Summary: This thread is about integrating an inexpensive BMS with a Solectria Force. The Force has an AC motor. A work in progress.
Now, I want to relate my exploration of Battery Management Systems with lithium packs.
A fact of life with lithium battery packs is that the battery packs need to be managed. By that, I mean looking after the condition of the batteries. The battery pack in the Tesla is as good as they get. It's battery pack has 6800 cells, which are managed. It is a massive parallel - serial combination of cells. Series to get the voltage, and parallel to get the amp-hours. Here is a paper on some of the Tesla design considerations.
In the cost analysis above, I compare Thundersky batteries to Headway batteries. From a cost perspective, Thundersky is the winner. But from a raw amp perspective, the Headway is a winner. AC drive motors draw less amps than DC motors, and can rev higher RPM's. If you are building an EV with a DC motor, then the Headway may be the logical choice for a battery pack.
Some characteristics of Battery Managemant Systems: Lithium cells may overheat if overcharged, or over discharged. So the BMS senses the parallel cell voltages and either manages charging, or allerts the driver of undervoltage / hazardous condition. For the BMS under consideration, the electrical function is described here. On the cell low voltage end, when discharging, the BMS sends out a signal when any cell reaches the Low Voltage Cutoff. I assume that the driver should re-charge the battteries at this point. My Force has an accumulated amp-hour gauge that tells the battery pack state of charge. So you see it coming. On the cell high voltage end, when charging, the BMS bleeds power off of parallel sets of cells, when they are charged beyond the High Voltage Cutoff. A LED switches on, and the charger current is cut back / regulated. It is assumed that most of the cells reach this point in the same interval of time. Testing will prove this out.
The concern I have is the following: The working voltage range of my Force is maybe 143 to 180. The AC controller does not like to see higher voltages, as it may blow a capacitor or some other power component. On the low voltage side, the controller goes into a creep home mode and limits the controller amps and the vehicle speed. I look forward to communicating with the controller wiith a program called AMCmon.zip available here. Hopefully it will tell me more about the controller parameters. So, the voltage range is from 143 volts to 180 volts ( 143 v+ 25%). It is stated in the Thundersky literature that the working voltage is 2.5 to 4.25 volts (2.5 v+ 70%). This is a problem, because the range is somewhat more than the 25% for lead. I can either not charge the batteries to 4.25, or not discharge them to 2.5. Hence, I am not using the full amp hours in the pack. And the advantage of going to lithium.
Upon further examination of the discharge curve , at the bottom of the page, the voltage excursion, where the majority of the watt-hours is located, ranges from 3.0 to maybe 3.3. This translates to plus 10%. So maybe I am OK. Even if it were from 3.0 to 3.5 the excursion is 16%. The additional area, at full charge is not much. So my plan is to set Low Voltage Cutoff at 3.0, and the High Voltage Cutoff at 3.5. I think these values can be set through some precision resistors.
BB has a thundersky battery pack on a Solectris Force - the performance is as expected with no BMS. Longterm results pending.
Aside from a BMS, having a display that shows battery voltages, in bar graph form, would be of interest. maybe a few temperature readings also. Like this display. The drawback is that the optos are not quite linear.
More to come.