I'm having a bit of trouble coming up with the most practical fix for my buck/boost converter PCB. To give a bit of an overview, I'm using a current sensor to read the current value through an inductor, that output is fed through an op-amp to amplify the signal, and then that is finally fed into comparators that controls a circuit to switch two MOSFETs on/off to maintain the inductor current between a min and max value.
In short, the circuit goes: current sense -> op-amp -> comparators -> MOSFET driver.
This worked well in simulations, however, on my PCB, the current sensor's default voltage output when it doesn't sense a current is 2.5V. Because of this extremely high voltage, the value goes right past the comparators and keeps one MOSFET on, without ever switching the other. The normal voltage values that the current sensor should output is between 0-500 mV. I only need to get past the initial default 2.5V on the output by pulling the line to GND briefly (to get the other MOSFET to turn on and allow current flow to the inductor) to get my circuit to start working.
I've brainstormed two ways to go about this, but I would love to get additional feedback.
-
Find a diode with a forward voltage greater than 500 mV, and solder it directly to the output of the current sensor. This way, whenever the voltage on the line goes above 500 mV (would never get this high during normal operation), the diode will short straight to GND.
-
Use an additional comparator on the current sense output that pulls the line low whenever it detects 2.5V at the output and 2.5V on the other input pin. Since my PCB has already been fabricated, I would need to run about 8 wires that connect to this IC all over the board. Not ideal, but I think it will get the job done.
What are your thoughts? Thanks for all your help. I've included a link to the part I'm using below.
https://www.digikey.com/en/products/detail/allegro-microsystems/ACS712ELCTR-05B-T/1284593
Best Answer
It seems like there is a fundamental conceptual disconnect here. The ACS712 output is offset to mid-rail to facilitate reading voltage when the current can be either positivist or negative. The ACS product literature calls this "bipolar." Some ACS current sensors are unipolar (although they may be offset slightly above ground anyway).
It seems like your best bet to fix up your circuit will be to remove the 2.5V offset at the op-amp. Please note that the ACS devices are "ratiometric" meaning that the offset will be very close to mid-rail (VCC/2). So rather than use a precision 2.5V reference to generate that offset at your op-amp, you would want to use a voltage divider from the same VCC that powers the ACS712, if possible.