Low-side battery charge current measurement.

Thread Starter

mspider65

Joined Sep 20, 2021
7
I'm going to design a system to start/stop and control from remote the charge process for my eBike battery.
I am stuck on the current sense circuit as, due to chip shortage, I have not found a suitable current sense amplifier with a high common mode voltage from my PCB supplier.
I changed my design in order to go with low-side measurement and ended designing this circuit.
ChargerContorl.jpg
The sensing resistor is 4 mOhm and since the max current is 5 Amps the max voltage across the shunt will be 20mV.
I used a precision opAmp with a low input offset voltage (typ. 2uV). The resistor values give a gain of around 110 and add an offset of 0.2V to the output since my ADC is linear between 0.15 and 2.45V and any voltage below 0.15V is sampled as 0.
The charging process is controlled with a pair of Mosfets (to avoid reverse current through the Mosfet body diode).
The circuit is not isolated and the ground is common also with the MCU (an ESP32 board).
I'm not a very big seasoned expert and I'm not entirely sure if this circuit will fit and what the drawbacks might be when measuring the current from the low side.
Any comment or suggestion is really appreciated.
Thanks
 

KeithWalker

Joined Jul 10, 2017
2,295
It looks as though it will work well as long as the charger ground does not accidentally connect to the bike and battery ground. It would not be dangerous. It just would not sense any current.
 

Thread Starter

mspider65

Joined Sep 20, 2021
7
It looks as though it will work well as long as the charger ground does not accidentally connect to the bike and battery ground. It would not be dangerous. It just would not sense any current.
Thanks for the feedback, then I will continue with this solution.
 

Thread Starter

mspider65

Joined Sep 20, 2021
7
I'm going to design a system to start/stop and control from remote the charge process for my eBike battery.
I am stuck on the current sense circuit as, due to chip shortage, I have not found a suitable current sense amplifier with a high common mode voltage from my PCB supplier.
I changed my design in order to go with low-side measurement and ended designing this circuit.
View attachment 255267
The sensing resistor is 4 mOhm and since the max current is 5 Amps the max voltage across the shunt will be 20mV.
I used a precision opAmp with a low input offset voltage (typ. 2uV). The resistor values give a gain of around 110 and add an offset of 0.2V to the output since my ADC is linear between 0.15 and 2.45V and any voltage below 0.15V is sampled as 0.
The charging process is controlled with a pair of Mosfets (to avoid reverse current through the Mosfet body diode).
The circuit is not isolated and the ground is common also with the MCU (an ESP32 board).
I'm not a very big seasoned expert and I'm not entirely sure if this circuit will fit and what the drawbacks might be when measuring the current from the low side.
Any comment or suggestion is really appreciated.
Thanks
I realized the MOSFET switch part of my first schematic was wrong.
I should use a couple of P-Mosfet on the High Side or i could use the two N-Mosfet on the Low Side.
I opted for the second solution.
This should correct (I hope...)

Schematic.jpg
 

crutschow

Joined Mar 14, 2008
29,466
Q10 is connected as a source follower, so its source voltage will be no more than 3.3V minus its Vgs(thres) voltage, which will not be enough to turn on Q11 and Q12.
What's the Vgs(thres) of those transistors?
What is the Charge Control (On/Off) voltage?
 

Thread Starter

mspider65

Joined Sep 20, 2021
7
Q10 is connected as a source follower, so its source voltage will be no more than 3.3V minus its Vgs(thres) voltage, which will not be enough to turn on Q11 and Q12.
What's the Vgs(thres) of those transistors?
What is the Charge Control (On/Off) voltage?
Yes, you are right, the circuit is wrong.
I could directly drive Q11 and Q12 from the MCU as they have a low Vgs threshold (about 1v).
However, I realized that, even in this condition, the circuit will not work because, when the charger is powered and the battery is discharged, the charger will have a higher voltage than the battery (up to 16v difference in the worst case) and the -Charger will be negative (+ Charger and + Battery are connected together) and the drain of Q12 will have a negative voltage and Q12 and therefore Q11 will also be turned on with 0v on the Gate.
I haven't found a solution to make a low side bi-directional power switch for this situation and will probably have to go to the high side and try to find a low Rds P-Mosfet.
 
Top