what is psophometric noise and how to measure it. please help meee.

[Teja.p]

i have one SMPS ... my asking it psophometric noise in mv ...
please help how to measure it and what is psophometric noise ...

 

[Warpspeed]

Psophometric noise measurement goes back to the original roots of the telephone industry (early 1900's) where it was found that very long distance telephone circuits suffered from different kinds of noise and interference, hum, whistles, crackles etc....

But how to measure it in a meaningful way ? It was pretty obvious that some interfering frequencies would be inaudible, say below 20 Hz, or above perhaps 18 Khz. But other frequencies such as 1 Khz could be particularly annoying, even at quite low amplitude.

So Bell Telephone Laboratories came up with a test that involved measuring noise in a telecommunications circuit after it has passed through a very carefully shaped frequency selective filter.

This filter has very high gain around mid range frequencies that the human ear is most sensitive to, and are the most subjectively annoying, and much lower gain at the lower and upper ends that can hardly be heard even at high amplitudes.
So noise measurements could be taken and referred to as "a psopometricaly weighted noise measurement".

Its only usually ever used on communications channels used for speech, and sometimes for noise on the output of dc power supplies used for telecommunications purposes. Its sometimes seen in specifications for telecoms equipment.

The quality of voice channels these days is now so good, that this measurement method has become long obsolete.
Very few people these days even know what pshophometric noise measurement even is.

Once in a moment of complete madness, I bough a Hewlett Packard 3556A psophometer and its user manual at an auction for fifty cents.
Have it here with me now.
Nobody else at the auction seemed to even know what it was.

As far as I know its the only instrument ever made that can make these special measurements, it looks like this:
https://www.tulsa-ads.com/Red-Fork-/Children-/Clothing-/Hewlett-packard-hp-3556A-psophometer.JSP

 

[Teja.p]

Thanks u sir
i have question ..how to measure noise in mv by using this meter

 

[betwixt]

Your simplest method would be to measure the voltage from a microphone through a tunable filter then apply a psophometric multiplication factor according to the filter frequency. In other words, take a linear measurement then convert the results mathematically using one of the uLaw or A-Law (or other hearing) response curves.

Brian.

 

[Warpspeed]

Its 3Am here right now.

I will get back to you later if I can locate the user manual and post a frequency response curve.
There should be a schematic there too, if I remember correctly.

Its basically just a shaped bandpass filter followed by a millivolt meter that reads in either millivolts or dbmv on the scale.
So you hook it straight up across the output of your dc power supply through a coupling capacitor, and just take a meter reading.

You can then quote your power supply has xxx millivolts of noise on the output psophometricaly weighted.
There is really no substitute for having the correct instrument to do this.

 

[FvM]

Or make your own A weighting filter which is defined in the standard as follows

Four differentiating zeros 
Two poles at -20.6 Hz (-129.4 /s) 
One pole at -107.7 Hz (-676.7 /s) 
One pole at -737.9 Hz (-4636 /s) 
Two poles at -12200 Hz (-76655 /s)

See also https://www.beis.de/Elektronik/AudioMeasure/WeightingFilters.html

 

[Warpspeed]

The slim operators manual I definitely have here somewhere I cannot find right now.

But I hit the jackpot, the full service manual is available on the internet, here :

https://www.testequipmentconnection.com/specs/Agilent_3556A.PDF

Anyhow, these are the two different frequency response curves.

 

[FvM]

I realize that psophometric measurements uses ITU-T weighting rather than common A weighting, it's e.g. specified in this ETSI standard under E.2