2
\$\begingroup\$

My company is designing a 4 cell BMS. We recently did a redesign of our cell voltage measurement circuits and are having major issues.

The first stage of each amplifier is a differential opamp with unity gain. For cell 1 we take the difference between gnd and CV1, for cell 2 we take the difference between CV1 and CV2, so on and so forth.

For cells one and two the differential amplifiers are working perfectly. However, for cells 3 and 4 the amplifiers are outputting voltages much high than the actual cell voltage. The output from the cell 3 differential op-amp is around 7 volts and for cell 4 around 11 volts.

The second stage of the amplifiers is meant to lower the voltage to an acceptable range for the ADC. Unfortunately, we used a non-inverting op-amp and have a gain greater than one. A gain of around 0.7 is desirable.

Thanks in advance for any help!

Schematic

\$\endgroup\$
10
  • \$\begingroup\$ If you need to debug anything , simply record the voltage on each pin and compare with datasheet for I/O range , gain and offset, then verify R values. " 0.7 volts " is not a gain spec. Be more careful on design specs, & measurements \$\endgroup\$
    – Tony Stewart EE75
    Feb 15, 2022 at 1:14
  • \$\begingroup\$ Hey Tony, thanks for the comment. 0.7 volts was a typo. We're looking for a gain of 0.7. We've scoured the datasheet and it seems that we're well within range with our voltages. The highest input voltage that it will ever see is 16.8 which is why we chose the 20V supply voltage. \$\endgroup\$
    – Ryan Coleman
    Feb 15, 2022 at 1:21
  • \$\begingroup\$ If you had chosen 1st stage to be 30k:20k you would have had gain = 2/3 with no need for 2nd stage. For Av=0.7 Rin would be 20k / 0,7= 28.57k . or choose better values with a better design criteria or "design spec" \$\endgroup\$
    – Tony Stewart EE75
    Feb 15, 2022 at 1:40
  • 2
    \$\begingroup\$ This is a perfect example why people should buy ready-made BMS, or at least use ready-made BMS ICs, rather than design their own from scratch. For one thing, this circuit draws so much current that it will over-discharge the cells and damage them irreparably in a relatively short time. For another thing, it's so inaccurate that it will balance the wrong cell based on an incorrect voltage measurement. For an IC that has already solved the problems you haven't yet even thought about, look at the great offerings from Texas Instruments. \$\endgroup\$
    – Davide Andrea
    Feb 15, 2022 at 1:51
  • 1
    \$\begingroup\$ Cell 3 and 4 is a construction error and I pointed out how to simply diagnose. List all pin voltages and compare with spec. which I read now is Vdd= 5.5V max not 20 for LMV324 .. oh oh. I also concur with @DavideAndrea \$\endgroup\$
    – Tony Stewart EE75
    Feb 15, 2022 at 7:02

1 Answer 1

Reset to default
3
\$\begingroup\$

(1) Your schematics says "LMV324" for channel 3 and 4 -- but LMV is a low voltage part, rated up to 5.5V.

(2) You want to be aware of "common mode voltage" value, which is average voltage on both inputs. for LMV324, it is 1.7V. For LM324, it is V_cc - 2 volt (which you will be violating in channel 4 if you use it for measurement).

I recommend that you use a pre-made BMS. The BMS'es have an important protective function, and installing the wrong part may cause battery fire which is very nasty and hard to extinguish. It is much safer to use pre-made one.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.