Of course, Karl, this has now opened the door, wide, for the subjectivists to say, "See, I told you changing the cables alters the sound!" ...
Absolutely
You see, many things occur at the speaker/amp interface, the cable being just one of them. You have a "perfect" voltage source in the power amp and then you stick a fixed 0.22 ohm (as here) in series with it and then some speaker cable on top of that. The load is complex and has an impedance from around 3 ohms to perhaps 20 or 30 ohms depending on frequency. So the voltage at the speaker terminals is modulated by that non linear load interacting with the resistance in series with it.Now where it gets interesting... each speaker works just as well as a microphone, a microphone of considerable current delivery. So your speakers are "feeding back" incoherent audio into the amplifier output terminals. How the amplifier deals with that is one of the big questions in the "subjective" debate imo. It all defies analysis but the one thing we can say for sure is that certain amplifiers seem to aquit themselves very well on a subjective level while others do not.
Member
Joined 2006
This is interesting Mooly : )
Would be great if you happen to have an amp that has no global feedback to try out with as well. How about adding a tube amp with an output tranny in this test?
Absolutely
Now where it gets interesting... each speaker works just as well as a microphone, a microphone of considerable current delivery. So your speakers are "feeding back" incoherent audio into the amplifier output terminals. How the amplifier deals with that is one of the big questions in the "subjective" debate imo. It all defies analysis but the one thing we can say for sure is that certain amplifiers seem to aquit themselves very well on a subjective level while others do not.
Getting closer I suspect!
Now the Reactive Load components of the 'Speaker can have varying effects upon the output of Amplifiers as in the results of some fairly basic tests using only an Inductor posted over on Speakerplans.com Forum here:
Amplifier Damping Factor - Dynamic Testing - Speakerplans.com Forums - Page 1
As to how much this might affect the subjective performance of Amplifiers needs much more work, but it certainly does demonstrate differences in performance between the samples tested at the same levels, which otherwise are unpublished.
Perhaps as well as looking at the output voltage of a test subject Amplifier, the output current could be measured and then compared?
Simple method uses a low value series resistance of ~0.05 Ohms and transformer coupling with the TX primary looking at the voltage drop across the sense resistor.
A Current Probe may be better though.
Using the same test for each Amplifier should null errors.... or at least be the same error for each measurement series.
Now there are numerous parameters which can be derived including:
VI Phase variations
Distortion - both Voltage and Current
Amplifier Output Impedance
No doubt others...
any of which or in combination could cause subjective differences
**Not sure about "a "perfect" voltage source in the power amp" bit though
Mik
Assuming an amplifier does not have a directional coupler at the output then all the returning signal sees is the amplifier output impedance, with all its frequency dependence, reactance and non-linearity. This is the same impedance you measure as an output. A particular point in a circuit has the same impedance 'looking out' as it does 'looking in', apart from when directional couplers are used.Mooly said:
I suspect that a real issue is that people spend a lot of time ensuring that the amplifer does linear amplification, but very little time considering the linearity and frequency dependence of the output impedance.
There is one proviso in this: overly aggressive output current protection could simulate a directional coupler, because it means that both voltage and current are being sensed so input and output signals can (at least in principle) be distinguished (possibly by accident!).
I have tried the same thing as Karl. Please feel free to download the resulting sound file
https://www.dropbox.com/s/7az52pby3ih7ul4/complete_audio_chain.wav
and I would appreciate your thoughts on "sound".
https://www.dropbox.com/s/7az52pby3ih7ul4/complete_audio_chain.wav
and I would appreciate your thoughts on "sound".
Thanks. Yes, that looks very workable and would resolve common grounding issues.
Well, if only... I have nothing like that for sure. It would be interesting to know how they come out of such a test.
Getting closer I suspect!
Now the Reactive Load components of the 'Speaker can have varying effects upon the output of Amplifiers as in the results of some fairly basic tests using only an Inductor posted over on Speakerplans.com Forum here:
Amplifier Damping Factor - Dynamic Testing - Speakerplans.com Forums - Page 1
As to how much this might affect the subjective performance of Amplifiers needs much more work, but it certainly does demonstrate differences in performance between the samples tested at the same levels, which otherwise are unpublished.
Perhaps as well as looking at the output voltage of a test subject Amplifier, the output current could be measured and then compared?
Simple method uses a low value series resistance of ~0.05 Ohms and transformer coupling with the TX primary looking at the voltage drop across the sense resistor.
A Current Probe may be better though.
Using the same test for each Amplifier should null errors.... or at least be the same error for each measurement series.
Now there are numerous parameters which can be derived including:
VI Phase variations
Distortion - both Voltage and Current
Amplifier Output Impedance
No doubt others...
any of which or in combination could cause subjective differences
**Not sure about "a "perfect" voltage source in the power amp" bit though
Mik
There are many variables unique to each set up and installation and not all designers deliberately aim for the "piece of wire with gain" when it comes to amplifiers. For those that do though, I think there is merit in such a procedure.
Perfect voltage... many amps come pretty close, but that closeness would be annihialated by the typical connections and the distance added from the feedback take off point to the speaker terminal. Let alone adding series resistance on purpose.
I'd never thought of that aspect Dave. In the case of these two amps, neither has additional protection circuitry as such. There is a 4 amp fuse (fast I think) in the Rotels speaker feed and the FET amp relies on me not shorting it out and the intrinsic robustness of the devices.
"They" do
I don't Your track... I'm listening on speakers... hmmm
I don't know what to say tbh. My first impression is of a slightly "dark" sound that lacks ambient clues. Let me relax more and come back to it. My memory of the original that you posted is that it had more life and sparkle.... maybe that is just expectation and memory playing tricks.
On headphones it sounds quite good... I'll come back to it.
The original track is in this thread
http://www.diyaudio.com/forums/ever...ning-test-math-added-harmonic-distortion.html
(music2)
There is definitely a degradation, because now the original data file goes through CD player, preamp, power amplifier, divider, link trafo, A/D converter. And, the power amp output level is at least 14 dB below full scale, so we loose S/N here. That's how it goes.
When comparing to original, please keep in mind that the sample recorded through audio chain is about -3.5dB below level of the original file, so volume adjustment is necessary.
http://www.diyaudio.com/forums/ever...ning-test-math-added-harmonic-distortion.html
(music2)
There is definitely a degradation, because now the original data file goes through CD player, preamp, power amplifier, divider, link trafo, A/D converter. And, the power amp output level is at least 14 dB below full scale, so we loose S/N here. That's how it goes.
When comparing to original, please keep in mind that the sample recorded through audio chain is about -3.5dB below level of the original file, so volume adjustment is necessary.
Pavel, just tried it out - and still had your music2 file on hand as a reference. Slight difference in volume, but not a problem.
The loss in quality was far less noticeable, if music2 was 10/10, then complete_... was, say, 7/10. Tonality was all in order but there was a slight loss of sparkle, a tinge of greyness added to the sound - as usual, this was all in the treble, some of the life, verve was missing.
But, overall, very good indeed ...
The original seems to resolve to a more pin sharp image. I close my eyes and slightly turn my head to the left and then to the right. The vocal on the original seems to never come from one or other speaker... its placed in its soundstage between the speakers at all times wheras on the complete chain it still has a central image but I'm also "aware" of the sound coming from the speakers to the left and right. There is a bit more zing and life to the original but not as much difference as I imagined.
Very interesting... I won't put the Rotel back... to much hard work... but I could record your original through my system and post it.
Very interesting... I won't put the Rotel back... to much hard work... but I could record your original through my system and post it.
Hmmm interesting idea Mooly, use of Pavels test track to compare systems! Of course it isn't just comparing the playback system but also the recording...
I've not had a chance to listen to any of the posted files.. perhaps tomorrow night
Tony.
Its a bit of an eye opener and a lot of fun. It would be good if some of the many chip amps could be put through something like this. It might be non scientific at this stage but its informative.
I have a 100 watt mosfet amp based on the Hitachi app note (2sk134 2SJ49) and an LM3886 P2P build based on brianGT's schematic. My speakers are also probably on the unforgiving side... maybe if I have the time I can do some recordings.... I would like to work out how to use the differential inputs on my focusrite 2i2 to record it though. I would play back through one PC and record with another (on the laptop running off battery) to avoid any earth loops.
don't hold your breath though! (I still haven't got around to listening to the new sort iether )
Tony.
don't hold your breath though!
(I still haven't got around to listening to the new sort iether ) Tony.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.