PMA said:This is what I get from class A amp at 1W/8ohm. It is considerably less distortion than Ayre or Pass. And only 2nd and 3rd components. It does not have global NFB. Would it sound better, or worse than Ayre and Pass?
Pavel,
What soundcard and what software are you using to make this measurement?
Thanks,
Bob
Since the soudncards do not interface with the PC and Windows in a syncronous manner, there is not assurance of syncronized data. Be best you can do is aligned data which statistically would be around half a sample off. I do agree that if you have the data aligned, you could average the number of periods sampled to filter out the noise. This is technique is used much when using MLS signals to measure speakers. So now I understand.gedlee said:
Bob Cordell said:
Hi,
I should have made my point more clearly. No problem
with obtaining distortion spectra at low levels...
what I was more getting at is the ear's ability.
If the ear can discern stuff -60 db below .001 watt out,
can it also discern stuff -60 db bellow .0001 watt out,
and -60 db below .00001 watt, etc.? I would
think the ears range would become compressed at low signal
levels, but gedlee seems to be saying it doesn't.
Maybe the ear's range is even more than -60 db at
these low levels?
Mike
mfc said:
It is absolutely true that the masking decreases with signal, not increases. This means that the ear can resolve more harmonic content from distortion at lower levels than at higher ones.
It does not seem to be well understood that the harmonics of a signal are simply a modification of that signals waveform - we call it timbre in other contexts. So most definatley the ear can perceive these waveform modifications more readily than it perceives incorent noise even if that noise is at a level greater than the harmonics. PLEASE stop thinking that the ear acts like an FFT. It doesn't and so these discussions which use FFT analysis are misleading at best and completly wrong at worst.
The previous image was for 8ohm load. But for 4ohm, we can see transition out of class A at some -12Vp and 12Vp. For the usual class AB low iddle current design, this would happen 1-2V from zero. Now we have inflexion points of transfer curve, source of higher harmonics. But still amp mostly works in linear portion of the transfer curve and does not annoy listeners as the low iddle current designs do.
Attachments
gedlee said:
I'm late to this thread, but I've done my share of these measurements. With an external A/D one could phase lock a good analog test signal (I've done ~-140dB at 1kHz) to the A/D clock forcing sychronicity. I mean here multiplying up the analog sine wave frequency to an appropriate clock frequency. If you're doing serious research there is no need to limit one's self to using a sound card only.
Also, even though they are maligned for no good reason to me, bridge null techniques remain a powerful tool. You can get a good -120dB null and amplify the residue. Put this into a nice low frequency spectrum analyser and you can get down to the thermal noise limit easily.
gedlee said:
Never meant to say you weren't serious. I guess I missed the point, is it enhancing the resolution of sound card measurements or measuring the true THD at reduced output powers?
BTW one of those amplifiers appears to also have some serious power supply problems (120Hz sidebands on the harmonics).
scott wurcer said:
The point that I most want to make is that we must look at
the true THD - mostly the higher harmonics - at reduced output levels. This is never published and seldom done. I use a soundcard, so its natural to look for techniques that work with them.
One amp was made in Asian for 220 volts and may not be working correctly on 110.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.