1983 Porsche 911 SC Targa

Sunday, September 30, 2012

The Batteries Arrive

After a little excitement with UPS losing my batteries for a day – they were loaded onto the wrong delivery truck – the batteries are here. I’m really glad they got here in one piece. With all of the cells laid out beside the car, I am wondering if they are really going to all fit.

The battery is considered fully charged at 3.6 volts, and empty at 2.8 volts. These cells shipped at 60% full, at 3.3 volts. There is a little variation in voltage from cell to cell.
Before the cells can be installed into the car, they need to be balanced. The process is to wire up all of the cells in parallel, and charge up the pack. The goal is to ensure that each cell has the same voltage when the pack is fully charged. When this occurs the pack is said to be in balance. Balancing is required because each cell has a slightly different capacity, and as the pack is discharged, cells with less capacity will have lower voltage than cells with higher capacity. If the cells are balanced when full, the charger will be able to more reliably stop the charge process when the pack is full. There is less risk of a single cell reaching the full state ahead of the other cells and being over charged. The problem with top balanced packs is that on discharge it is difficult to know when cells are empty. The voltages will vary a lot, and there is risk of discharging a cell too low. For long battery life it is important to be conservative with charging and discharging the pack.  I will also have a circuit on each cell (battery management system) that monitors for a low voltage condition and gives a warning to stop driving before damage occurrs.


  1. 60 Cells in the Porsche may be a bit ambitious, but give I say give it a go! You may find some very creative ways to get them all in there.

    You know where I come down on top vs bottom balancing, but I'm glad to read you understand the dangers of top balancing. When it comes down to it, top balancing or bottom balancing both have their advantages and disadvantages. You just need to know what they are and act accordingly.

    The CA cells seem to exhibit only a 7% sag in voltage under high loads. Of course that was from a nearly fully charged cell. I don't know if state of charge affects that sag, but that's something to watch out for. If your battery management system alerts you that you have a low cell, that alert is likely to come when the cell is under load, giving you plenty of time to get the car off the road and prevent any damage (depending of course on what you set the alarm threshold for).

    In terms of charging, I'd encourage you to stay away from the 3.6 volts CALB says is the top end. If you were to watch the charging cycle with a good amp/hour counter, you'd see that only one or 2 amp hours actually goes into the batteries between 3.45 and 3.6. Enough energy for a lap around the block. With nothing good happening anywhere above 3.6 volts and virtually no difference in the energy stored between 3.45 and 3.6, I'm just not convinced it's worth pushing them that them to the full 3.6V. But that's my 2 cents.

    Looks like the build is coming along nicely. Thanks for keeping us all up to date Joey.

  2. Hi Tim,
    Good to hear from you again.

    You mentioned a lot of important battery issues. CALB recommends to charge the cells at constant current, and then constant voltage, and to terminate the charge when the voltage is 3.6 volts and the current falls to 0.05C (C is cell capacity), or in my case 180 Ah *0.05 = 9 amps. When the charger is switched off, the voltage will relax over a day or two, and should settle at a resting voltage of 3.4 volts. This is considered fully charged. I agree that pushing the cells to full charge is not best practice for long cell life. I plan to charge to 3.45 volts per cell and terminate when the current drops to 9 amps. The resting voltage should be safely below 3.4 volts. If you tried to charge at 3.4 volts, there would be little voltage difference between the cell and the charger, not much current would flow, and it would take forever to get to 3.4 volts.

    I will have a battery management system (BMS). Details are not finalized yet, but shunting would kick in at 3.5 volts and the BMS will stop the charger if a cell reaches 3.6 volts. With these parameters, the expected routine during charge is that the BMS will not do anything, so long as the cells stay in balance. The charger should handle the charge, and the BMS is only a backup. I know many people hate active top balancing systems, and so far the only argument against that makes sense to me is that you can’t accurately balance a cell under load (during the charge) because the resting voltage will be much different than the loaded voltage. But my planned system will not actively try to balance with every charge, only if a cell is out of balance. I may yet clip out the shunts and use the BMS only to end the charge early is a single cell exceeds 3.6 volts.

    With a top balanced pack I will be dependent on watching the amp-hour meter and the BMS to provide a warning to know if any single cell is getting low. I read your blog about how your pack is bottom balanced and I was impressed that your testing showed that when all of the cells reach the bottom together, the power drops dramatically and there is very little risk of over-discharging you pack. If only a single cell is empty, the driver probably wouldn’t notice until it was too late. It almost made me change my mind about bottom balancing. But this is why I stayed with top balancing – I will not be present when the charger is running. So as the pack ages, and cell capacities decrease, you can’t be sure that they all decrease at the same rate. A bottom balances pack has more voltage variation when the cells are charged, and over time I would be depending more and more on the BSM to protect the pack every time I charge the pack. A conservative charge cut-off is recommended for a bottom balanced pack, but the margin erodes over time. With a top balanced pack, the voltage varies most when the cells are empty, and I feel I have more control over how the pack is discharged. I don’t think there is an issue either way; you just need to know the pitfalls and nuances of each method.
    I’ll post soon on the balancing process that is now underway.

  3. Joey, I think you hit the nail on the head on many points. Top balancing the cells while under load is folly. The voltages will inevitably be different after the cells rest. I've seen cells charge to the same apparent voltage, but then after resting for a day show voltage disparities. It's entirely likely that the act of trying to balance them under load actually puts them out of balance. But like you pointed out that doesn't negate the merits of top balancing.

    But what does that ultimately mean? It only means you have a pack that's out of balance at the top (by a very small margin) and at the bottom (by a larger margin). As opposed to a bottom balanced pack which is as evenly matched at the bottom as the person doing balancing has patience to do, and wildly unbalanced at the top. In practical terms having a top balanced pack means very little unless you discharge the cells enough you get near the bottom.

    Like you'd pointed out, all the cells have slightly different capacities. So if you bottom balance, and then charge until the first cell hits 3.45V, after a rest, cell voltage in the pack is all over the place, It's ugly, but harmless. Top balancing moves that unevenness to the bottom of the charge curve. But if you have a BMS that's watching each cell and reporting a low voltage condition, you've mitigated the risk.

    With that system in place, there is one argument for top balancing that you can't deny. You're going to charge the pack to full far more often than you discharge the pack to empty. Since you're going to spend more time at the top, there's an argument to be made for keeping them top balanced.

    But one thing I've noticed, and I think you'll notice (before you do any active balancing that is) is that these batteries are rather like a rocking chair. You put energy in and it comes back out: in - out - in - out - in - out, and like a rocking chair, there's a lot of movement, but it doesn't go anywhere. Meaning cycle after cycle, they really just don't drift from each other. I have one cell in my pack that I know to have the lowest capacity. For the last 1.5 years, I can count on it hitting 3.45 volts just before any other cell in the pack does. Like clock work. No cell has ever drifted or moved in state of charge relative to the others and hit that magic number first. And I've watched for just that and it doesn't happen.

    Like I'd said and you echoed, both methods have their advantages and disadvantages. The trick is watching for and avoiding the pitfalls no matter which you choose.


  4. my 2 cents.... do bottom balancing and do not overcharge (3.45v )considering all your cells are good and very similar in capacity. If you do top balancing, when the first cell hit SOC 20% that's when you range ends. otherwise that's when you start having cell going bad, and it happens faster than you think. My friend has lost 20 cells out of 80 already. He is using top balance. Also, if you live in areas where you have temperatures below 50F, you will need a battery heater. Your gray cells are have much better cold weather performance.

  5. I'm using the Orion to top-balance. When cells hit the charge limit, the Orion begins passively discharging them until they reach the same voltage as the lowest cell. What's great is that it's CAN-enabled, and so all pack data is available to me while driving in real-time on an Android device via an OBDII to Bluetooth interface using the Torque app. I can see high cell, low cell, resistances, pack voltage, amphours, state-of-charge, discharge current limit, remaining "safe" amps, etc. Torque can also be set up to sound alarms if any limits are hit on any of the parameters.


    1. I'll look into the Orion. In general I prefer the BMS to be more of a monitor and less active in the management of the cell. My main concern is system reliability.