pre standard two preamp, and development of measurement methods

[BV]

You want see stepped spectra vs. amplitude, or not? It is possible only single ton measurement, like this (-80to 0dBV, 10kHz)

 

[tryonziess]

Pavel and Wavebourn,
Thanks for that link. They have a good assortment of stuff. It is also compatible with Vista. About time.

Tad

 

[Wavebourn]

You want see stepped spectra vs. amplitude, or not? It is possible only single ton measurement, like this (-80to 0dBV, 10kHz)

Yes, I want to see spectra vs input level, on the whole frequency band. Spectra must be shown in respect to input level.

 

[jan.didden]

You guys rock. Keep going!

Jan Didden

 

[Wavebourn]

You guys rock. Keep going!

Thanks!
Actually, it would be easy if to have source codes of something close to what we want to see.


Let me install and try what Pavel suggested, then may be I will contact developers suggesting them valuable improvements.

For now, which sound card is one of the best for such audio measurements?

 

[PMA]

Lynx L22 (BV has one), ESI Juli, EMU (number forgotten).

 

[KSTR]

I would like to back the suggestion of Spectralab, this is one of the few truly universal software FFT-analizers/toolboxes with lots of functionality and it has control features for automated measurements, too.

What you need to get waterfall plots is to pass each data slice as blocks of a FFT-length, normalized to ref. signal level beforehand if required, in a .WAV file (in 24-bit fixed point resolution, for example). However, the true effort is how to get/record and extract the data to pass.

A software were you really get something good for your $$$

- Klaus

 

[KSTR]

And if the task is only 2D/3D visualization of existing data:

gnuplot, but that is NOT point&click stuff.

Note: this is a "backseat-driver" recommendation, I haven't played with gnuplot myself, but it is an widely accepted standard in all kinds of science and engineering.

 

[syn08]

For now, which sound card is one of the best for such audio measurements?

EMU 1212M, you can't beat that for $150. Avoid M-Audio, they are not worth the money (as a measurements tool). Lynx are very good but, to me, not worth the money ($600-$800).

Here's what EMU 1212M can do in balanced mode. In single ended is much worse, but still acceptable for the DIYer.

I've looked into that (have a couple of Sun Fire V100 blades myself), not a chance in hell.

 

[PMA]

A proper place to let it die. Who cares? We can still discuss 40 years old stuff. Do not get me wrong, I am not young at all. Anyone can learn and grow!!

 

[Wavebourn]

Sure, a technical discussion about measurements of amps in "everything else", while JC's personal blog in transistor amps...

I am out of the forum from now.

 

[PMA]

I would like to start a thread on my new preamplifier project, pre standard two –
http://web.telecom.cz/macura/prestd2_en.htm
and up-to-date measurement methods related to solid state amplifiers/preamplifiers.

The preamp is intended as a study how to get the best possible suppression of interferences, both EMI field and PSU line, and the best sound of course

I do not plan to show a detailed schematics, but I would appreciate a discussion on measurement methods, with results shown here.

I would also like to ask moderators to move recent discussion on 3D measurements from the dead zone
http://www.diyaudio.com/forums/showthread.php?s=&threadid=146830
to this thread, if possible. This thread would be definitely solid state related.

 

[PMA]

Thanks.

Possible topics might be IMO:

- cleanest measurements as possible (mains components suppressed as much as possible)

- external amplifier to see real noise floor, not measurement card noise floor

- 3D waterfalls, as already shown

- UHF spectral analysis to see suppression of HF interference components

- sine/square method

- spectrum of square to see if even harmonic components were generated

 

[PMA]

Something to start with -

amplitude response of quite usual audio design, extended to 1GHz:

 

[PMA]

Compared to amplitude response of a 100MHz amplifier, designed according to HF rules:

 

[PMA]

Here's what EMU 1212M can do in balanced mode.

This is pretty good. I expect that the higher frequency components get worse with increased amplitude?

 

[BV]

This is pretty good

Yes, it is. But I suppose, calibration is not done correct, and level is in respect to dBFS, not dBV rms..Noise floor under -150dBV at showed settings seems unrealistic.

 

[syn08]

This is pretty good. I expect that the higher frequency components get worse with increased amplitude?

Yes, about linearly. Unfortunately the EMU drivers are not allowing using the internal loop over 48KHz sampling, so an external generator is required.

I have posted quite some time ago the reasons why a sound card is though not the best tool for serious measurements, I'll find the reference and repost.

 

[syn08]

Yes, it is. But I suppose, calibration is not done correct, and level is in respect to dBFS, not dBV rms..Noise floor under -150dBV at showed settings seems unrealistic.

It is done correctly. Look in the picture for the averaging, sample size and RBW.

 

[PMA]

I calculated that, and it may be done correctly. Sound card inputs differ in opamps used and topology. The values shown are not in contradiction with theory.

My card input is worse, the best noise floor result I obtain from preamp measurement is attached, and it is a noise floor of the soundcard. Related to dBVrms.