I have designed a power supply which has a front-end buck, which is current-mode controlled. In my simulations at the moment my current sensing is implemented by a current sense transformer. I also see a lot of people using resistors, but I have came across the Allegro current sensor surfcae mount package: https://www.allegromicro.com/en/pro...egrated-conductor-sensor-ics/acs730#Packaging.

What are the benefits of using a hall-effect sensor compared to the other techniques? Which is best for high frequency, 500kHz-1Mhz signals? Benefits/drawbacks/ etc.

Thanks in advance!
J

 

The benefit of such a device is isolation. The downside, depending upon how you want to use the current information may be the phase shift. It is hard to beat a resistor for simplicity as a 1 MHz current sensor.

 

The benefit of such a device is isolation. The downside, depending upon how you want to use the current information may be the phase shift. It is hard to beat a resistor for simplicity as a 1 MHz current sensor.

Okay, that makes sense. So it may be good for sensing current on a high voltage secondary of a transformer?
My issue with 1MHz for a CS transformer was the AC losses, but I don't know enough to know whether that will be an issue?

 

What power supply topology are you considering? Isolated or non-isolated?

 

What power supply topology are you considering? Isolated or non-isolated?

The supply is a buck converter followed by an isolated push-pull stage. The push-pull is "uncontrolled" as of now, but I was thinking to add a current sensor to those switches also, in case I want to add control complexity at a later point.

 

So the buck stage provides the regulation and the push-pull stage provides the isolation?

To what end would you sense current in in the push-pull stage?

Sorry about all the questions but so far I don't understand the basis of your question.

 

So the buck stage provides the regulation and the push-pull stage provides the isolation?

To what end would you sense current in in the push-pull stage?

Sorry about all the questions but so far I don't understand the basis of your question.

No I understand the confusion. I have designed a high-voltage generator which must operate at fixed frequency. Therefore the buck output current is regulated to the push-pull converter which provides the isolation, yes. The transient response of the regulator is very important (it is a radar transmitter application).Therefore I imagined that it could be used in the future for tighter control. But I have not really thought about that much yet.