How to measure the Vpp, Vrms of a low frequency signal (<10Hz)

[crossrulz]

---

@Rajashekar wrote:

How to capture the signal for exact integer multiples of one full cycle? Can you provide some details please.

---

I was planning on releasing this to OpenG.  It's just not quite ready for that.  In the mean time, try this.  In simple terms, it calculates how many samples of a waveform to keeps and cuts down the data array accordingly.

[johnsold]

Rajashekar,

1. I recommend that you use DC coupled measurements, not AC.  For frequencies less than 1 Hz, or more realistically < 10 Hz, treating the singal as a slowly varying DC voltage avoids many of the problems associated with the typical assuptions made by the designers of AC instruments.

2. The snippet posted by crossrulz will work if you know the frequency. If your signal has varaible frequency or the frequency is not known, then you need to use other techniques.  Zero-crossing techniques are relatively easy and work well when the noise level near the zero crossings is low.

Please tell us the nature of the signal you are trying to measure and what information you are trying to determine about the signal.  Perhaps someone can suggest a better way to do what you want to do.

Lynn

[Rajashekar]

Hi,

The following are the two types of testing conducted where in each type a filter is present in the circuit and need to verify the Frequency Response for the same. At each input frequency the output will be monitored and checked for lying with in the range.

1. The test is conducted by varying the frequency of the modulating signal with all other parameters fixed. For each input frequencies the lower and upper limits for the final ouput signal (amplitude Vpp or Vrms) are defined. At each input frequency the final o/p will be monitored for the amplitude of the signal and checked for the signal lying within the range. Refer Diagram 1.

For Ex: 1) Mod freq: 1Hz to 200 Hz, Carrier Freq: 2000Hz

            2) Mod Freq: 0.16Hz to 16Hz, Carrier Freq: 400Hz

2. The following is the flow of signals. The test is conducted by varying the frequency of the input sine signal with the amplitude fixed. For each input frequencies the lower and upper limits for the final ouput signal (amplitude Vpp or Vrms) are defined. At each input frequency the final o/p will be monitored for the amplitude of the signal and checked for the signal lying within the range. Refer Diagram 2.

For Ex: Sine freq: 1Hz to 2000 Hz depending on the cut-off frequency of the filter.

We are using the instruments PXI 5412 - 2 Nos., PXI 4070, PXI 5122, PXI 2532.

Thanks & Regards,

Rajashekar

[Rajashekar]

Hi,

You mean to say AC Volts DC Coupled configuration in DMM.....

In Digitizer i have configured for DC coupling only..

Regards,

Rajashekar

[crossrulz]

It sounds like you know what frequency should be coming out.  Therefore, I would digitize with the DMM at some high rate and truncate the waveform using the method I gave above.  Now that you have an integer number of cycles for the output frequency (you willl need to calculate this frequency for the modulation), calculate the RMS, Mean, and whatever else you need to.

[johnsold]

Rajashekar,

Thank you for giving a more complete description of your process.

I am suggesting that you do not use the AC Volts mode because the lower frequency limit seems to be 1 Hz even when DC coupled.  Run the digitizer in the DC Volts mode with the sampling rate high enough to get enough samples per cycle to meet your error requirements for the Vpp or Vrms measurements.

Are you digitizing only the modulation frequency signal or the modulated carrier signal?  Are all the signals sine waves?

I am not familiar with the instruments you are using. If they have the ability to use external clocks, you may be able to synchonize both the generation and digitization precisely.

Lynn

[arienai]

You can use a lock-in amplifier (like SR830, NF Li5640, Signal Recovery 5209, or Zurich instrument if you have money) that is very precise (nV up 1V rms) if you use sinusoidal reference signals (I mean if source of your circuit is plugged with sin(wt) waveform).

Or you can use DMM like 34401 or 34410 but it works from 3Hz.

Hope it helps.  

[arienai]

Usually, we use a lock in amplifier for this purpose, because it can accurately filter the signal and measure only at the requested frequency. Some popular models like SR830, Ametek, NF Li5640, of Zurich instrument, or any, are recommended. 

To my knowledge, DMM like Agilent 34410A (6.5 digits), can measure Vrms by setting the bandwidth to the lowest value, that is 3Hz for this particular instrument (that I used today for the same purpose than you). That means you cannot measure below 3Hz. 

Indeed, if you want to measure below 1Hz, the time constant, or integration time must be at least 10 times higher than 1/f. 

Another way is maybe to acquire the signal with an oscilloscope, and then to use mathematical calculation to output the rms value. 

Hope it can help,