ground.

Intermittent contact with the cRio casing as reported here should be fixed if you notice it... but everyone should also be aware that you aren't going to be able to find and fix ahead of time every little problem on the robot. Something like that intermittent contact may occur during a match without your knowledge. If your frame is isolated from your electrical system, however, it shouldn't cause any problems.

[IndySam]

We had a strange problem occur in a similar fashion.

We have a vacuum powered by a FP motor. We were practicing in a hallway and every time the vacuum would grab onto a ball we would loose communication. Wait a bit and the communication would re-establish itself.

We took the bot back to the shop and could not reproduce the problem.

We were practicing in a different area on Saturday and never had the problem even with a very dead battery.

We have not had time to go back to that hallway to try and reproduce the problem.

Our cRio is mounted in a lexan box and I will make sure to check that the mounting bolts or something else in not contacting the frame.