RMS to DC Converter using Passive Components

[bouncetherabbit]

Hi all,

I've got a PWM signal and i need to convert it to a straight line dc voltage. Is it possible to do it with passive components? I know there are ICs out there that do the conversion. Are there ICs that do the conversion without a supply voltage?

Thanks!

 

[biff44]

A simple series resistor followed by a shunt capacitor will do that. You have to play around with the values to get it right, based on the pulse frequency.

 

[bouncetherabbit]

Something like a simple low pass RC filter? Are there any specific formulas for me to decide the values of R and C?

I kind of tried using the RC (random values) filter with a 50% duty cycle but it gives me max. value of the square wave. If the duty cycle isn't 50%, it doesn't seem to be in a straight line. Is it due to having wrong RC values?

The dc value should be equal to Vavg = V(2d-1)

d= duty ratio
V= max value of PWN signal.

Thanks!

 

[dick_freebird]

If you were clever and had a slow measurement allowed,
an electrical-thermal scheme would work across a wide
range of duty cycles / crest factors and so on. This seems
in fact to be what is done for RF power meters (high enough
frequency that active processing is not reliable for high
accuracy).

You'd have a "heater" resistor, a close-coupled (thermally)
thermistor or temp sense diode, and a well characterized
thermal path (maybe a second thermistor in a bridge to get
a delta-T (equals power). Getting back to volts would need
a square root.

Cheapo voltmeters do RMS conversion with rectification
and a 0.707 scale factor but this only works for sinusoidal.
"True RMS" meters are more elaborate.