Low-distortion Audio-range Oscillator

[RNMarsh]

Originally Posted by RNMarsh

Comparing specs from Audio instrument makers ($$) --- SRS model SR1 the residual thd+n is spec'ed at -106 to -112dB. The AP 2700 is spec'ed at -112 to -115dB. I want better, unfortunately. Thx-RNMarsh

Noise floor (22KHz BW) is -118dB spec on the SRS. Now there is some fudging going on with both examples... like using 3-4 volts of output as ref. In any event, it looks like < -120 is pushing things pretty hard in analog domain for wideband sources/testers. -RNM

[scott wurcer]

Originally Posted by RNMarsh

Comparing specs from Audio instrument makers ($$) --- SRS model SR1 the residual thd+n is spec'ed at -106 to -112dB. The AP 2700 is spec'ed at -112 to -115dB. I want better, unfortunately. Thx-RNMarsh

The Christmas tree bulb was better than that out of the box.

[1audio]

Originally Posted by richiem

Hi Demian -- something isn't correct here -- I RMSed the first 9 products using my little Excel worksheet and got 0.00035%, -109dB. I assumed that the fundamental was at 0dBFS. If it is lower, and it seems to be lower than -4dBFS, then the THD is even higher than I calculated.

Edit -- with fundamental at approx -5dBFS, the THD is -104, 0.00064%

You are right, the calculation is off somewhere. I got different numbers the second time I tried it. Beta software. . .

Can you post a link to the spreadsheet?

My point was that the broadband noise can make even really low numbers not so low. And the THD+N meters (any of these older units) only measure the combination.

[1audio]

Originally Posted by scott wurcer

The Christmas tree bulb was better than that out of the box.

Here is an idea, use a DC voltage added to the AC across the bulb to steer the absolute output level.

[davada]

Originally Posted by 1audio

Here is an example of THD vs. THD+N. This is a measurement of my KH4400A. The THD is -117 dB/ .00013% but THD+N is -97/.0014%, a 20 dB difference and the noise floor is in fact quite low. Its just the effect of the energy inside the frequency band.

I will collect the different protection solutions I have docs for and we can see what there is to learn. There are anumber of different tricks used in different products. To meet the CE spec most measurement instruments need to be able to withstand something like 300V without damage (or becoming a hazard to the user) so there are a lot of good solutions.

Hi Demian,

What OS are you running QA400 on?

[richiem]

Not sure this will download for use successfully, but it's worth a try. I left the last run of numbers in to give an example.

[davada]

Originally Posted by richiem

Hi Demian -- something isn't correct here -- I RMSed the first 9 products using my little Excel worksheet and got 0.00035%, -109dB. I assumed that the fundamental was at 0dBFS. If it is lower, and it seems to be lower than -4dBFS, then the THD is even higher than I calculated.

Edit -- with fundamental at approx -5dBFS, the THD is -104, 0.00064%

I't works out to .000607% THD.

[1audio]

Originally Posted by davada

Hi Demian,

What OS are you running QA400 on?

Win 7 64 on I5 w/ 6 GB of ram.

I'll try it on the other systems in the next few days. It seems to crash a lot on this system.

[RNMarsh]

Originally Posted by scott wurcer

The Christmas tree bulb was better than that out of the box.

that one and several others as well. Whats up with these companies not being SOTA in lowest harmonic distortion attainable... and at their prices, too?? [no need to answere] -RNM
When I used the thd+noise instrument from Demian, it was reading on my little amp... .0005% . Too close to the test equip residual? The FFT output was uncalibrated so that didnt help. can it be easily/accurately calibrated? Looking for a test system to easily and accurately measure harmonics in the .00001 - .0001% area. Thx-RNMarsh

[davada]

Originally Posted by 1audio

Win 7 64 on I5 w/ 6 GB of ram.

I'll try it on the other systems in the next few days. It seems to crash a lot on this system.

Smoking.

What version of software do you have?