pic: Sparked SPARK



We discovered this after our first quarterfinal match in Peoria. 4500 loaned us another one though. [strike]The lesson: Don’t screw in your terminals too tight![/strike]

Edit: As of now, we are unsure of the cause.

Update: REV did replace the SPARK, in case any future people see this thread. None of the other controllers failed like this and the problem was on our end.

Hi bobbysq,

Yikes! Send me an email at contact[at]revrobotics.com. We’d like to take a look at the SPARK and send you a replacement since this type of failure is unique and we haven’t seen it before. To be honest, we haven’t seen many SPARK failures at all.

-David

It actually SPARKED

Team 6640 had a couple of Sparks “spark” in Arkansas at the Rock City Regional. They also earned the Rookie All Star award.

I only know because we gave them the two sparks we got in the KOP this year as replacements.

Reach out to them if you need more spark failures to look at.

Thanks! However, we suspect it may have been a bad connection on our part and the original text of the post was inaccurate. (I was trying to edit it in time, you guys responded so fast!) We’ll still send out an email, (assuming that the effects are unique) and if you want, I still have the failed terminal that somehow desoldered itself.

Once again, thanks for doing this. The SPARK is still a good product, and the other controllers have survived through more intense situations.

We will reach out to them.

Yes, we would still like to get it back to inspect if we can.

A note for everyone reading this. If you have any spark (or other REV products)failure Pretty Please reach out to us. We try to continuiously improve all of our products and if we don’t hear about issues we can’t do anything about them. In our experience most of the spark failures are due to things like hooking them up backwards or other electrical issues on the robots, but we still try to log every one so we can do the engineering analysis required to see if they are user errors or manufacturing issues.

Thanks,

Greg

Did the the metal piece (terminal) that is soldered to the board come off with your wire and lug? Can you post a picture of it? Was the screw holding the lug onto the terminal of the SPARK loose? What was this particular SPARK used for in your robot? Was it operating at high current for a long time?

We had a similar failer at out comp in Duluth, however we believe it was a result of the spark previously being wired in reverse by our freshmen who where doing wiring practice (they would turn on the board, listen to the 40 amp fuses click and wounder why the CIM wasn’t getting power, I heard the clicking across the shop about 5 minutes into their testing and came over to fix the problem, in hind sight we shouldn’t have used that spark on the final robot, live and learn) the spark has now gained a place of honor on the back wall of our pit. :stuck_out_tongue:

I forgot to ask the OP if the SPARK was able to function correctly for some period of time.

It is a coincidence the OP made this post yesterday. I was asked by a mentor for a team that works at the same facility to look at a 45 A Anderson connector that melted. My diagnosis was that one of the contacts was deformed (probably when being crimped) and the contact pressure with the other contact was reduced, leading to high contact resistance and heating at high currents.

I didn’t have time to take a picture of the wire, since we had to work quickly. The terminal did desolder itself and likely made that burn mark. I do still have the terminal available. Nothing special with that though, it’s just blackened. We used it on our drivetrain with a 40A breaker. (the “5A” on the label refers to half of a PWM Y cable on port 5)

It lasted from it being put on the robot during build season to yesterday.

I am presuming that by “It lasted from it being put on the robot during build season to yesterday.” you mean that it did function properly for some time. If so, it is most likely that the screw holding your crimp lug onto the - terminal of the SPARK was not tightened sufficiently since the area around the + terminal looks fine. If this is true, Rev might find that the SPARK will still function. Your screws should be tight enough that it should not be possible for you to make the crimp lug pivot around the screw with your fingers.

If it had been damaged by a reverse connection (as Ginger described) causing damage in the SPARK which then causes a short circuit when you connect it the right way, that short circuit current would cause both the + and - terminals to heat up and burn.

Guys,
this is a typical failure when the terminals are not tightened. Drawing stall currents on high current motors cause tremendous heat and some sparking. The only way that over tightening could cause this failure is for a team member to have replaced the original screw with a much longer one. There would be evidence of the screw contacting the circuit board (scratches) if that was the case.

If the terminal is the type I think it is, it might be possible (hypothetically) to crack the pins that go through the circuit board by over-tightening the screw. The crack would reduce the cross sectional area of the metal leading to localized heating. Having said that, I have never seen one actually fail this way so the loose screw is still the most likely failure mode.

From what I can tell, the terminal is perfectly intact and we only used the standard screw. (It’s metric anyway, we don’t have those around) I agree that the most likely cause at this point was a loose connection on our part. This does bring up the question though: was the screw terminal loose to start with or did it loosen up over time?

As I have said for years, a loose connection can raise temperatures to extreme. I can see in the photo that the terminal unsoldered itself from the board for at least the terminals visible. The connection only needs to reach 400 degrees for the solder to release.

Was there the possibility of the wire being pulled tight so that it pulls on the lug making the screw turn counter-clockwise? It is most likely that it was never tightened enough. Many manufacturing facilities will have torque specs for all screws and will use calibrated electric screwdrivers to get it right. As a general rule, if you can push on the end of the crimp lug with your finger and cause it to pivot around the screw, even just a little, the screw is not tight enough.

Solder starts softening 5-25 degrees below the melting temperature, depending on the particular alloy. If there was tension on the wire, it could pull the terminal out below the melting point. I used the wrong solder last week and my test setup fell apart in the oven when it hit 175 C/345 F.

Speaking of Duluth, while I was a CSA during the competition I troubleshooted a team whose spark controller caused a spark too. They had the metal part of the crimps for the wires from the battery and the main breaker (that go into the PDP) rested on a spark controller, which caused a minor spark. I thought it was quite ironic too :stuck_out_tongue:

There was a decent amount of slack in them (enough that we just stuck the replacement on top of the bad one without any wire changes) but there is a very good chance it may have just not have been tightened enough.

I am also going to attach the driver station log file from the match I suspect the motor controller died in. Check PDP 2 and 3, the ports the motors on that side of the drivetrain were on. This probably won’t be much, but it’s something.

This is your most likely failure mode.

When I teach electrical assembly to various teams, I tell them to do a test immediately after a crimp is made or a screw is installed. I also tell them to find someone meticulous to go over their work. Afterward, I will check their work and I sometimes still find bad crimps or loose screws. I have also been asked by the head mentor on a past team to check the electrical system of teams that we picked. You can guess what I find…