Power Distibution Board - Broken stud

[de_]

Our 2013 power distribution boards stud broke on the first usage BEFORE the nut was even tight (power cable still loose on stud). It was installed by a multi-year seasoned electrical guy. Yes it was the correct original metric nut. I presume there is no warranty. Thats 3 power distributions boards we have with damaged or broken studs. Thats $570 of not usable in the competition boards and I suspect a minimum of one out of every two boards we own.

If we are in that situation, chances are a lot of other teams are. So there could a thousand or more of these boards in this condition in the FIRST community. I do know the apparent reason the boards use a brass metric stud / pcb mount is there is not US equivalent. And its brass so it can be soldered to the PCB and handle the high currents. But with an ongoing 50% failure/damage rate we are seeing, its time for FIRST to work harder for a solution that will have a reasonable MTBF in the FIRST environment.

We have in past cut the studs off, drilled and tapped the brass block for a 1/4" steel bolt which seems work fine and handle the 120amp max current without any issues. Too bad FIRST will not allow its usage in the competitions.

Our 2013 Digital Side car was DOA this year so between the two boards thats over $350 after taxes and shipping we have to spend to replace the boards which both had a MTBF of 0 minutes apparently with no warranty.

[Ether]

Originally Posted by de_

Our 2013 power distribution boards stud broke on the first usage BEFORE the nut was even tight (power cable still loose on stud). It was installed by a multi-year seasoned electrical guy. Yes it was the correct original metric nut.

Not meant to be argumentative, just wondering: was he using a torque wrench when this happened?

[Jon Stratis]

The PDB was introduced in 2009 for Lunacy, if I remember correctly. Since then, we've only had to buy one replacement a couple of years ago, for a board that got burned out after being mis-wired. We've never had any of the studs on our boards break off.

Now, for our story this year on a broken PDB:

This year, we're using an old board. About 4PM the day we bagged the robot one of our mechanical guys wanted to run a motor manually - we have a rig set up that allows us to safely (this part can't be emphasized enough for teams that want to do something similar!) do this with our Anderson PowerPole connectors, and it even has variable speed control! Anyways, when the mentor went to power the motor, he plugged into the wrong side of the PowerPole connectors, feeding 12V into the PDB through the Wago connectors while the robot was turned off. When they powered on the robot, there were a lot of red LEDs on the PDB and stuff didn't work.

While all this was happening, I was in the car heading to the build space from work. I got a phone call from the guilty mentor that started with "I think I did something very bad..." After getting the full description of what he was seeing, I told him the board needed to be replaced with one of our old "spares". Seemingly at the same time, the electrical team came to the same conclusion after doing some searching on CD.

So, take this lesson to heart: mechanical mentors should never touch anything electrical!


Anyways, what I'm getting at is you're seeing a 50% failure rate, but most teams aren't. My team hasn't seen any failures that we didn't directly cause through misuse. If there really were that many failures of the studs across all FIRST teams, we would have heard a lot more about it than the occasional breakage we hear about now.

[FrankJ]

I talked to a representative from Total Electronics at worlds last year. He seemed to be a stand up guy. If the issue really is a defective part & not operator error. (I do not judge these things over the internet ) I expect they will stand behind it.

I have found AndyMark as well to stand behind their products.

[Alan Anderson]

Originally Posted by de_

Thats 3 power distributions boards we have with damaged or broken studs. Thats $570 of not usable in the competition boards and I suspect a minimum of one out of every two boards we own.

If we are in that situation, chances are a lot of other teams are.

The other possibility is that your team simply doesn't understand how to tighten the nuts properly. I have first-hand experience with only two broken PDB studs out of literally hundreds of boards, one caused by an unsupervised and untutored student cranking down the nut too hard, and the other caused by a cross-threaded nut being overtorqued when it was obviously stuck. I have heard of fewer than a dozen in the history of this control system, including your three.

My sample and yours have greatly different qualities. Is there any way we can get a reliable survey of PDB breakage across all of FRC?

[EricVanWyk]

Originally Posted by Alan Anderson

Is there any way we can get a reliable survey of PDB breakage across all of FRC?

The single hardest part of my job is that I have no way of knowing about errors that go unreported. The worst feeling is finding out that there was a fixable issue that has gone unresolved because of lack of information.


That said, this really does sound like user error to me.

[IndySam]

We have every PD board we have ever received and they are all still completely functional, including a prototype board. In all the time FIRST has used these boards I have never seen a broken stud that wasn't from OE.

[de_]

If you only install 1 board a year and a mentor does it as opposed to a random student, then sure you are less likely to have a problem.

But in our team, random students installed at least 5 boards on various desktop and test & competition robots this year alone (many when I am not there).

What is the probability a student will realize the stub is brass and its tensile strength is 19K PSI versus the grade 8 fasteners we use which is 150K psi. Zero. We don't teach metallurgy in our high school. We don't own a torque wrench. What student could possibly recognize the metric nuts Zero - looks just like a 1/4 nut. Standard 1/4" 20 nut fits on the stud and tries quite happily to cut a thread into the soft brass. That is how a student trashed our first PDE.

This is a $190 hidden mine waiting to explode, especially in a dynamic high student to mentor ratio environment. Count your self lucky if you have not had it happen yet.

The particular student in this case has install a lot of PDEs over the years. He's installed a lot of brass pneumatic fittings in aluminum with no issues. He has no history of brute force. He used a small wrench. So don't blame the student.

Our robot is probably 50% over budget this year and now we should be buying a new PDE and a new Digital side cars as both did not survive even a minute in our student environment. Is it the students fault or is it a design fault? I can't find fault with the students.

What I will do between trying to get a robot built and buying replacement parts is try to find the time to make a warning label for future PDEs that warns the studs are brass and metric and will break 7.5 times easier than a grade 8 bolt. Something FIRST might consider. I will also send a link of this post to FIRST.

[Jon Stratis]

Originally Posted by de_

If you only install 1 board a year and a mentor does it as opposed to a random student, then sure you are less likely to have a problem.

But in our team, random students installed at least 5 boards on various desktop and test & competition robots this year alone (many when I am not there).

What is the probability a student will realize the stub is brass and its tensile strength is 19K PSI versus the grade 8 fasteners we use which is 150K psi. Zero. We don't teach metallurgy in our high school. We don't own a torque wrench. What student could possibly recognize the metric nuts Zero - looks just like a 1/4 nut. Standard 1/4" 20 nut fits on the stud and tries quite happily to cut a thread into the soft brass. That is how a student trashed our first PDE.

This is a $190 hidden mine waiting to explode, especially in a dynamic high student to mentor ratio environment. Count your self lucky if you have not had it happen yet.

The particular student in this case has install a lot of PDEs over the years. He's installed a lot of brass pneumatic fittings in aluminum with no issues. He has no history of brute force. He used a small wrench. So don't blame the student.

Our robot is probably 50% over budget this year and now we should be buying a new PDE and a new Digital side cars as both did not survive even a minute in our student environment. Is it the students fault or is it a design fault? I can't find fault with the students.

What I will do between trying to get a robot built and buying replacement parts is try to find the time to make a warning label for future PDEs that warns the studs are brass and metric and will break 7.5 times easier than a grade 8 bolt. Something FIRST might consider. I will also send a link of this post to FIRST.

It really comes down to student training. You don't have to go into too much detail or talk about metallurgy. We have a small bag in our electrical kit labeled "PDB Nuts" with spares of the correct nuts in it. The students on the electrical team know its there, and we make it a talking point every year that it takes different nuts than the rest of the robot. When we go over all of the electrical components, we stress that they need to make things tight, but they can't crank down on them like they would on frame members - plastic will crack, and even if it doesn't it will compress and damage circuitry. Our students do all the install work for us, and we have never had a problem. We typically install 2-4 of them each year, as part of practice robots, demo's, bench-top test boards, etc.

[de_]

Its not student training: The last one broke with a well trained student. We will not get a better student.

[Ether]

http://www.chiefdelphi.com/forums/sh...90#post1240690

Are these two problems related?

[de_]

Kevin, Where are you getting your data re other teams experiences re the stud ? Have you surveyed them ?

I have already stated I know exactly why how the current PDE ended up as is. So don't go there.

I can tell you 0 studs failed on the previous design (that was a $30 part). They were not brass.

[Kevin Sevcik]

Originally Posted by de_

Kevin, Where are you getting your data re other teams experiences re the stud ? Have you surveyed them ?

I have already stated I know exactly why how the current PDE ended up as is. So don't go there.

I can tell you 0 studs failed on the previous design (that was a $30 part). They were not brass.

Nothing scientific, just subjective impression based off the teams I work with here in Texas, and the lack of outcry on CD. If it was a very common problem, you'd see a lot more about it here. Look at the kerfuffle on the BAG Motors. By Paul Copioli's count, it's something that's affecting a very small percentage of motors, but there's several teams on here talking about it.

[FrankJ]

Here is a data point: We have six of these boards on various robots. We have one a broken stud. Relatively untrained student screwed on a 1/4-20 nut with an experienced mentor looking over her shoulder. Definitely operator error in our case.

We are now powering that PD board through a Wago connector for a test bed.

[FrankJ]

So where did it fail? Is it at the button of the stud where it transitions to the square? Does the nut turn freely on the stud? Was the stud bent or the threads damaged?

You are never going to get a grade 8 fastener in this application. You can teach the difference without going into metallurgy.

Occasionally even the best of us make mistakes. I would like to say it wasn't so.