I'm thinking of making a power meter circuit which will have a high-side current-sensing amp (LTC6102), a resistive voltage divider for sensing the voltage, a dual op-amp (OPA2365) as a buffer for the current and voltage signals, a 2 channel simultaneously sampling ADC(LTC1407), an FPGA for multiplying and keeping a running sum and a blue-tooth UART module for transmitting to a phone.

The ADC and the rest of the digital circuit will be powered from 3.3V, and will draw 70mA. This will be measuring an e-bike's consumed power with the maximum supply voltage of 54.6V. I would use a liner regulator but then I need a radiator because I need to dissipate 3.5W considering the high voltage drop.

So I'm considering using one of these DC-DC converters with a built-in inductor to save space with less dissipated power.

My concern with these switching converters is noise.

I can estimate the impact of the power supply ripple based on the converter's specs and the PSRRs of the op-amp and the ADC (the current-sense amp will powered directly from the supply)

What I don't know is how to estimate is the impact of the radiated EM noise from the inductor in the converter.

My understanding is that when it comes to EM-induced noise I should be concerned about the places with low currents/high input impedances.

In my case these are the connections:

  1. from the output of the current-sense amp to the buffer op-amp (1LSB is 2.5V/4096=610uV or 1mA/4096=244nA for the current sense amp's output)
  2. from the resistive divider to the buffer op-amp (same 610uV LSB)

I understand that the actual currents into the op-amp inputs will be lower, but not sure how to estimate those.

This might be a broad question but as a ball-park estimate is 0.5 LSB of noise achievable in my case with a switching DC-DC converter considering the EM noise from the inductor of the converter?

(Now I'm thinking maybe the bluetooth module is a bigger EMI concern)

  • \$\begingroup\$ After searching for actual DC-DC converter modules it looks like they are relatively rare with such a high input voltage (and quite bulky, comparable to the 40x40mm heat-sink I was considering), so I may have to use a linear regulator with a heat-sink after all just for this reason... \$\endgroup\$
    – axk
    Mar 20 '19 at 23:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.