don’t i need to measure forward or reverse current with lowside currentsensei? This is a very confusing issue. Can we get a smooth vector in clark transform by measuring only positive values? I have an inventer. It is connected with low side current sense and has mc33174d opamp. And it is connected as an inventing amplifier opamp. There is no offset circuit. Is it possible to do FOC control with this? If I could understand this incident, I would feel relieved, but I couldn’t understand. Is it necessary to read the low side bidirectional current? Or is it valid for both lowside and inline. Thanks for reply.
Hey @drkome,
The current passing through the motor windings has some dynamics and can be both positive and negative going through the shunt resistor.
Simple example
Consider the case of a simplified case where you have only two phases a and b:
By setting different Ua and Ub values you can get positive and negative currents. As the current dynamics of the motor are much slower than the PWM changes, typically 10-20 times slower, the current cannot change as fast as the PWM does and stays (more or less) constant within one PWM period. Even if the all the mosfets are connected to the ground (low-side mosfets active). The same is true during the period when all the mosfets are connected to the VDD (high-side mosfets active).
BLDC motor currents within a PWM cycle
Now consider a BLDC motor with a diagram like this:
And if we set voltages Ua = 2Ub = 2Uc
We can divide this PWM duty cycle in sectors:
And these sectors can be represented with three different electrical circuits
As you can see within one PWM cycle the topology of the electrical circuit changes drastically, but the currents do not have time to settle within those circuits. As they have some dynamics the overall currents I_a, I_b and I_c will be relatively static within one PWM cycle. Additionally, their sum is always zero
I_a+I_b+I_c = 0.
It is worth noting that even though the currents I_a, I_b and I_c stay more or less static, the current from the power supply I_{Vcc} and I_{gnd} does not. In sectors 1 and 3 there is no current going to the power supply I_{Vcc} = I_{gnd} = 0 (as I_a+I_b+I_c = 0)
In this specific case (as Ua = 2Ub = 2Uc) the currents will settle to values of
I_a = \frac{2U_a}{3R} , I_b = I_c = -0.5I_a. So I_a will be positive and I_b and I_c will be equal and negative.
Takeaway
So you do need a bidirectional current sense for the low-side current measurement because the currents can have both signs depending on the voltages applied.