[time-nuts] How to measure regulator noise?

Bruce Griffiths bruce.griffiths at xtra.co.nz
Tue Jan 8 14:54:53 EST 2008

Don Collie wrote:
> I`d just hang an AC millivoltmeter[or microvoltmeter] across the regulator`s 
> output.
> I use my H/P 400H, which will give readings down to about 50uV. If your 
> regulator produces less noise than this [say a 723, with
> 2uV], then you`ll need a more sensitive meter.
> It might be wise to short the input wires of the meter, to check the 
> inherent noise in the measurement setup - for meaningful results this should 
> be, perhaps, no more than a tenth
> of the expected reading [depending on how riggerus you want to be]. Bear in 
> mind that most analogue meters are average responding, and calibrated for a 
> sinewave - so to get the RMS value for white noise [or whatever] a 
> correction factor should be applied [depending also on what sort of accuracy 
> you need/want].
>      If you want to *look* at the noise, I understand that one of the Tek 
> 7000 series plug-ins is able to display very small amplitudes of this order. 
> Again, because you may be dealing with comparitively small voltages 
> [compared with those which might be induced due to hum fields etc],
> it would be wise to check the residual noise by shorting the probe[s] at the 
> regulator. This is valid since the output impedance of the regulator is 
> nearly zero.
> Go to it!,..........................................Don C.

This method doesnt give the noise spectrum which can be important for
use when powering VCXOs etc.
The better discrete low noise regulators have very low noise so one
still needs a low noise preamp.
For noise measurements at frequencies below 10Hz a preamp with low noise
and a low cutoff frequency can be challenging.


More information about the time-nuts mailing list