using a volt meter to measure duty cycle

[circuitmangler]

Is there a simple way to use a volt meter (DMM or analog) to measure duty cycle? Assume the input is very clean -- steady and "rail to rail" signalling. Duty cycle is then just the time average of the input divided by the supply voltage.

My gut tells me there ought to be a simple way to do this. Would an analog meter automatically do it (just divide the meter reading by the supply voltage)? Would a DMM do it itself, or would it need some "help"?

Cheers,
circuitmangler

 

[OperaHouse]

Used to do that to measure dwell. Remember when cars used to have points! Zener and resistor fixed the max voltage and the meter movement integrated it for you. 100% is full scale, nothing more to it. You can use an op amp in a long integrating mode if all you have is a digital meter now.

 

[felixc]

If your circuit can be loaded a bit and if the drive is symetric, a simple RC (with a time constant about 10-20 times the pwm base frequency) should convert the oscillation into a DC voltage which you can read with a DMM.

 

[MMurphy]

circuitmangler,

Are you looking at a fixed frequency or does it vary? What kind of accuracy do you want? Do you need to calibrate this or is some quick and dirty/cheap solution adequate?

OperaHouse and felixc have some good suggestions. You might also consider a monostable multivibrator if the source can't take the loading of felixc's suggestion.

If you are considering buying a new meter anyway, some DMMs have frequency/period/duty cycle functions included.

 

[circuitmangler]

Thanks for the tips.

The frequency will change but only when I deliberately
change it, so t would be nice if I didn't have to re-tune
the circuit every time I changed the frequency, although I
wouldn't mind hitting a "reset" button or some-such. Or if
the circuit could average over the last N seconds for some
small value of N.

I've looked into DMM's that have duty cycle capabilities,
but I really can't justify the expense right now. I'm just
looking for something really cheap so I can get a rough idea
of what the duty cycle is (say, to within +/- 5%).

Thanks again for the suggestions.

 

[biomedicalEng]

Hi,
If your input signal is a pulse and you know it's peak voltage and it's frequency is below the filtering rate of your DMM, then you can measure the duty cycle by your volt meter.

_V1_
| | |
| | __V2__|
<-t1-> <-t2->

Due to above values and asume that VDC is the displayed value on the volt meter, we have:

VDC=((V1*t1)+(V2*t2)/(t1+t2))

if V2=0 then VDC=V1 * t1/(t1+t2) where t1/(t1+t2)=Duty Cycle

Hence:

Duty Cycle= VDC/V1

and Duty Cycle% = (VDC/V1) *100

Best regards,

Kasra Ravanbakhsh

Email: info@kdi-co.com

Web:

http://www.kdi-co.com

 

[lcsjk]

If your frequency is 60 Hz to about 200Hz, an analog meter, D'Arsonval, will work. If the frequency is too slow the needle will vibrate. With a low cost dvm, use a 10k resistor and a 10-100uF capacitor in series and measure the voltage across the capacitor. This will give a voltage directly proportional to duty cycle. If this is too slow reading or too fast reading, adjust the R x C so that RC is about 10x or more times the period of your input. This approach was mentioned earlier in the discussion. If your frequency is in the MHz range, wiring and lengths will start giving errors.