Our bikes are used quite hard at times, and the electrical system is mostly homemade, so we can expect a few problems occasionally. This week I have not done well at all!
It started on Monday morning, I had just descended Whitford Brown Drive (100 metres of descent on a good wide road with a bike lane) and was turning right into Papkowhai Rd, gave the throttle a twist to keep up with the other turning traffic (and because I keep it in a high gear ready for the slight downhill across the junction). Nothing happened – no power, stuck in a high gear traffic getting away from me! I dropped it down a couple of gears, and limped it a couple of kilometres to the shop. I checked all the connections, everything looked good with a cursory glance, and I thought the worst, either the motor had packed up or the controller was playing up. When Lynn arrived in the afternoon, I started the process of swapping the controller from her bike to mine to test just what was wrong. Starting with my bike, I started taking all the connectors apart, throttle, pedelec, brakes, power, motor hall sensors, motor power cables. Oops, motor power cables would not come apart! Had a good look, turned them over, and there was obvious swelling and melting of the connectors. Since I had the supplied bullet connectors melt on me within a few hundred kilometres, I had switched both bikes to colour coded 45amp Anderson Connectors. They had been good so far, but I must have made a mess of crimping one of the connectors, causing a high resistance connection, which had finally moved to the point where it was arcing and causing large amounts of heat.
I cut the connectors apart, replaced what needed to be replaced, put it back together again, and off we went, with a nice power level again.
Tuesday night, I set off from the shop and because of an idiot in a tin box I got trapped into having to ride over a vicious man hole cover, the bike, battery and controller all bounced quite hard, and the motor shut down. As it was showing exactly the same symptoms, I had a look at the Anderson connectors, and one of the other connectors had popped apart! The wire had pulled out of the crimped connection. A two minute ride back to the shop, 10 minutes with the terminal crimper and a new connection, and it was all as good as new.
Finally, I hope, last night (still Tuesday) I decided to improve the pannier retention system to stop the bouncing or at least reduce it. This involved taking the batteries out of the panniers. There are 6 Headway 40152 cells on each side, nicely packaged with wiring looms that come together at the BMS which is in one side, along with the the charging port and the power socket. So I put it back together again, re-installed everything and switched it on, and it all went quiet. The voltage was showing 19.6 instead of 39+. This was similar to a problem I had been getting a couple of weeks ago, when the battery was mysteriously cutting out at the top of the big hill on the ride home. Consistently, for about a week, it would cut out at the same point every night. After a couple of minutes, the voltage would rise again, and everything was good for another day. It went away when I did some work on the panniers. So I tested everything again, and the 19.6 v was at an amazingly low ampage. This made me think that the BMS was actually switching the battery off. I thought this was the case when it happened on the road, but as I could not test it as thoroughly, it was difficult to confirm. I would have thought that when the BMS switched off the battery, the three mofset transistors would have got the voltage a bit closer to zero. But the BMS isn’t the best piece of equipment I have bought…
For the BMS to switch off the battery, one or more cells must be getting down below the low voltage threshold, and at the time I suspected one of the cells of dropping voltage while under load up the hill. This is why I am building the the multicell voltage meter/logger.
Anyway, I tested all the connections and voltages at all available points, and was getting the right voltages out of each cell, and getting unbroken circuits from the connectors to the BMS. I can if I am feeling brave test the voltages getting to the BMS, but as it is a very tightly made PCB, I always worry I am going to short something once too often and blow up another BMS. I re-installed the panniers and connected everything up securely, making sure the multipin connectors all clicked, and everything was fine again. I think one of the 13 BMS connections was not quite connected properly, maybe reducing the voltage slightly or actually disconnecting it.
I think I maybe need to rethink the battery build, keeping wire runs from the cells to the BMS as short and as simple as possible. Having 6 cells on each side of the bike has its advantages, but reliability may be more important.