Major Problem - Stability

12/24/2015

Click a picture to enlarge it



Our second tournament showed major instability issues. The team was fighting them even before the match.

The debug process was very stressful… We went through a process of testing every wire, taking everything off the robot,
and even hard resetting the phones and swapping them. Eventually we determined there was something wrong
with our PDM(Power Distribution Module) and with our wire connectivity. We fastened all the wires in, and secured
our PDM’s connection to the battery and voila, the robot is now more stable than it has ever been!

- Stability got worse during the match
- Things worked slightly better in the days following the match
- Then we fell of the cliff... on 12/21/2015, robot completely unstable, won't work for more than 3 seconds
- 3 solid days, after much problem investigation and isolation, several steps back, then several forward, 2 trips to DDC, 3D printing, and targeted repairing... problem fixed
- Robot more stable than ever. Doing weelies. No failures.
- Not only back in business, better than ever


Some of below is part of the saga

Lessons learned
- stability is a bear
- stability is critical
- problems, even difficult ones, can always be solved
- we can solve whatever we put our minds to



Robot became completely unstable. This was not fun for anyone. But bound to happen in such a complex machine. We were plagued by this at the last tournament as well. It somehow got worse.
Suspecting the phone and associated connections to it, we decided to secure it. We decided to 3D print a chassis/holder for the phone, to secure it. The other one we had been using was flimsy, and also broke. Our new one will be much better.
We begin 3D printing
Took 5 hours in total. There were 2 components, a base chassis, and a cover.
Worked perfectly.
Problem solved, or so we thought... after a few bouts of thinking we solved the problem (it worked o.k. for a while), we reverted back into chaos... NOT working well at all.
A lot of this going on while the team pondered/freaked-out/sweat
Next day... start over, disassemble everything. Laid out all of the electronics of the robot off of the robot, and tested individual components and ran 2 motors, the PDM, the battery, a motor controller, and the simplest wiring to see if the problem was contained there. Realized... Connection between PDM and battery caused failures, and the LEDs on it to glitch... unstable
Also realized this wasn't getting done
So we cut the intermediate connections between the PDM (Power Distribution Module) and the battery, as there was a power take-off there, as well as a heat-shrunk connection. Not certain what the issue was, we also suspected, it could be a bad connector. We decided to start over with this connection, and do a direct connect to the battery.
These connectors were suspect. We requested more be brought home from DDC. Unfortunately, received wrong type.
We then direct connected the battery. While we basked in the wisdom of this debug method, confident our process will take us forward... This led us astray since the battery itself turned out to have problems. We determined this by checking the voltage (which looked fine at first), and then jiggling the wires around (and it went to 1 volt... arghhhh!!!!). So that led us off the scent for a couple of hours. We labeled the battery "sporadic"
We finally determined that 3 things solved our overall isue: 1.) phone holder, which has a built in shim for the cord not to move 2.) The PDM to battery connection was flaky when the cords were torqued, replacing them with a single hard-wire connection 3.) the USB connection to the motor controlllers (the final culprit that could be discovered when the other issues were corrected
For now, this is the method (electrical tape the usb wire to the chassis, near the connection) to secure the motor controller wire. It is working well. Robot can take a lot fo jarring and abuse (as in wheelies) with no hiccups
This is a shim we will likely 3D print and make use of as a final solution.for securing wires
A fine job sweating thru this and solving it. Everybody happy now...


Created on 12/24/2015 11:52:13 AM
with PhotoPageGen version 3.7