I was looking at the latest hunk of hogged from solid stereophile has reviewed and noted IMD was some 20dB worse than the humble LM3886. X-ref the JC-1 , which is about the same (so still 20dB worse than the chipamp). But both are around 30dB better than the minimum in Dave's presentation, so hopefully good enough.
Shame AP can't lend stereophile the latest tester so they can do multitone tests as would be interested to see how they perform on that.
Shame AP can't lend stereophile the latest tester so they can do multitone tests as would be interested to see how they perform on that.
Perhaps this is a really dumb question stemmed from looking at problem in the wrong way, but won't the local supply bypass (e.g. energy storage/capacitance sitting right on the output transistors) greatly reduce the bandwidth of this problem, as seen by the remainder of the power supply/wiring?
What? I don't think so. I have an AP that can do multi-tones. I suggested it a long time ago here. AP suggests it also instead of single tone sine wave tests.
You should try it.
Stereophile should do it also. They can rent one or buy one.
-RNM
Last edited:
Depends on the physical layout (ground plane if you do a output transistor pcb, I suppose), but essentially, yes. And that becomes the take-off for the feedback as well.
Yes. ?? Unsighted, Bob Jones will tell you when you just turned on the ultra sonic alarm and when you turned it off. Karen Richardson, too.
??
-RNM
Last edited:
Yes you did
I would argue being able to measure distortion whilst playing actual music into actual speakers would be a BIG step forward in what we can find out about an amplifier.
I've posted the multitone measurement for the power amplifier I am using many times. You have the AP, where is the plot for Damir's? I expect it will be superbly good, but not seen it.
Would something along the lines of this have any potential? ESP SIM (Sound Impairment Monitor)
There is a residual error unaccounted for if you do this with a CFA. You know, the feedback current that does not exist.
Making a bridge out of the input and the output divided by the gain (with phase compensation) gives a better accounting for more cases.
Two things,
using cymbals to illustrate the problem has a problem in propagation. A cymbal mostly represents a point source of acoustic energy. Very few HF or UHF speaker sources can begin to approximate a point source within a room. The power spectra is not even close and will dramatically change the perception within that space. ie, not a valid test.
As was noted (I believe Scott) condensor mics have large resonances in the top of their range. (HF) To move away from this in condensor mics one has to get smaller so as to move the inherent resonance up in freq, 1/2", 1/4" or even 1/8th inch. They get progressively noisier and flatter higher up, but still have damped resonances in the HF passband. (needed to make it usable/acceptable) To me this colors the sound a fair amount, this seems to be desirable to many producers, witness the preponderance of large diaphragm condensors.
The mics that do not have this upper resonance are ribbon mics and to a lesser extent dynamics(they have many other problems). Ribbon mics tend to have their inherent resonance at low frequencies (say 5- 20 Hz). Dynamic mics are a more complex bag and have many things going on, making them less interesting a source vs the less "complex" condensor or ribbon mic.
Just a couple comments to limit the choices of off the cuff testing.
Cheers
Alan
using cymbals to illustrate the problem has a problem in propagation. A cymbal mostly represents a point source of acoustic energy. Very few HF or UHF speaker sources can begin to approximate a point source within a room. The power spectra is not even close and will dramatically change the perception within that space. ie, not a valid test.
As was noted (I believe Scott) condensor mics have large resonances in the top of their range. (HF) To move away from this in condensor mics one has to get smaller so as to move the inherent resonance up in freq, 1/2", 1/4" or even 1/8th inch. They get progressively noisier and flatter higher up, but still have damped resonances in the HF passband. (needed to make it usable/acceptable) To me this colors the sound a fair amount, this seems to be desirable to many producers, witness the preponderance of large diaphragm condensors.
The mics that do not have this upper resonance are ribbon mics and to a lesser extent dynamics(they have many other problems). Ribbon mics tend to have their inherent resonance at low frequencies (say 5- 20 Hz). Dynamic mics are a more complex bag and have many things going on, making them less interesting a source vs the less "complex" condensor or ribbon mic.
Just a couple comments to limit the choices of off the cuff testing.
Cheers
Alan
And HF loss to air meaning what the drummer hears on his/her cymbals is a far cry what the audience hears (at least in an unamplified venue; amplified, we run into the very thing Alan discusses)
This problem only gets worse with frequency, too.
Also someone mentioned needing to shut off ultrasonic alarms/etc: a lot of times those aren't even ultrasonic, but rather in the 15k range (again, HF loss). The "ultrasonic" rodent repellents (of dubious utility) that family friends have at their beach cottage make me nauseous, albeit no one else in my family seems bothered by them. Whatever sound they're putting out was definitely audible, and also probably in the 15k range.
This problem only gets worse with frequency, too.
Also someone mentioned needing to shut off ultrasonic alarms/etc: a lot of times those aren't even ultrasonic, but rather in the 15k range (again, HF loss). The "ultrasonic" rodent repellents (of dubious utility) that family friends have at their beach cottage make me nauseous, albeit no one else in my family seems bothered by them. Whatever sound they're putting out was definitely audible, and also probably in the 15k range.
Last edited:
May I ask what is the purpose of injecting a fraction of Amp-out past the nulling stage?
Also how phase adjustment is done when the test signal is not of a single frequency (say a dual tone 60Hz and 7kHz)?
George
Attachments
Thanks Daniel for answering.
I had a very simple distortion magnifier from some decades back to use with an oscilloscope (2 transistor long tail pair, single ended amp, resistors & capacitors, Wireless World, Circuit Ideas).
Now, when using a sound card to record the input and output signal, everything can be done in SW through a wave editor.
Phase invert one channel (if needed), synchronize (thus the question for phase adjust), equalize amplitudes, save as mono, sw amplify. It is around a 10 minutes work.
George
I had a very simple distortion magnifier from some decades back to use with an oscilloscope (2 transistor long tail pair, single ended amp, resistors & capacitors, Wireless World, Circuit Ideas).
Now, when using a sound card to record the input and output signal, everything can be done in SW through a wave editor.
Phase invert one channel (if needed), synchronize (thus the question for phase adjust), equalize amplitudes, save as mono, sw amplify. It is around a 10 minutes work.
George
- Status
- Not open for further replies.