Skip to main content
Electrical Engineering

Return to Question

deleted 33 characters in body
Source Link
Anindo Ghosh
Anindo Ghosh
  • 50.7k
  • 8
  • 107
  • 203

I have this sort of odd requirement and I'm wondering if someone with experience might have some advice on a circuit to use. I have a chip that draws 40 Amps of current at 0.85V with +/3% tolerance. Now normally I would use a DC/DC supply with a differential remote sense feature to counteract the IR drop from the switcher output to the core power balls at the middle of the chip.

However this time I'm forced to use a specific switching controller that does not have differential sense pins, just a single vsense pin. I was hoping I could maybe use an opamp setup as a differential amplifier. Then run the inputs as a diff pair under the chip and have the output drive the vsense of the regulator. Or maybe drive it through a resistor and keep the vsense hooked up locally.

Oh and one more thing I can't use any supervisor or power manager chips like the LTC2974. So it has to be a circuit I put down on the board. Crazy design constraints, I won't go into the reasons behind it.

Anyway does anyone have advice on what kind of opamp circuit I could use to accomplish this? Am I missing something critical?

Thanks for any help you can give.

I have this sort of odd requirement and I'm wondering if someone with experience might have some advice on a circuit to use. I have a chip that draws 40 Amps of current at 0.85V with +/3% tolerance. Now normally I would use a DC/DC supply with a differential remote sense feature to counteract the IR drop from the switcher output to the core power balls at the middle of the chip.

However this time I'm forced to use a specific switching controller that does not have differential sense pins, just a single vsense pin. I was hoping I could maybe use an opamp setup as a differential amplifier. Then run the inputs as a diff pair under the chip and have the output drive the vsense of the regulator. Or maybe drive it through a resistor and keep the vsense hooked up locally.

Oh and one more thing I can't use any supervisor or power manager chips like the LTC2974. So it has to be a circuit I put down on the board. Crazy design constraints, I won't go into the reasons behind it.

Anyway does anyone have advice on what kind of opamp circuit I could use to accomplish this? Am I missing something critical?

Thanks for any help you can give.

I have this sort of odd requirement and I'm wondering if someone with experience might have some advice on a circuit to use. I have a chip that draws 40 Amps of current at 0.85V with +/3% tolerance. Now normally I would use a DC/DC supply with a differential remote sense feature to counteract the IR drop from the switcher output to the core power balls at the middle of the chip.

However this time I'm forced to use a specific switching controller that does not have differential sense pins, just a single vsense pin. I was hoping I could maybe use an opamp setup as a differential amplifier. Then run the inputs as a diff pair under the chip and have the output drive the vsense of the regulator. Or maybe drive it through a resistor and keep the vsense hooked up locally.

Oh and one more thing I can't use any supervisor or power manager chips like the LTC2974. So it has to be a circuit I put down on the board. Crazy design constraints, I won't go into the reasons behind it.

Anyway does anyone have advice on what kind of opamp circuit I could use to accomplish this? Am I missing something critical?

Source Link
confused
confused
  • 1.8k
  • 1
  • 23
  • 38

Differential remote sense circuit for 0.85V DC/DC supply

I have this sort of odd requirement and I'm wondering if someone with experience might have some advice on a circuit to use. I have a chip that draws 40 Amps of current at 0.85V with +/3% tolerance. Now normally I would use a DC/DC supply with a differential remote sense feature to counteract the IR drop from the switcher output to the core power balls at the middle of the chip.

However this time I'm forced to use a specific switching controller that does not have differential sense pins, just a single vsense pin. I was hoping I could maybe use an opamp setup as a differential amplifier. Then run the inputs as a diff pair under the chip and have the output drive the vsense of the regulator. Or maybe drive it through a resistor and keep the vsense hooked up locally.

Oh and one more thing I can't use any supervisor or power manager chips like the LTC2974. So it has to be a circuit I put down on the board. Crazy design constraints, I won't go into the reasons behind it.

Anyway does anyone have advice on what kind of opamp circuit I could use to accomplish this? Am I missing something critical?

Thanks for any help you can give.