But are low sig in stages of ic amps less affected or more affected by the bias control and outputs being close to sig in, in relation to discrete's and there slugish thermal tracking.
Which is better??
Which is better??
"Maybe I should refrase my question (or remark rather). I've noticed many times certain people reference to DS, however, I fail to understand why he would be treated as some sort of authority in (high end) audio?"
Probably has nothing to do with his 20 year + history of publishing tecnical articles in Wiress World, Electronics World, etc.
While I can't say I agree with everything he says, I do consider him to know what he is doing.
I had an amplifier modulate the bias with the signal. I had to re-build it twice before noticing the bias IC had lost contact with the sink.
Probably has nothing to do with his 20 year + history of publishing tecnical articles in Wiress World, Electronics World, etc.
While I can't say I agree with everything he says, I do consider him to know what he is doing.
I had an amplifier modulate the bias with the signal. I had to re-build it twice before noticing the bias IC had lost contact with the sink.
Konnichiwa,
Well, I treat him as an authority in a very narrow field (aka a Specialist), where is work is prodigous and where most people would benefit from looking at it. When it comes to audio in genral I cannot but consider Mr. Self as fairly limited in expertise, as evidenced by some of his commercial designs.
Mr. Self may be an excessively rabid and activist Objectivist, who does not even considers making listening tests and who's definition of audibility is informed by ancient and misleading views (long put to rest in research among other areas into perceptual coding and hearing prothestetics) and he may not be able to design a mixing desk without majorly lousing up the ground bus; but when it comes to evaluating distortion mechanisms in the standard "Lin" Amplifier (meaning the classic LTP/VAS/EF ntopology), a topology I might add that is among the worst possible by far for audio amplification, it's widespread use non-withstanding.
Some of these distortion mechanisms clearly apply to other topologies as well, some others will be easily eliminated by different topologies.
Back to thermology - Ideally a "High End" Chip Amp would be a 2-Chip Split, one driver Chip, containing low level stages linearised as much as possible and an Output Stage Chip containing protection circuitry, biasing and output transistors. Sadly such an approach is not economical.
One might treat an LM3875 very much like a normal "power buffer" Op-Amp and place it into the feedback loop of another Op-Amp with exceptional AC performance (like OPA627?) and perhaps make the Amplifier fully differential ("bridged") to minimise remainig issues, including thermal ones, but at the same time, these issues may very well be involved where the "sound" of the Chip Amp is involved.
Sayonara
Peter Daniel said:
Well, I treat him as an authority in a very narrow field (aka a Specialist), where is work is prodigous and where most people would benefit from looking at it. When it comes to audio in genral I cannot but consider Mr. Self as fairly limited in expertise, as evidenced by some of his commercial designs.
Mr. Self may be an excessively rabid and activist Objectivist, who does not even considers making listening tests and who's definition of audibility is informed by ancient and misleading views (long put to rest in research among other areas into perceptual coding and hearing prothestetics) and he may not be able to design a mixing desk without majorly lousing up the ground bus; but when it comes to evaluating distortion mechanisms in the standard "Lin" Amplifier (meaning the classic LTP/VAS/EF ntopology), a topology I might add that is among the worst possible by far for audio amplification, it's widespread use non-withstanding.
Some of these distortion mechanisms clearly apply to other topologies as well, some others will be easily eliminated by different topologies.
Back to thermology - Ideally a "High End" Chip Amp would be a 2-Chip Split, one driver Chip, containing low level stages linearised as much as possible and an Output Stage Chip containing protection circuitry, biasing and output transistors. Sadly such an approach is not economical.
One might treat an LM3875 very much like a normal "power buffer" Op-Amp and place it into the feedback loop of another Op-Amp with exceptional AC performance (like OPA627?) and perhaps make the Amplifier fully differential ("bridged") to minimise remainig issues, including thermal ones, but at the same time, these issues may very well be involved where the "sound" of the Chip Amp is involved.
Sayonara
"Back to thermology - Ideally a "High End" Chip Amp would be a 2-Chip Split, one driver Chip, containing low level stages linearised as much as possible and an Output Stage Chip containing protection circuitry, biasing and output transistors. Sadly such an approach is not economical. "
25 year old receivers are built that way.
STK 050 (EF), 050II (tripple EF), 070II , 0100II were common output modules. MKII versions had more current gain.
STK3042, 62, 82 were stereo voltage gain stages. STK3156? had a J-fet front end, enough CCS and CM to shake a stick at.
The combination of the STK3156 + STK100II looked like someone had ripped the basic schematic for a Mark Levinson ML3.
25 year old receivers are built that way.
STK 050 (EF), 050II (tripple EF), 070II , 0100II were common output modules. MKII versions had more current gain.
STK3042, 62, 82 were stereo voltage gain stages. STK3156? had a J-fet front end, enough CCS and CM to shake a stick at.
The combination of the STK3156 + STK100II looked like someone had ripped the basic schematic for a Mark Levinson ML3.
If you are talking about mechanical resonance (which could indeed influence electrical parameters) then I would assume that this won't happen at low frequencies. I even doubt that a die would show resonant behaviour in the audible range at all.
I don't want to claim that it won't generate audible signal degradation.
But chances are very small that it would lead to an increase in THD at lower frequencies.
Regards
Charles
djk said:
are you telling me that what suggested as "ideal high-end" chip amp by our "expert" is a 25-year old design that had been proven inadequate in the real world, and based on a topology that by our "expert's" own account is the worst for audio amplification?
That cannot be right, can it?
🙂
Just because it is
1.) 25 Years old and
2.) not commercially attractive
doesn't mean that it is technically inferior.
Regards
Charles
1.) 25 Years old and
2.) not commercially attractive
doesn't mean that it is technically inferior.
Regards
Charles
Originally posted by phase_accurate
If you are talking about mechanical resonance (which could indeed influence electrical parameters) then I would assume that this won't happen at low frequencies. I even doubt that a die would show resonant behaviour in the audible range at all.
I don't want to claim that it won't generate audible signal degradation.
But chances are very small that it would lead to an increase in THD at lower frequencies.
Regards, Charles
I have read of physically bending pcbs to cause changing DC offsets in surface mounted ic's.
Mechanical vibration/resonance of the pcb assembly would cause audio modulation/intermodulation in this case.
Eric.
If you are talking about mechanical resonance (which could indeed influence electrical parameters) then I would assume that this won't happen at low frequencies. I even doubt that a die would show resonant behaviour in the audible range at all.
I don't want to claim that it won't generate audible signal degradation.
But chances are very small that it would lead to an increase in THD at lower frequencies.
Regards, Charles
I have read of physically bending pcbs to cause changing DC offsets in surface mounted ic's.
Mechanical vibration/resonance of the pcb assembly would cause audio modulation/intermodulation in this case.
Eric.
Millwood, your criticisms of the technical content of messages is valid and valuable. Please lose the personal stuff; it's against the forum rules and does nothing to advance your points. I've pulled two of your posts which contain the latter without the former.
SY said:
Millwood, your criticisms of the technical content of messages is valid and valuable. Please lose the personal stuff; it's against the forum rules and does nothing to advance your points. I've pulled two of your posts which contain the latter without the former.
Duly noted.
I wrote that in response to the post from a certain poster so that it was tone-matched, 🙂
If a specific topology is implemented using inferior parts you can't blame the topology, can you ?
I think IC design has improved significantly since the early STK modules.
Regards
Charles
Regarding output transistors and temperature sensing transistors: I happen to design analogue IC's for a living (mainly RF stuff, no audio power amplifiers), and for what it's worth, it is standard practice to put transistors forming a translinear loop or something similar to a translinear loop in close proximity. I would be very surprised if any audio power amplifier IC manufacturers would put the temperature sensing devices at the other end of the die.
mrfeedback said:
Flexing the die a little, huh? Sounds very similar to how a strain gauge transducer works.
- Status
- Not open for further replies.