rms measurement of frequency varying sine signal

[nm2007k]

how to build a circuit to measure rms value of the sine wave signal with frequencies vaying from 1 Hz to 1 KHz.

the voltages is of range +/-20 mv(if it is unamplified), +/- 10 V if it is amplified.


Thanks in advance.

 

[Audioguru]

Since the input signal is a sinewave then you can use a simple peak detector circuit and reduce the gain by 0.707 times for the RMS reading.

 

[KlausST]

Hi,
There are ICs that do the RMS to DC conversion.
with a capacitor you can adjust cutoff frequency of the output signal.

Klaus

 

[D.A.(Tony)Stewart]

Unless there you don't mind measuring Signal + Noise, the appropriate method is to convert Peak to RMS assuming Sine wave and RMS=Peak/Root(2) or 0.707 x Vp as Audioguru stated.

If you want to eliminate noise, then a tracking filter is needed.

 

[crutschow]

The effect of signal noise can be minimized if the rectified sine wave is filtered to give the average value instead of the peak. Then just multiply by the RMS/Ave ratio to convert to RMS. The conversion ratio depends upon whether you half-wave of full-wave rectify the signal. For best accuracy use a precision full-wave rectifier before doing the averaging.

 

[jay1691]

hello,

As stated in upper thread. I have same problem. I want to find RMS for sine wave signal with frequencies varying from 1 Hz to 5 KHz and the input voltage range is +/- 10 V.

Here as said in thread by crutschow i am using full-wave rectifier circuit for rectifying single before averaging but i cant understand how to decide capacitor value for whole frequency range(1 Hz to 5 KHz) or can any one provide averaging circuit using op-amp.

full wave rectifier circuit is attached in image with its reference link.

https://www.analog.com/library/analogdialogue/archives/44-04/absolute.html

Please suggest me further circuit for finding average of that signal.

 

[crutschow]

If you want a single filter with settling time the is significantly faster than an RC passive filter, you could use a low-pass 2nd-order active filter at the rectifier output.
For a ripple of no more than 1% of the average value at the lowest frequency of 1Hz use a 2nd-order Butterworth filter with a corner frequency of about 0.13 Hz.

For easy design of such a filter you can use the free FilterPro program from Texas Instruments.

 

[FvM]

A second order filter would be the standard solution. Alternatively you can think about an analog or digital signal processing that averages the rectified signal over one (or an integer multiple of) signal period(s) by means of a zero crossing detector.

As a side remark, I would prefer to put "rms measurement" in quotes because you are describing an averaging precison rectifier, not a rms measurement. The circuit can be calibrated to display rms values for sine input voltage, but can't give correct rms measuremet for other input waveforms because it's no true rms circuit.

 

[schmitt trigger]

On the other hand, if you really require true RMS sensing, go to the Analog.com or Linear.com websites and search for "RMS conversion".
Both have single-chip solutions for true RMS sensing.

 

[jay1691]

Dear schmitt trigger,

As you said i search for true RMS sensing IC for this application. but i cant found any IC which will work for full range (1 Hz to 5 KHz ) of frequency. As i want to use same IC work for 1Hz as well as 5KHz input signal.

- - - Updated - - -

Dear FvM Sir,

Thank you for replay. As input signal for my circuit is only sine wave so "averaging precision rectifier" will work for me there is no need to true RMS measurement. Am i right ???

- - - Updated - - -

Dear crutschow,

As you said low-pass 2nd-order Butterworth filter with a corner frequency of about 0.13 Hz. Can you please give any practical or reference circuit for this design.

 

[crutschow]

..........................
As you said low-pass 2nd-order Butterworth filter with a corner frequency of about 0.13 Hz. Can you please give any practical or reference circuit for this design.

Use the free Filterpro program from Texas Instrument that I mentioned in my previous post.

 

[schmitt trigger]

Dear schmitt trigger,

As you said i search for true RMS sensing IC for this application. but i cant found any IC which will work for full range (1 Hz to 5 KHz ) of frequency. As i want to use same IC work for 1Hz as well as 5KHz input signal.

Analog Devices has a very comprehensive app note (which actually made it into a short booklet) for RMS conversion.
Essentially it outlines that the lower cutoff frequency can be made as small as desired, as long as you are willing to tradeoff either settling time and/or ripple.