Mr Self has used an AP system at some real circuit so I don't doubt his results. I admit that it had been uplifting to see some pictures and other valuable info. But as suggested, best thing is to get in contact with the author then you won't have to speculate.mikek said:
peranders said:
No speculation involved...the circuit is given, and distortion curve ascribed to it.....If the results, (in simulation, and with a prototype), are not reproducible by whomsoever...where ever....whenever, then they may be considered invalid...that is one of the oldest sceintific principals around.....
Try it for yourself.....thats what it's all about after all......why take Self's word, or indeed, mine or anybody else's word for anything?
Simulate...and BUILD the circuit for yourself,.....to determine whether or not it makes sense......
weeghel said:
He never seems to respond to e-mail..
Anyway, i just figured perhaps those with experiance of his work may find this discussion fruitfull....
PMA said:
yes...but this would double open loop gain.....and require emitter degeneration, or a reduction in diff. stage bias current to make the whole stable ...the resistors would increase noise...making the whole point of a discrete op-amp. invalid...
I think folks are better off trying the JE-990 here:
http://www.psyber.com/~dibsed/CIRCATS/je-990.pdf
...Reproducible...tried and tested...
Apart from the feedback, I suppose 🙂
Self has been writing stuff, that whilst not entirely untrue, misses the point by such a huge amount for years, he only gets the credit he does due to the prolific nature of his public output.
Try building any of his so-called 'blameless' designs and tell me what you hear.
I didn't like the results, anyway.
He's never covered some seriously fundamental issues w.r.t. push-pull amps and this prevents any of his designs from working optimally.
Andy.
ALW said:
His designs as published are not entirely 'blameless'.....for instance the description 'load-invariant' he ascribed to one of his designs is hardly accurate...🙁 ......
ALW said:
what seriously fundamental issues?
To Mike K
the similar topology was used by Self in Advanced preamp in the late 70th, the unit have less then 20 ppm. There MPSA06/56 were used
Jensen opamp and Renardson opamp have very similar topology and the same order disto. I found the disto figures produced by various CADs as the most unreliable
To Kuei Yang Wang
Please could you mention, for which company he did his mixer design?
To Andy LW
During last decades I made a lot of simulations and measurements, and the results (at least shape of curves) looks like Self data. I also made several Advanced preamp and MC Head preamp for my friends without any problems.
Please could you specify what troubles have you faced with his design(Blameless amp)? What are “some seriously fundamental issues”? May be I also miss something?
Thank you
the similar topology was used by Self in Advanced preamp in the late 70th, the unit have less then 20 ppm. There MPSA06/56 were used
Jensen opamp and Renardson opamp have very similar topology and the same order disto. I found the disto figures produced by various CADs as the most unreliable
To Kuei Yang Wang
Please could you mention, for which company he did his mixer design?
To Andy LW
During last decades I made a lot of simulations and measurements, and the results (at least shape of curves) looks like Self data. I also made several Advanced preamp and MC Head preamp for my friends without any problems.
Please could you specify what troubles have you faced with his design(Blameless amp)? What are “some seriously fundamental issues”? May be I also miss something?
Thank you
dimitri said:
Nothing to do with CAD per se.....i found the the design as shown on his web-site to be unstable....in simulation....calculation.....and prototype......
I have made simulation for Fig 3 An improved discrete opamp
http://www.dself.dsl.pipex.com/ampins/discrete/discrop.htm
with MPSA06/56 and found the same shape frequency response as Mike K got, but the peak is +12dB instead of +30dB(from Mike K). The circuit was stable.
http://www.dself.dsl.pipex.com/ampins/discrete/discrop.htm
with MPSA06/56 and found the same shape frequency response as Mike K got, but the peak is +12dB instead of +30dB(from Mike K). The circuit was stable.
Mike,
I have to agree with both dimitri and sonnya above, the design appears stable, with about 35 degrees of phase margin, in my simulations. Not a brilliant phase margin and likely to have some variation in real-world implementations, but nonetheless, it's potentially stable.
For someone purporting to design commercially available amplifiers the statement above though astonishes me. Surely you didn't mean that when referring to stability?
Your semi choice is a significant part of the problem.
So credit where due, on this occasion Mr Self seems more capable of getting the basics right. As for the other issues, I'm not commenting publicly, sorry.
Andy.
I have to agree with both dimitri and sonnya above, the design appears stable, with about 35 degrees of phase margin, in my simulations. Not a brilliant phase margin and likely to have some variation in real-world implementations, but nonetheless, it's potentially stable.
For someone purporting to design commercially available amplifiers the statement above though astonishes me. Surely you didn't mean that when referring to stability?
Your semi choice is a significant part of the problem.
So credit where due, on this occasion Mr Self seems more capable of getting the basics right. As for the other issues, I'm not commenting publicly, sorry.
Andy.
If I interpret this thread correctly, Self's circuit is the subject of
severe criticism of what sounds like a very unusual effect.
Unusual enough that there is probably a mistake somewhere.
I think it's very appropriate at this time to elicit some response
from Self, who is not a dummy, and is certainly capable of
defending his designs. If Doug doesn't offer up a response,
maybe somebody else (there are several of you) can offer up
some additional info.
severe criticism of what sounds like a very unusual effect.
Unusual enough that there is probably a mistake somewhere.
I think it's very appropriate at this time to elicit some response
from Self, who is not a dummy, and is certainly capable of
defending his designs. If Doug doesn't offer up a response,
maybe somebody else (there are several of you) can offer up
some additional info.
I do not know if it is clear, but i have nothing against Self. He have maked some wonderfull books. Covering a lot of subjects in amp design. Not in the depths but enough to give you some valuable info.
Also on stability, he have maked a small chapter where he takes this subject up to debate.. He also states that he will not go into depth with all the subjects, because of all the brilliant books covering the subjects..
So he is a wiseguy.....
Also note, he has made measurements with a AP system. So it is not fantasy ...
Regardless of what other people says, Choise of transistor or fet has a great impact on phase margin, distortion, noise and bandwidth.
The models in spice are not perfect, they show less gain linearity than real life regarding "hfe" or "gfs".
Distortion measurement have to be done real life. Also note that caps can distort a lot, and mask the opamp's performance a lot...
Go find wireless/electronics world over the last year, where there have been a article on capacitor distortion... worth reading.....
Also on stability, he have maked a small chapter where he takes this subject up to debate.. He also states that he will not go into depth with all the subjects, because of all the brilliant books covering the subjects..
So he is a wiseguy.....
Also note, he has made measurements with a AP system. So it is not fantasy ...
Regardless of what other people says, Choise of transistor or fet has a great impact on phase margin, distortion, noise and bandwidth.
The models in spice are not perfect, they show less gain linearity than real life regarding "hfe" or "gfs".
Distortion measurement have to be done real life. Also note that caps can distort a lot, and mask the opamp's performance a lot...
Go find wireless/electronics world over the last year, where there have been a article on capacitor distortion... worth reading.....
sonnya said:
True!🙂
sonnya said:
True...
sonnya said:
True.....but not if the circuit breaks intermittently into oscillation....However, THD in simulation that gives WORSE results than those ascribed to a circuit in prototype, constitutes grounds for the gravest suspicion...methinks...
I used MPSA06/56 models used by Dimitri, and can confirm the 12db peak...however the gain margin is just 8 degrees.
In my experiance Zetex's models tend to be more realistic than on-semi's........they were the first to introduce BSIM3 models for their devices.
Moreover my prototype of Self's circuit bursts joyously into oscillation at every opportunity..
.......merely confirms the theoratical, and simulated results i have obtained.... 🙁
- Status
- Not open for further replies.