As the title says, I have a signal which is mostly DC but has some RF ripple on-top of it.

The signal comes from a pickup coil that receives a 13.56Mhz signal (RFID). That signal gets rectified by a high-speed diode bridge rectifier. This is followed by a leaky low-pass filter, e.g. a small capacitor and load resistor in parallel. (Note: I can't change any of this acquisition circuit).

Now I want to get an accurate measurement of the RMS voltage.

Currently I just hook it up to a digital oscilloscope and use the internal measurement functionality. This however gives slightly different readings depending on the time-base and voltage range selection.

Since I want to automate this measurement in my lab I thought about building a little box that does a more precise RMS to DC conversion in analog, so I can use a precision voltmeter to do the measurements. However, all the "true RMS to DC" chips that I've seen so far won't deal with signals of higher frequencies than a megaherz or so.

Any ideas how I could accomplish my goal?

  • \$\begingroup\$ Do you want to measure the RMS of the RF signal without the DC, or the average of the DC value ? \$\endgroup\$
    – bobflux
    Sep 2 '17 at 11:34
  • \$\begingroup\$ The average of the DC value. \$\endgroup\$ Sep 2 '17 at 11:56
  • \$\begingroup\$ what is the burst length? and what accuracy do you need? (absolute) \$\endgroup\$ Sep 2 '17 at 19:47
  • \$\begingroup\$ Precision rectifier for each bit needed? or average DC in 1ms? \$\endgroup\$ Sep 2 '17 at 20:07

OK, so from what you say

  • You have RF on top of DC
  • You want to measure the DC voltage, averaged over "some delay", and ignore the RF
  • The RF makes your multimeter (or other equipment) misbehave.

If this is what you need, the answer will be a simple filter, for example a second-order LC filter with optimum damping, so L - R - Cap to GND. This will remove the HF, smooth the voltage, and average it a bit, depending on cutoff frequency. You could also use a notch filter, but this would not remove the harmonics very well.

Now, a filter will of course slow the transient response, so if the "DC" voltage you are trying to measure is not really DC, but varies at say 100kHz, then you'll want your cutoff frequency to be:

  • High enough above 100kHz for good transient response
  • Low enough below the 13.56MHz you want to remove, so it is actually filtered out.

I'll update when you give more info about this.


Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.