Bypassing BMS for High Current Application

[magudaman]

So I have a regular 70 amp BMS with balancing that I want to use for Low Voltage cutoff, High voltage cutoff, and charging but during my discharge I can see upward of 300 amps. My original intention was to run the BMS output + into my controller's key switch port but it turns out my BMS controls the negative only (most bms do, totally gapped it).

So I want to have some low power bypass that will turn the keyswitch on and off when the battery hits LVC. I will be charging directly through the BMS so the charge side is taken care of. I am guessing I can use a FET based switch but I am not sure how to wire that up, any help or ideas.

 

[olaf-lampe]

I had the same problem with electrifying a scooter with a ping-BMS. I used solenoids that where switched by the BMS power output but the power-lines were wired directly to the battery.
-Olaf

 

[amberwolf]

Since the BMS switches the negative side, you can do either of these:

Wire directly from the pack's + and - to the controller.
Setup a small NPN transistor that can handle the full pack voltage (Vce => pack max charged voltage) with base wired to a voltage divider from pack + to pack -. Base itself goes to BMS - output thru a smaller resistor. Collector goes to keyswitch line, with a pullup resistor from collector to pack +. Emitter goes to pack -.

When the BMS is running normally, it grounds the base, so the transistor is turned off, and teh keyswitch is pulled high by the pullup resistor.

When the BMS triggers LVC, it disconnects the base from ground, and the voltage divider inputs the proper voltage (current) to the base to turn the transistor on, and it then grounds the keyswitch wire on the controller, turning it off.

Make sense?

 

[magudaman]

Yah I do sort of understand and that does make sense. My only concern is the pull up resistor and output to the controller. I believe that key switch line powers the controller electronics so there is a little bit of load. So so I just calculate accordingly and what happens when I short to ground, doesn't it suck up some power then too.

Couldn't I just use the positive of battery and negative of bms to switch positive of battery on and off through a transistor. I just am not sure I understand transistors well enough to understand how I calculate resistors for their required voltages.

 

[olaf-lampe]

I just am not sure I understand transistors well enough to understand how I calculate resistors for their required voltages.

Thats why solenoids like the tyco czonka 3 are much easier to use. A bit overkill, but pretty common:
http://relays.tycoelectronics.com/datasheets/ev200.pdf

-Olaf

 

[amberwolf]

I believe that key switch line powers the controller electronics so there is a little bit of load.

There is, but not much, most likely (it could be measured with any multimeter set to amps, though--just stick the meter leads between the keyswitch line and the controller power input).

So so I just calculate accordingly and what happens when I short to ground, doesn't it suck up some power then too.

Yeah, it would be using up power then, depending on your pack voltage vs the dropping resistor. That could be avoided by using a second transistor that is switched opposite the first, so it is only connecting the keyswitch to the first transistor when the first one is off. Might need more than one extra transistor, though; I haven't pondered and visualized that one yet.

Couldn't I just use the positive of battery and negative of bms to switch positive of battery on and off through a transistor.

You'd want to use a contactor, like the EV200 mentioned above, to do that, rather than a transistor. Transistor (or FET) would have to dissipate a lot of power thru it, dependng on it's internal resistance, so either multiples in parallel or some bigger heatsinked ones might be needed, depending on your pack voltage and on the peak currents your setup requires.

I just figured that doing it via the single transistor was the simplest/lowest-part-count method.

To do it via contactor, you need the transistor circuit (or a variation of it) plus the contactor, and the transistor stuff must be setup to only deliver the right amount of current to trigger the coil, and not over-drive it. (full pack voltage would probably destroy most typical relay coils, unless they are rated for that high a voltage--most common ones are 5 or 12 or 24V coils, AFAIK).


The calculations can be done with web-based stuff; I somewhere have a shortcut to a page that does basic transistor circuits by just plugging in the numbers you do have, and it figures out what resistors you need.