Wednesday, February 17, 2010

Well it has been a while since I posted. I now have about 2,800 miles on my Triac who I have named Mack.
Mack is performing very well. I am driving him daily, and usually never get below 45% charge. I am going to try and get some video done and put on YouTube as soon as the weather gets better. Somehow video always looks better when it is not pouring down rain.

I have seen a lot of debate going on in the EV community on battery management systems, and I would like to put my comment out there. The biggest issue I have seen is over charging or over discharging. Lithium batteries will be damaged if you charge them to high, or drain them to far. The chargers I have seen appear to do a good job at charging without going to high. The big issue I see is with Battery management systems not allowing the voltage to drop below a certain voltage, and they are doing a bad job. And by bad job I do not mean they are damaging the battery. What they are doing is stopping or limiting the amount of power because of false readings.
Here is my understanding on lithium batteries. I am going to use Mack (my Triac) as an example, but it applies to all Battery management systems I have seen.
My cut off voltage is 2.5V per cell. That means if one of the 45 cells goes to 2.59V it will shut off to protect the batteries from being damaged. The battery can go below 2.59V under load as long as it is still above 2.59V at rest. When you are going up a hill drawing about 150-200 amps the voltage in the batteries will drop. I have seen it drop as much as .3V. If my voltage is at 2.9V, going up a hill could fool my BMS system into thinking my batteries are to low, and shut the car off, when in fact I have almost half a charge left. I have not run into this yet because I usually only drain my pack to 50%, and I do not have any hills at the end of my trip.
There is my long winded two cents on BMS systems. I think they are great used as a visual aid, but I would much rather not have it making the decisions for me until it can do a better job. The big problem is the small range of voltage from full to empty. Mack is usually at 3.3V per cell when charged, and 2.5V when empty. That is only .8V difference.


  1. What battery chemistry does Triac use, and are you actually observing the voltage as you drive?

    If your batteries actually reach 2.9V at 50% discharge, then your batteries are very different from mine, and your BMS should be designed for your particular batteries.

    For LiFePO4 or LiFeYPO4 batteries, the at-rest voltage stays fairly constant until the battery is substantially drained. 2.9V at rest is very low, and at that level the BMS *should* err on the side of protecting the batteries. So, the BMS you describe would do a good job for that kind of battery.

  2. I am using LiFePo4 batteries. The batteries are not at 2.9V @ 50% SOC. The 2.9V was an example of what would happen if I was at that voltage, and needed to go up a hill. The point I was trying to make is that @ 2.9V my batteries are fine and not in danger of damage, even under load. The trick is how long can I be under load before the voltage @ rest drops to 2.5V.
    I always watch the voltage while I am driving to get a feel for how the batteries are doing. Not that that really tells me anything. The discharge curve is so flat it is not really a good gauge. It also tells me which battery is the highest, and lowest voltage, and at what voltage they are. That is good to know, because it will let you know if the Pac is getting out of balance. Mine has always been within .02-.03 volts of each other.
    My BMS does looks at the voltage under load, and treats it as if it was at rest. That is the problem. At rest, and under load are two very different things. I think a BMS system should be able to do the math of amount of load, and voltage drop. If I am drawing 100 Amps, the voltage will drop less than If I am drawing 200 Amps. I do not know if it is the same all the time, and could be accurate, but there should be some kind of adjustment made in the BMS to account for this.
    I am really anxious to see the results of the Nissan Leaf when it comes out, and how it deals with this. Everything I have read says that it gets 100 mile range, but it was done at an average speed of I think 19.6 miles an hour

    "And that 100-mile range is based on a driving cycle, devised by the Environmental Protection Agency, that consists of a pattern of starts and stops meant to replicate driving in Los Angeles, and that produces an average speed of 19.6 miles an hour. "

    Thanks for your comments :)