Joshua_G said:
But people here seem to only quote ONE actual measurement which is hardly the worst case for even a plain triple darlington with diode stack biasing. I admit asking nicely is better. 🙂
And I guess we just have to keep trying till they do.. 🙂
I did what I said in a earlier post and performed the Krill Vs.
EF in a more fair manner. Using a known low distortion
input source (my APT/holman FA2 .001%) and not a
simulated source , I put the 2 "neck to neck".
In the test I had to run the krill with NFB taken from the VAS
(like Steve's original) and I ran the triple at twice the power
with full GNFB.
Gk was right on , the triple , even at twice the power , beats
the krill . The difference in the krill , by its biasing scheme,
seem to be what type of distortion it produces (mainly H2)
and the EF (H3/5). This seems to fit in nicely with what
DX and I have heard.(smoother tubelike sound) Both amps , at least in my simulations,
can easily be made to have very low distortion.
I feel bad , because If I took the NFB off the output, the krill
might match the EF or even beat it.. 😎
Attachments
syn08 said:
I needed to wait a while before replying to this. At first I was angry. After having more time to digest what you have written, I can only say you should hope you never know as much about challenge as I do. If you feel you can tell me about challenge, then you do not deserve my anger. You have my sympathy.
Steve Dunlap said:
Sorry about that. Obviously, you don't owe me any explanation. This is just a virtual community on the Internet, it's not worth getting angry, just ignore my message and happily live ever after.
Bob Cordell said:
I believe I stated that the output stage in question was a modified version of the one I have posted. There is no feedback around the output stage, with the exception of emitter degeneration.
The 400W amps were fully balanced.
Since the output stage does not rely on the VGS for anything but an input signal, any number of VGSs can be used. Pick one you like that can swing about 100V PP. There is no shortage of them.
Do you accept Charles Hansen's claim of no feedback? I haven't seen his schematic so I cannot comment.
syn08 said:
From many posts of yours, my impression is that you try to annoy people intentionally.
Re: Well... i would never have money to buy such kind of amplifiers
🙂
Yes please ... completely agreeddestroyer X said:
🙂
Steve Dunlap said:
Hi Steve,
Yes, but I was interested in the complete specific arrangement of the one you actually built and measured.
I have not seen Charles Hansen's schematics, but I do believe that he uses no feedback other than emitter degeneration. However, I could be mis-understanding his claim.
BTW, one thing about output stage distortion is that it can go all over the place with level and frequency. As you know, crossover distortion can be worst at a variety of different operating levels. Can you post a curve of THD-20 vs. level?
Thanks,
Bob
Steve Dunlap said:
I trust Charles' integrity with such a claim.
I have not seen his schematic either, but he has given a lucid
description of the circuit, and I have no problems understanding it.
Also, the performance is within the bounds of being achievable with
only degeneration.
😎
Bob Cordell said:
I will consider posting that. It will take me a while to draw it up again in my current software.
So, you are willing to accept Charles Hansen's claim, but not mine. May I ask why?
I cannot produce curves of anything. I can not even lift an amp on to my test bench. I know, that is a "convenient" out for me. I really wish I wasn't.
The reason I measured at 20KHz was because another designer told me my measurements at 1K didn't prove anything. If I measured at 20K I would see much higher distortion. I did not. In fact, at all frequencies, at rated power, distortion figures were very low.
I see no way to resolve this if you and others will not accept what I have said. Perhaps someone could come over for a couple of weeks and build one for you (and a couple of others) to test for yourselves. I still have all the parts, except the case to put it in.
ostripper said:
I was looking closely at the two curves you posted... the difference between the 100w Krill and the 200w EF is 3dB, but the scales on the left side of the graphs show a different story. The Krill seems to show a noise floor at -120db while the EF is up at -68, and even taking that into account and drawing a line at the same point in the Krill graph, the peak of the 2nd harmonic is then at what appears to be -16db or so lower??
Am I being confused by something?
These are simulations, not the real thing, or??
_-_-bear
PS. again I may be misinterpreting something, but it looks like the first harmonic that appears is at 30khz? Isn't that not the second harmonic of 10khz?
Steve Dunlap said:
Steve Dunlap said:
I am not trying to be Steve's defender, but just want to draw the attention of those who are perhaps, not aware that Steve is not in a position to test his amps nor draw graphs. The above quotes are from posts 467 and 472 in this thread.
So let us be mindful of Steve's limitations as we seek to push the envelope. Thanks everyone.
Steve wants to cooperate, to see his amplifier constructed and used
this is what makes him happy.... and this is so easy, have only to attract and to give special support to people that will build and not to waste time and energy with people that will simulate, check, inspect, discuss, agree, disagree, say that made first the circuit, that can make it better...that works... that do not works..and nobody have build ....nobody listened... virtual listeners.
But he decided to be answering questions, this attracts theory addicted people...not the real "population" he needs.
this people, detail inspectors, loves to exercise themselves and some loves to show themselves... the last thing they want is to be happy building things.... they want to conclude sonics performance watching simulators.
So, in my point of view he used the wrong strategy.... sitted there alike a consultant he would have people inspecting the amplifier in details, some challenging him, some trying to understand and others trying to say they could do better.
What means the small distortion when the speakers distorts many times more?...so... discussion about small numbers are non sense in my point of view...just academic exercise... reality is another thing.... we listen reality...if small distortion will means nothing when speaker distorts more... this is alike discuss the anjos sexuality when it is possible they have no sexuality and we are not in the heaven to worry about that.
What is the point to reduce distortion from 0.1% to 0.0001% when the speaker will be distorting 4 percent?.
What means bother with 20 Kilohertz?....no one listen this frequency into normal listening levels...you have to be near the airplane jet turbine to listen that frequency... i see...harmonics...but have someone tested cutting high frequencies and comparing...to see if someone can perceive absence of tones higher than 20 Kilohertz into the music spectral distribution?.... maybe they did...
Now was made this way.... i tried, more than once to explain Mr. Steve that possibility, to select "builders"....not theorists...not to stay in that position to answer questions...better to say.... build and listen!
regards,
Carlos
this is what makes him happy.... and this is so easy, have only to attract and to give special support to people that will build and not to waste time and energy with people that will simulate, check, inspect, discuss, agree, disagree, say that made first the circuit, that can make it better...that works... that do not works..and nobody have build ....nobody listened... virtual listeners.
But he decided to be answering questions, this attracts theory addicted people...not the real "population" he needs.
this people, detail inspectors, loves to exercise themselves and some loves to show themselves... the last thing they want is to be happy building things.... they want to conclude sonics performance watching simulators.
So, in my point of view he used the wrong strategy.... sitted there alike a consultant he would have people inspecting the amplifier in details, some challenging him, some trying to understand and others trying to say they could do better.
What means the small distortion when the speakers distorts many times more?...so... discussion about small numbers are non sense in my point of view...just academic exercise... reality is another thing.... we listen reality...if small distortion will means nothing when speaker distorts more... this is alike discuss the anjos sexuality when it is possible they have no sexuality and we are not in the heaven to worry about that.
What is the point to reduce distortion from 0.1% to 0.0001% when the speaker will be distorting 4 percent?.
What means bother with 20 Kilohertz?....no one listen this frequency into normal listening levels...you have to be near the airplane jet turbine to listen that frequency... i see...harmonics...but have someone tested cutting high frequencies and comparing...to see if someone can perceive absence of tones higher than 20 Kilohertz into the music spectral distribution?.... maybe they did...
Now was made this way.... i tried, more than once to explain Mr. Steve that possibility, to select "builders"....not theorists...not to stay in that position to answer questions...better to say.... build and listen!
regards,
Carlos
Distortion figures aside, if Steve's novel output stage manages much lower noise figures, even if only by 10dB and it's likely to be far more than this, then the consequence will be greatly improved resolution which is a very good thing indeed.........
And it is precisely what one would expect from a non-switching output stage.
Bravo, Steve. Thank you for your contribution. Might even build one myself!
Hugh
And it is precisely what one would expect from a non-switching output stage.
Bravo, Steve. Thank you for your contribution. Might even build one myself!
Hugh
V (c) on the krill 15Db to -75db H2= -93db
V (b) +28 and -58db= -86db ,not 16 but a 7 db difference
LT just sums the h2 - h9 , gives a error report showing percent
(.009 -.003%) difference. The point most relevant was not
whether the EF beat the krill , but how bias generator
affects the ratio of distortion components. BTW, something about
GNFB vs. NO nfb affected the noise floor. totally different amps.
this bring me to DX's set of comments...
Having many prototypes around , veroboard, and the simulator
one CAN know : A - exactly what values will give the right
operating conditions (Ic) so you don't end up with a smoking
mess
and get it right the first time.B - A rough idea of how the amp will sound , how to compensate
it,(sure beats swapping out 10 different components) ,and
right away you know what will NOT work , (oscillate).
C. how to improve on a design.. better devices and how they will
affect the circuit and how the design compares to others (to
choose which you should attempt).
Who are you to determine who the real population is ??
By far the most and best of the amps (Leach , symasym
blameless, etc.) started out on the simulator backed
up by theory and numbers, the devices we use were not
guessed at !! , but were the result of long nights of
numbers and calculation.
How a amp behaves at 20 Khz and at what distortion it
can do that at , has a definite affect on how it will perform
in the audio range.
Now that is udder BS... When I started at the forum I just
wanted a simple amp to replace the Walmart garbage I
had. I saw DX amps and quasi amps but chose the quasi
and built them because he (quasi) would answer questions
so we , the builders,would learn why as well as build them.
Is that not the purpose of the forum , to make designers,
not just sheeplike builders ??(to learn). The ESP site is
a prime example of this , rod elliot give both plans to build
and all the theory that goes along with them.
Because of this , I personally don't need someone elses
plans or diagrams anymore. I can just see any idea ,topology,
theory.. simulate it , prototype it , If i like it , make PCB's
and live with it.
OS
- Status
- Not open for further replies.