Yes but, "There IS a C node in my schematic, I have no idea where it went in the one you downloaded. Every number, from THD, to each harmonic, is different between your results and mine." Please take the circuit in #341 and without doing anything sim it, then add node C and compare the .four to the one I posted. Maybe you weren't around for the first round of 50ppm distortion at 25mA of bias.
It sounded pretty good, John? Can you tell us?
My feeling is that regardless if it switches or doesn't switch, if it sounds good, it's a positive step forward. Proving someone right or wrong merely incenses people, but says nothing of the quality of the product...... we play the man a lot on forums. It's a blood sport for the 21st century.
Steve, I still think it's very clever. I don't care if it switches, right or wrong. But if it sounds good, and it sounds different, there is probably a place in the market for the Krill.
My feeling is that regardless if it switches or doesn't switch, if it sounds good, it's a positive step forward. Proving someone right or wrong merely incenses people, but says nothing of the quality of the product...... we play the man a lot on forums. It's a blood sport for the 21st century.
Steve, I still think it's very clever. I don't care if it switches, right or wrong. But if it sounds good, and it sounds different, there is probably a place in the market for the Krill.
How would I scope the emitter current without an isolation transformer for my scope? I really didn't need to anyway - all of this has been played out before, you missed rounds 1-10 I guess.
I respect Steve but he had his 15 minutes and now I think he misses it.
No, apparently the important thing is that you ignore the detailed explanation of what the circuit is actually doing (because you don't like where the explanation is coming from) and just have faith. Have faith.
Or build it and know that it is sublime. Just listen and have faith.
keantoken,
I believe I am using the latest version of LTspice. It was updated no more than 3 weeks ago. I would also like to say that I do not normally use LTspice. I only posted my files in that format because it was insisted on by a person that is no longer on this forum.
Now to address some of the other posts. My schematic, the one I actually use, not the ones with changes other people who are no longer here insisted I must make for the output stage to work "properly" is posted on post 302. If you are not using THIS EXACT SCHEMATIC, you will get somewhat different results.
http://www.diyaudio.com/forums/solid-state/151295-krill-next-generation-16.html#post2019171
A number of people insist on either using a different schematic, or changing things like resistor values or bias levels, post their results, and claim that since they are different from mine, I must be lying. For example:
The second line in that example is a blatant lie. I has been brought up several times before and refuted each time. It is simply an attempt to discredit me by using carefully selected portions of sentences strung together to make it sound like I said something I did not say.
There should be no reason to do that. That is the same file I posted on post 302, but with the added 20 ohm resistor and the change in the bias resistor. If it is different in any other way, I have no idea why.
Learn to read schematics.
This amp has been built. Here is a link to another thread where it's sound is discussed, if that is aloud.
http://www.diyaudio.com/forums/solid-state/159954-krill-complete.html#post2061736
I will no longer entertain the foolishness of those who claim this doesn't work. I'm not going anywhere. I will not be run off by a lack of comprehension on their part. If I no longer reply to their foolish ramblings and outright lies, it is because I am ignoring them, not the rest of you.
I believe I am using the latest version of LTspice. It was updated no more than 3 weeks ago. I would also like to say that I do not normally use LTspice. I only posted my files in that format because it was insisted on by a person that is no longer on this forum.
Now to address some of the other posts. My schematic, the one I actually use, not the ones with changes other people who are no longer here insisted I must make for the output stage to work "properly" is posted on post 302. If you are not using THIS EXACT SCHEMATIC, you will get somewhat different results.
http://www.diyaudio.com/forums/solid-state/151295-krill-next-generation-16.html#post2019171
A number of people insist on either using a different schematic, or changing things like resistor values or bias levels, post their results, and claim that since they are different from mine, I must be lying. For example:
The second line in that example is a blatant lie. I has been brought up several times before and refuted each time. It is simply an attempt to discredit me by using carefully selected portions of sentences strung together to make it sound like I said something I did not say.
There should be no reason to do that. That is the same file I posted on post 302, but with the added 20 ohm resistor and the change in the bias resistor. If it is different in any other way, I have no idea why.
Learn to read schematics.
This amp has been built. Here is a link to another thread where it's sound is discussed, if that is aloud.
http://www.diyaudio.com/forums/solid-state/159954-krill-complete.html#post2061736
I will no longer entertain the foolishness of those who claim this doesn't work. I'm not going anywhere. I will not be run off by a lack of comprehension on their part. If I no longer reply to their foolish ramblings and outright lies, it is because I am ignoring them, not the rest of you.
This deserves to be repeated.
Krills have been built.
They work.
They sound good.
IIRC, this all started when Carlos heard a krill amp, was stunned by it's performance, and wanted to learn more about it - "how does it work?", "why does it sound so good?", "where can I get one?" etc.
The designer (Steve) was good enough to provide schematics of his amps and explain them as best he could.
Since then arguments have flared regarding sims and Steve has been called a liar (if not worse).
I find this behavior absolutely disgusting.
While there might reasonably be discussion about how it works or why it sounds good, there can be no doubt that it does so.
Sadly, those who live by the numbers and confuse simulations with reality have not only learned nothing themselves, but have prevented the rest of us from learning.
Those who wish to build boring amps with triple-ef output stages because "everyone else is doing it" or because they "know" it's "the best" are, of course welcome to do so but I, for one would really appreciate if they would (or could) abstain from disrupting discussion of anything different or unusual.
Cheers - Godfrey
Look closer.
To add to what godfrey said here:
I also offered to give, free of charge, all the parts to build two channels of the 100W or 50W amp to anyone that would come by and pick them up. The parts were picked up by c2cThomas who arranged to send them to Carlos. Many people here have been very kind and helpful. I do appreciate that.
I also offered to give, free of charge, all the parts to build two channels of the 100W or 50W amp to anyone that would come by and pick them up. The parts were picked up by c2cThomas who arranged to send them to Carlos. Many people here have been very kind and helpful. I do appreciate that.
Its voltage is set to -34V (minus thirty-four volts) so there is no problem there. However, the file from post 341 doesn't have a "C" node, just as Scott said. I get the error message too when trying to simulate it. Could the label have been deleted by mistake when adding the resistor perhaps?
For the file from post 341, with the output renamed to "C" and the frequency changed to 20kHz I get:
Code:
Circuit: * C:\joakim\spice\krill\Krill20\Krill20.asc
Direct Newton iteration for .op point succeeded.
Fourier components of V(c)
DC component:0.00348593
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 2.000e+04 2.875e+01 1.000e+00 -0.02° 0.00°
2 4.000e+04 4.201e-03 1.461e-04 -84.23° -84.21°
3 6.000e+04 9.634e-03 3.350e-04 4.27° 4.29°
4 8.000e+04 2.170e-04 7.546e-06 -103.50° -103.48°
5 1.000e+05 8.697e-04 3.025e-05 -11.36° -11.33°
6 1.200e+05 4.476e-05 1.557e-06 75.58° 75.61°
7 1.400e+05 1.091e-04 3.794e-06 33.54° 33.56°
8 1.600e+05 4.157e-05 1.446e-06 100.64° 100.66°
9 1.800e+05 5.227e-05 1.818e-06 -167.91° -167.89°
Total Harmonic Distortion: 0.036687%For the unchanged file from post 302 I get:
Code:
Fourier components of V(c)
DC component:-0.00994689
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 2.000e+01 2.829e+01 1.000e+00 -0.00° 0.00°
2 4.000e+01 8.461e-03 2.991e-04 -90.16° -90.16°
3 6.000e+01 6.187e-03 2.187e-04 -0.18° -0.18°
4 8.000e+01 6.272e-05 2.217e-06 87.42° 87.42°
5 1.000e+02 2.729e-04 9.646e-06 0.03° 0.03°
6 1.200e+02 3.728e-05 1.318e-06 -89.88° -89.88°
7 1.400e+02 1.008e-04 3.564e-06 0.03° 0.03°
8 1.600e+02 7.917e-06 2.799e-07 89.43° 89.43°
9 1.800e+02 6.475e-06 2.289e-07 0.87° 0.87°
Total Harmonic Distortion: 0.037071%However, for the same devices rearranged into a triple EF (KSC3503/KSA1381, 2SC4793/2SA1837 and 4 pairs of Q2SC5200/Q2SA1943), and only 100mA (half of what is used in the post 302 amplifier) of bias per pair, I get less than a third of these distortion numbers:
Code:
Fourier components of V(c)
DC component:-0.000219481
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 2.876e+01 1.000e+00 -0.00° 0.00°
2 2.000e+03 2.723e-03 9.467e-05 -88.97° -88.96°
3 3.000e+03 3.303e-04 1.148e-05 8.74° 8.74°
4 4.000e+03 4.723e-04 1.642e-05 -88.91° -88.90°
5 5.000e+03 1.531e-04 5.321e-06 -3.55° -3.55°
6 6.000e+03 1.681e-04 5.844e-06 -88.33° -88.33°
7 7.000e+03 3.353e-04 1.166e-05 0.41° 0.41°
8 8.000e+03 6.110e-05 2.124e-06 -88.68° -88.67°
9 9.000e+03 2.128e-04 7.399e-06 -0.18° -0.18°
Total Harmonic Distortion: 0.009809%The 8 ohm load is quite light for 4 pairs of output transistors in an amplifier powered by +-34V... But transistors are cheap nowadays, so using lots of output transistors isn't such a bad idea for lowering distortion.
Last edited:
BUT,
Does lower simulated distortion always and inevitably correspond to better sound?
THAT is the question, not whether Steve is a liar or whatever...... I think much more is going on in a power amp than merely the simulated numbers. Sure, they are helpful, and give some indication of where weak points are, but THD in particular is a unilateral measure of a multifaceted issue. To me, H2 and H3 are almost unimportant, I look at H5, H7, and H9 - they are the critical harmonics, and the THD gives these artefacts no more importance than H2, which is crazy.
After all, tubes amp have high THD, and they sound pretty good. In fact, most of their distortion is H2, H3 and H4, all relatively benign compared to high, odd order.
Hugh
Does lower simulated distortion always and inevitably correspond to better sound?
THAT is the question, not whether Steve is a liar or whatever...... I think much more is going on in a power amp than merely the simulated numbers. Sure, they are helpful, and give some indication of where weak points are, but THD in particular is a unilateral measure of a multifaceted issue. To me, H2 and H3 are almost unimportant, I look at H5, H7, and H9 - they are the critical harmonics, and the THD gives these artefacts no more importance than H2, which is crazy.
After all, tubes amp have high THD, and they sound pretty good. In fact, most of their distortion is H2, H3 and H4, all relatively benign compared to high, odd order.
Hugh
I don't think anyone has claimed that the Krill would sound bad. What was said was that Scott's simulation results were off. I posted my results to show that Scott's numbers from post #354 could be reproduced.
BTW, by increasing the triple EF bias to the same amount as in the Krill, the 7th harmonic goes down and the THD goes up. Like that, the numbers look just like the Krill.
I was going some where with this amp before all the shouting started. There is a reason for four pair of outputs.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.