Night Train wrote:Oi Tim! This is your thread! Where are you!
Thanks for your write-up, photos and videos, Night-Train.
That was a good weekend. On Sunday, after Jon and Night-Train had gone home, Ian and I connected the BMS to the pack.
We weren't sure which way the wires went into the Hall-effect current sensor, so left that out. The cell sensing wire ring tags supplied were too small to go over the posts, so ended up clamped between washer and strap.
My BMS system uses 6 slave modules to monitor up to 10 cells each. The slave looms were fitted, then the slave modules, then the main harness to the central computer.
Switched on the computer and ... low voltage alarm. Zero volts on every cell. What had gone wrong?
Check the voltage on a random cell with a multimeter. Ow! That was hot! No, terminal is cold. Try again. Ow! Pop! That would be the 10 Amp shunt going pop because I had the red meter lead in the wrong hole. Doh! My multimeter is now a voltmeter.
Back to the BMS mystery. Yes, each cell still has 3.3v across it. We haven't destroyed the pack by leaving it connected overnight. Phew! So why no indication on the BMS? Ian wondered if the BMS understood minus voltages, so we carefully disconnected the last 3 cells, then reconnected the last cell the opposite way round. Bingo! 3.33v
When we laid out the cells on Saturday morning it took 2 or 3 goes to get the strap pattern right to connect the end-of-rows without needing wires. In doing this we had ended up with cell 1 as negative and cell 51 as positive. Then when we put the BMS loom on we followed the cell numbering, so all the slave looms were the wrong way around. To correct this, we tried to think of a short-cut, but in the end there was nothing for it but to remove the slave looms and start again. I'm glad we weren't working in the car at this point.
Plugged in the computer only to see "low voltage alarm" again, this time with a pack voltage showing of 160v. Using a different screen on the BMS, I could see that all but 2 cells had 3.3v across them. Looking better
We traced the 2 low volt cells and found one badly crimped pin in a slave multiplug, and one cell on which I had got my wires crossed again. I'm really glad the failure mode for reverse connection is not dead short, or I would have had lots of melted wires that day. For installation in the car I intend every sense wire to have a fuse close to the cell terminal, but as this was a trial-run, off the car, only one fuse was fitted in the #51 sense line.
With the bad crimp temporarily pushed back into the plug, and the reversed wires corrected, we now had 168v showing on the BMS screen and no low-voltage warning. Success! It is a peculiarity of this system (but obvious when you think about it) that the pack voltage is calculated by summing the individual cells rather than a check from one end of the pack to the other. So if any of our intermediate fuses blow the BMS still shows full voltage without any continuity. In the trial set-up it would have been useful to take the sense wires the other side of the 5A fusewire, but in reality that will be a 200A fuse and if I blow that I will have other things to worry about than a bad BMS indication.
Many thanks again to Night-Train, Jon and Ian for giving up their weekend to progress this project. Next job will be trying out the balancing charger, checking what indication the BMS gives when the balancing spill loads are switched in. This will have to wait, as I am taking Mary away on holiday next week.