Our concerns about numbers, harmonics, distortion, square waves, THD, measurements

Status
Not open for further replies.
Jump to Latest
#83
I forgot to mention, for your 1v OP {sqrt(X)} change "X"
to a lower value.
And if you want to really drop the noise floor , do
.param num_fft_pts 65535 or any multiple (sloooow)

View/ "spice error log" will also give a summary of
the first 9 harmonics and total THD. (attached)😎
To "mess around" , throw your differential way out of balance
and see the H3 "soar" 😀

OS
 

Attachments

  • thd.gif
    thd.gif
    21.5 KB · Views: 739
#84
OS,

This is a very clever program, thank you for the tips. I don't understand any of it yet, but I'll get there in a few years, I'm sure.

Lumba,

Thank you for your comments, but why you say that bemuses me. There are moments when I think I would have preferred being a Columbian drug lord....... so much more exciting and profitable!

Hugh
 
#87
I wonder if there is some problems today simply because of history. For example, the manufacturers of cars tell us how many horsepower in their engines. It's a measurement, but too simplistic. Just like THD. For a car the power is quoted only at the peak, where nobody sensibly drives and it tells you only the maximum speed nothing about how the car drives and how powerful it feels (for this you need the torque curve as a function of rpm). THD may have been useful when amplifiers were all pretty awful as far as distortion was concerned and so as a simple number than Joe Public could get a handle on it allowed manufacturers to publish something scientific for comparison purposes. Now we know the truth that the sound is a complex function of the amplifier (it's environment and other things too) the THD number is just as useless as maximum horsepower in most cars.

I really like the idea of trying to correlate subjective listening experiences with something that can be measured even if I think it's only possible to do this imperfectly. It would be nice to know what to look for if I want an amplifier that has a certain sound. The comments that Carlos and others make when comparing amplifiers (e.g. A is so much sweeter at the hi end and B much firmer at the low end) tells me much more than THD.

The next step would be to know what do I have to tweak in the design or operating points. Perhaps I can then build one amplifier instead of being unable to choose who's design to follow. And I add some adjustments to the front panel which I can use to change the acoustics of the amplifier to suit my mood. One day I want to relax with some particular music and have a certain sonic signature from the amplifier yet another day I want something else. OR maybe this should all happen in the pre-amp.

I ramble.
 
#88
And I add some adjustments to the front panel which I can use to change the acoustics of the amplifier to suit my mood. One day I want to relax with some particular music and have a certain sonic signature from the amplifier yet another day I want something else. OR maybe this should all happen in the pre-amp.
Click to expand...

That is a grand idea, I have thought about switching a diode
out of the Vbe at low volumes (overbiasing) keeping in
class A , then as the amp thermally undercompensates (overheats)with increased volume :hot: :hot: , switch the last diode back in, allowing for Stable class B operation.

As far as a different "character" for each genre , best leave that
for the EQ or preamp.
OS
 
#89
Hello everyone

Wen I test an amp or audio circuit, I use the lowest level that my
FFT test can correctely use, just few db over the noise floor wen it is possible. And I look to see if there is odd distortions harmonics or high frequencies distortions harmonics and their respective level compared to the fundamental. And for more realistics tests results, those measurements should be done with a real loudspeaker load connected to the amp under test. But without listening tests all thoses measurements are a bit useless.

And for a listening test, I first listen at normal level for the definition, the soundstage, etc.., and after that I do lower the output to only 100 to 200 mw, and if the definition, the soundstage, etc.. are still at the same qualities at this very very low level output, it's a very good amp who really deserved to by named high-end amp.

Nelson Pass and Graham Maynard did some talk about those first watts output of an amp, these first watts are crucial, and you also need an amp NFB who won't go positive at pole frequency because of the load. That is why we need good compensations. There is also the driver transistors switching noises, but the right cap a the right place and you get better results.

Bye

Gaetan
 
#91
Re: Re: Back in the 80's

Bob Cordell said:
The MOSFET amp you bought surely did not have 0.001% THD at 20 kHz. The spec you are referring to is almost certainly 0.001% at 1 kHz (assuming you recall it correctly). This very different, and much easier to achieve. It sounds like you have possibly fallen into the trap of 1 kHz THD specs.

Cheers,
Bob
Click to expand...

I do not recall it being a MOSFET amp:scratch: I do not recall it listed at .001% at 20kHz either. I do not recall falling in a trap. I guess I don't remember any of those things...weird:xeye:

Any who, I'm just saying I don't pay much attention to specs like I once did. I think I can easily handle 1% THD without much bother. I would love to spend time studying the different harmonics in distortion, though. I envy guys like you Bob, who have dedicated so much effort and time to such things.

I love reading this stuff!
 
#92
Lumba Ogir said:
My understanding says that the value of simulation in that regard is zero, because the conjugate variables, position and momentum of a moving particle cannot be determined simultaneously. In other words, spectral components are not traceable at any given time instant.
Click to expand...

FFT analysis in SPICE gives the amplitude and phase of each individual harmonic (IOW, a full spectral analysis). It's not just a THD thing. You can do IM with it too - it's just a bit more effort. If you have questions about what it can and cannot do, just ask.
 
#93
Very good Andy has done a good job in this forum with simulators

In the graph fig. appears as the measurement IMD , two different frequencies are applied to input, 60Hz and 7Khz (Standard default SMTPE) the residual near the fundamental 7KHz is the product IMD (highlighted yellow in Figure).
 

Attachments

  • imd.gif
    imd.gif
    47.9 KB · Views: 742
#94
Re: Re: Re: Back in the 80's

mosfets said:
I do not recall it being a MOSFET amp:scratch: I do not recall it listed at .001% at 20kHz either. I do not recall falling in a trap. I guess I don't remember any of those things...weird:xeye:
Click to expand...

Given the time you bought it, and the description of how it sounded, I'd bet that it was indeed a MOSFET amp. I have one of those as well. It's sitting out back, and at least being useful as a home for field mice. The sound was gawdawful.

Seeing that no OPT is cheaper than the cheapest OPT, designers of these abominations simply adapted the SEPP-type output stage that works quite well with BJTs to MOSFETs. It's not a good idea, being that "complimentary" N-Channel / P-Channel devices are a good deal less "complimentary" than NPN / PNP pairs. For MOSFETs, use one type or the other, but not both in the same circuit.

Connect them to an OPT or use a Circlotron-type OTL final. Also, keep 'em well away from Class B operation. These thingies have hideous cross-over behaviour.
 
Status
Not open for further replies.
Top Bottom