20KHZ DC FREQUENCY SIGNAL

[DAVIDES]

HELLO

I have  20 channels 20V@20KHz DC signal with different dty cycle

I want to monitore with a presicition of  0.5% the output but I'm not sure if a multimeter can achieve this precition in VDC mode. I'm not interested in store signal just average value in a scope (mabye 2 values per second), as a regular multimeter, but i'm not sure if using a DMM with a multiplexor is enough due that is a DC signal but with 20KHz of frequency. If anyone has expierence with this please help.

Regards

[johnsold]

Your explanation is not clear.  DC is not 20 kHz and 20 kHz is not DC.  What do you really have?

Can you post an image (.png) or a VI with data saved as default showing the signal you are trying to measure?

Lynn

[DAVIDES]

Thnks for your answer sorry if i was not clear. It;s a 20KHz square wave PWM regulated. It is DC because just go from 0 to one direction either positive or negative but not both. Check Picture from a hanheld oscilloscope. What it's important is not capture the signal just the RMS value, but I'm not sure if a voltmeter can do it accurately.

[Dennis_Knutson]

You've got a strange definition of dc.

Why did you post to this board if you have a question about a volt meter? The specs for your meter should have the max frequency. A DAQ Device with the correct sample rate could do the job.

[johnsold]

I thought it might be something like that, but I did not want to guess.

DC voltmeters respond to the average value if the pulse period is shorter than the meter's response time.  At 20 kHz this will be true.  AC voltmeters fall into two categories.  True RMS meters and average responding meters calibrated to display RMS for sinusoidal signals.  For the image you posted a DC voltmeter would read about 2/3 of the difference between the high and low levels of the square wave.  It appears to be about 1.8 divisions, giving an average of about 1.2 divisions.

The precision and speed depend on the meter you are using.  A digital voltmeter with 2 1/2 digits (maximum reading = 1.99 * range) will have 0.5% resolution.  Its accuracy may be better than that or worse. A 3 1/2 digit meter will have 0.05% resolution and likely will have accuracy substantially better than 0.5%.  An analog meter might be that good it if is a laboratory grade meter with a mirror scale.

You mentioned a DMM with a multiplexer.  What is the model or part number of the equipment you are using?

Lynn

[DAVIDES]

I'm considering if I need a DAQ for obtain the full signal and then processing in order to obtain VDC, or if a multimeter with a multiplexor can do a thejob. I don't need the full signal this test is for about 8 hours long so I just need VDC

[DAVIDES]

johnsold.

I'm considering use Data Logger from Agilent HP34970 but I'm not sure if it is better use a DAQ acquire the full the signal and process with software, I prefer first option because looks simple but I'm not sure about frequency response of multimeter. The question is I don't have neither DAQ nor HP34970 and I want to be sure what is the best option in order to buy the right one.

[johnsold]

The 34970 uses multi-slope integration for its DC measurements, so the frequency should not be a problem.  The only precaution would be to avoid autoranging.  Keep it on a range which includes both the high level and the low level voltages of the pulses.  If the pulses are 0 to 10 V and the duty cycle is 0.5%, the average voltage is 50 mV.  If the meter was set to the 100 mV range, during the time the pulse is at 10 V the meter input would be overloaded.  The meter would need to stay on the 10 V range regardless of the duty cycle.  The meter can resolve 10s of microvolts on the 10 V scale so this should not be a problem for your application.

Lynn