Definitives

[robmil]

Hi All

i think it was the school I went to but I really enjoyed science and physics especially the science experiments 'cause, effect, result, reason'...loved it all.

The problem I am having here is that there are too many folks here measuring using their own 'methods'. Sometimes the results say one thing by someone and sometimes the same product/design will give a different set of results by someone else.

I'm trying to measure regulator noise. Tangentsoft shows his method and then goes to explain ALW super regs. The ALW regs give a set of results that are amazing BUT... and here is the BUT, Tangentsoft cannot verify or agree with the results because the measuring method used is very different to his own. That means that if ALW prints that the noise specs are -122dB @1Khz he could be lying or with someone elses test-bed it could measure 10dB better!

Has anybody tried to measure a LM7815 and ended up with the same results as the manufacturer?

Then you ask Aglient or TTi technical - 'how to measure regulator noise' and again the answer is VERY different again.

Are there any standards out there where x=x and y=y?

It is frustrating for a novice like ME who has the tools, is willing to learn but sees too many different ways.

There should be a Audio Measuring Bible out there or at least one that the diyaudio community should follow.

There's my rant.....sorry for wasting your time

Rob

[Nico Ras]

Are you sure that the results are different and not just normalized to some other quantity. one may measure noise power density per finite bandwidth, some would measure noise voltage if the impedance is ill defined and so on, but mostly can these measurements be converted or normalize to a common quantity.

[sakis]

he he he .....
beyond procedure be aware that when it comes to noise or distortion figures many of dear forum members post the results of their simulator ....

obviously in real life most designs are compromised by the pcb /choise of parts /and total implementation resulting by far diferent results than the ones originally posted ...

think, that this has nothing to do with the procedure .... that is another diferent ball game ....

i can assume that manufaturers also create specs for parts only at ideal conditions

happy regards
sakis

[sofaspud]

Manufacturer's often show their test circuit in the datasheet. They also often provide a range for specs. It isn't easy to make identical components given different wafers, batches, fabs, etc.
I'm not a member, but I would think AES has some kind of measurement standards for various aspects of audio.
I think my best advice would be to use the procedure that is most meaningful to you.

[robmil]

Sofaspud - QUOTE]I think my best advice would be to use the procedure that is most meaningful to you.[/QUOTE]

Sorry but this is the problem. This shouldn't exist because something meaningful to me may not be meaningful to someone else.

Nico... I'll send over the link re: tangentsoft/ ALW regulators and the Tangentsoft statement.

I think what we really need here is somewhere on the forum that lists test procedures and how-to or else how do we know what we are measuring is right?

[sofaspud]

Fair enough, but for whom are you measuring? For me?
I really can't help but wax philosophical on your statement, unfortunately.
But your test procedure idea has some merit, and has been proposed here in one form or another previously. I can't say how popular it would be however, and whether enough contributions would come in to make it worthwhile.

[robmil]

I do think this is as important as designing circuitry..... the difference is 0.002% THD is very different to 0.02% THD!

We really do need standardised test prodcedures, at least then we'll all know where our circuits 'actually' stand.