Amplifier this filter works with voltages less than 2 volts. This amplifier assembled from 2 opamp, they are covered by overall NFB (we call it "composite"). Therefore, this amplifier has low distortion.
You can go to this forum: Audio Perfection Forum and ask the person Audio Perfection Forum. He is the author of this filter and he lives in the United States.
I don't think the level of distortion for 1 kHz is a good indicator of the quality of the amplifier. Therefore, I did not measure distortion for this frequency - for the past 15 years.
PS: And I have even less brains than you. 🙂
Last edited:
I use the calculation TKHD 20K during the simulation of the amp models for quick evaluation of changes of the circuit. Because sound cards have a narrow band of frequencies, while measurement method is used with CCIF IM 19+20kHz.
IMHO you can't use IMD figures to calculate THD20k and vice versa.
http://www.diyaudio.com/forums/soli...s-ab-power-amp-200w8r-400w4r.html#post3945064
In the above example the measured THD 20k was about 0.001% @100W8R while the measured IMD was about 0.0005% @ 100W8R.
I and many others would really like to know how you (T117) measure such low THD values (sub ppm) with a simple sound card...
BR, Toni
http://www.diyaudio.com/forums/soli...s-ab-power-amp-200w8r-400w4r.html#post3945064
In the above example the measured THD 20k was about 0.001% @100W8R while the measured IMD was about 0.0005% @ 100W8R.
I and many others would really like to know how you (T117) measure such low THD values (sub ppm) with a simple sound card...
BR, Toni
All I can see is that your filter suppresses the 20kHz and also the harmonics above 20kHz ...
😱
Yes, I see that also, Dick Moore very easily obtained 60 dB nulls with only .5 dB attenuation of the 2nd harmonic. Working hard he says he was able to get 100 dB nulls but could not hold them for very long. Dick used good modern OP amps, no need for a composite.
Could you please post the LTspice *.ASC files for one of your <0.000,3% THD20k amps?
I'm sure many of us would like to have a play with such good circuits and this would be a great convenience as well as saving a lot of time getting it to work 🙂
If you have the SPICE models for the devices, we would be even more grateful.
Last edited:
This is what I'm hoping to persuade Bob to do.
I did this circa 1980 but it was very hard work for my small brain. I no longer can do this 'rithmetic & algebra 🙁
You have to make all sorts of assumptions about Loop Gain Response etc but it's possible.
Last edited:
if a "error" is measured by the negative feedback diff input...
then its reduced by the excess loop gain at the frequency of the error
IMD product errors are reduced by the excess loop gain at the frequency of the IMD product
for difference frequencies like the 1 kHz 2nd order diff of the popular 19, 20 kHz 1:1 test it is the excess loop gain at 1 kHz that matters
so higher order loop like TPC coupled with high gain in the small signal forward path break simple single dominant pole based calculations in a good way
and make "flat loop gain over audio" look very dubious when "musical" audio frequency distribution, human hearing acutity for errors vs frequency are considered
then its reduced by the excess loop gain at the frequency of the error
IMD product errors are reduced by the excess loop gain at the frequency of the IMD product
for difference frequencies like the 1 kHz 2nd order diff of the popular 19, 20 kHz 1:1 test it is the excess loop gain at 1 kHz that matters
so higher order loop like TPC coupled with high gain in the small signal forward path break simple single dominant pole based calculations in a good way
and make "flat loop gain over audio" look very dubious when "musical" audio frequency distribution, human hearing acutity for errors vs frequency are considered
You can try to play with this one, THD20k at 100W/8=0.000048% or 0.48ppm.
Just unzip and play, models included.
Damir
Yes, of course. This suppression must be considered.
I was interested in the coincidence of the results of simulations and measurements. The results of the simulation were confirmed by several harmonics.
I can't hear the second harmonic of the signal of 20 kHz and above. So I don't see much sense to measure harmonics above.
I do not hide these models. 🙂 I opened several topics, you find there a lot of the models. On the basis of these models were built good amplifiers.
For example:
Attachments
Last edited:
This model amplifier, with which I began. This scheme is as simple and just as effective.
2 ppm @ 20k @150 W @ 8 Ohm according to the simulation, 3 ppm according to measurements. 🙂
2 ppm @ 20k @150 W @ 8 Ohm according to the simulation, 3 ppm according to measurements. 🙂
Last edited:
Thx! I try to ask again: how do you measure the 3ppm THD20k@150W8R? With the filter shown be4?
H2 has a delta of -102dB
H3 has a delta of -85dB
IMHO this is far away from 3ppm ... it smells like >50 ppm.
Maybe I'm completely wrong - maybe one can do a calculation?