Tuesday, June 25, 2013

Bottoms Up Bottom Balancing Again

Well, I needed to once again do a bottom balance on my cells. Not because of any issues while they were in the vehicle but because I mucked and messed with them and knocked them out of balance from each other since I last had them in my car. I had 24 100ah Hi-Power cells to balance. I did 26 of them. The main issue here is setting up a little balancing station for your own build and if you decide to bottom balance this would be a very simple setup. The main issue of course is how do you drain the cell with reasonable speed and not burn anything down. My first attempt was pretty crude but it worked and took out power at nearly 100 amps but was not automated. It required me to sit for long periods babysitting my cells while they drained. In the process I had decided I had enough time to run in and grab a quick cup of coffee (pour, cream and sugar) then come back out. Ooops. I was pulling power out so hard the cell had drained to near zero while I had stepped in for a quick cup of Joe. Bummer. Not going to do that again. This video was 2 years ago.

So now we move on to a bit better setup with a very simple meter. This is where the JLD404 and JLD5740 shine. Setup the JLD404 to count Ah out and the JLD5740 to monitor your volts and to stop and start your setup when you reach your required voltage. The relays in these babies are just perfect. You do need your 75mv shunt. I am using a 1000 amp shunt in my setup. Once connected and tested that all the systems work in harmony I begin my discharge testing. Ok lets back step a bit here. First you really need a load to connect to your battery. I use my spare motors. This time I only used my old Military Starter Generator. I could have connected up another couple motors but I did not want to drain them at too high a current draw. So only one motor was used and it provided about 33 amp draw. It was pretty consistent so I used that. My volt meter was set to shut the contactor off at 2.5 volts then when the voltage bounced to 2.8 or above the contactor was once again turned on so the motor would create another draw and then when the batter hit 2.5 volts again it would shut off. This would and could go until you no longer get the cell to bounce back over 2.8 volts. In effect bringing the cell to a balance at 2.799 volts. I actually did not allow this to go to long as I needed to bring down 26 cells. But if I left and forgot the setup I would be safe knowing that my cells would never get below 2.5 volts and never over 2.8 volts until I shut it off. I plowed through 26 cells. The result was that most bounced back up to 2.9 to 3.2 volts even after emptying them to 2.5 volts under 33 amp draw. If done at a higher amp draw the bounce would have been higher. So in order to take out most of the rest I setup another setup using my PowerLab 8. I have it set up to drain the cells at a constant 10 amps until the voltage comes down to  2.8 volts again. 2.8 volts being my balance point. The lab is set with a CC/CV algorithm and to cut off when the amperage draw reaches C/10 or about .5 amps or some where in that range. I figured it would give the amperage time to settle well so the cells should not go past 2.8 volts. Well most of them actually rose to about 2.85 volts and a couple reached into the very low 2.9 volts. I did not however have time to read all the voltages in the morning after the final trimming because the new owners of the cells arrived at 6:30 am to pick up the cells. But my quick check on a couple revealed 2.8 was the range but higher than 2.800. Most were at least 2.85... volts that I actually tested.

Here is the list of cells from the bulk charge, Ah out, Resting voltage, Trim Ah out, and time to do the final trim. The reason you see a few blanks is because I forgot to log a few cells so those are blank. I also decided later to add two extra cells from 24 to 26. So this is what you get. Voltages are lousy to use for trying to figure out SOC. Now the cells are in balance for the new owners. However the owner will most likely undo all my hard work because they are in the top balance camp. But they have my reason and they are thinking about it. They did not really give much thought and just followed the monkeys. I am totally in the bottom balance camp. Bar none. 





57 Deluxe 23 Window Double Door Bus

1 comment:

  1. Thanks for sharing your balancing setup. Everyone says that voltage is a poor indication of state of charge. There are two reasons I can think why: flat voltage in the middle of the discharge curve, and voltage bounce after a charge/discharge event. So if you are at the bottom of the curve (or the top of the curve), and you have a resting open circuit voltage, you can get a reasonable state of charge assessment from the voltage. Overnight may not be long enough to let the cells rest, for your final tweaks during balancing. I found, even with low charge currents, the cells were still settling by tenths of millivolts, 10 days after the last charge/discharge. The relaxation followed an exponential function with a 7 day time constant (63% of the voltage change happenens every 7 days). I guess it all depends on how close you want your cells to balance. With your setup, all you would have to do is leave the cell connected a little bit longer, and you would automatically get a more balanced pack, because those cells that tended to creep up would get more discharge cycles as the JLD5740 cycled the load a few extra times. But 5 mV of variation at the end of the curve is probably negligible when you are in the knee of the discharge curve.