i took Fischer's class at 190 E. Delaware -- before UC had a real facility downtown -- I asked him what had caused the depression and he said RCA had produced too many radios! the following year Fischer and Myron taught a course together.
now I apply options to junk, still read JPE and JOF.
We are getting off track and I would hate to see this thread devolve into "Everything Else" by the moderator as I anxiously await to see who is the winner !
now I apply options to junk, still read JPE and JOF.
We are getting off track and I would hate to see this thread devolve into "Everything Else" by the moderator as I anxiously await to see who is the winner !
Hint? Possibly but maybe not......
________ low ____ value for ________ audio ________
This a guess as I don't know and I would have said PSRR. I think that has been guessed and that was not it as I understand. Andy hasn't told me the answer either!
________ low ____ value for ________ audio ________
This a guess as I don't know and I would have said PSRR. I think that has been guessed and that was not it as I understand. Andy hasn't told me the answer either!
Peter Daniel said:
Peter, that was to be expected with ANY cap. You changed the xover network, meaning the freq response etc of the speaker changed. It would be surprising if you didn't hear a difference.
Jan Didden
The mica cap was 0.01uF in parallel with 3uF cap😉
BTW, 3u cap is the only cap in a speaker's crossover.
BTW, 3u cap is the only cap in a speaker's crossover.
I like the power supplies in Spice a lot. Utterly-zero output impedance, perfectly linear.
Failing this:
-output impedance fairly low
-output impedance resistive over the frequency range of interest and beyond for safety
-linear, i.e. low distortion: imagine what a non-linear power source would do to a circuit that can be modelled as a current source with zero PSRR, carrying a wide-spectrum signal.
Failing this:
-output impedance fairly low
-output impedance resistive over the frequency range of interest and beyond for safety
-linear, i.e. low distortion: imagine what a non-linear power source would do to a circuit that can be modelled as a current source with zero PSRR, carrying a wide-spectrum signal.
Werner,
Pick one of those and you will find yourself with a prize.
1 out of 3, good odds 😉
Andy.
Pick one of those and you will find yourself with a prize.
1 out of 3, good odds 😉
Andy.
ALW said:
Drop magnitude of impedance. Feeding a sympathetic amplifier circuit you still get away even with 1R or so.
Two can be relaxed somewhat. Having reactance in the range of interest (and above and below for safety) can be OK, provided the total impedance isn't resonant.
Leaves us with linearity, then. Which sort of explains the reputation of
1) painstakingly-developed complex regulators (hi Jonathan!)
2) passive LCRC kind of things.
Ladeeeees and Gentlemen
We have ourselves a winner!
Werner,
Drop me an email (you know where to find me) and I'll sort something out for you.
Should keep the rest of you busy for a while 😉
Andy.
We have ourselves a winner!
Werner,
Drop me an email (you know where to find me) and I'll sort something out for you.
Should keep the rest of you busy for a while 😉
Andy.
Thanks, though I don't necessarily agree on this being the one and only Big Issue,
at any time, with any load/circuit.
But I forgot to mention the more imporant thing, which is that I took finance classes from a guy who claimed to be a mate of Modigliani's.
at any time, with any load/circuit.
But I forgot to mention the more imporant thing, which is that I took finance classes from a guy who claimed to be a mate of Modigliani's.
I'm certain many will disagree too!
The balance is partly to do with how the other parameters affect this one though, within the context here and assuming basically sensible figures elsewhere (e.g. a PSU with 10k o/p impedance isn't a very good PSU!), it will bring improvements of extraordinary magnitude.
IMVHO 😉
Andy.
Werner said:
I am positively surprised, 🙂 I would have never expected this many finance people on such a forum. first, you got a JOF-reading DIYer, then a M&M theoreom specialist. What's next? a micro-market structure guy? or a dividend signaling advocate?
Diversity is good, tho.
It still is about power supplies
Not all that glitters is gold.
Me just a lowly EE who hung around the economics faculty and the only named theorem they ever taught me there was M&M (BTW long since forgotten) and the only finance problem still relevant to me now is whether I will be able to convert that stock into Ferrari order forms.
Now let's shut up about $ing and try to keep this thing (the thread, not the Ferrari) on track...
You can regard the supply/load combo as a non-linear transimpedance driven by a signal-dependent current source, alongside the constraint that the supply rail must carry only noise, or failing that, a clean copy of the signal.
Take you simulators in transient/FFT mode and have a look at what an LM317 does to the rail when feeding a 1mA 1kHz current source.
Consider then a preamp circuit with zero PSRR, like Andy's Naim, and you have yet another source of HD and IMD.
So I can imagine that PSU linearity is less a problem in a circuit feeding a well bandlimited opamp operating at low to moderate gain.
Not all that glitters is gold.
Me just a lowly EE who hung around the economics faculty and the only named theorem they ever taught me there was M&M (BTW long since forgotten) and the only finance problem still relevant to me now is whether I will be able to convert that stock into Ferrari order forms.
Now let's shut up about $ing and try to keep this thing (the thread, not the Ferrari) on track...
You can regard the supply/load combo as a non-linear transimpedance driven by a signal-dependent current source, alongside the constraint that the supply rail must carry only noise, or failing that, a clean copy of the signal.
Take you simulators in transient/FFT mode and have a look at what an LM317 does to the rail when feeding a 1mA 1kHz current source.
Consider then a preamp circuit with zero PSRR, like Andy's Naim, and you have yet another source of HD and IMD.
So I can imagine that PSU linearity is less a problem in a circuit feeding a well bandlimited opamp operating at low to moderate gain.
SY said:
yeah but how about time? I got my amplifier construction book a few years back, got a copy of electronic workbench / microsim and my amp is still laying around in my garage in pieces, 🙂
I'm certain many will disagree too!
Andy,
Not a disagreement but, some might now ask how one achieves very good linearity in a power supply, and more importantly, how does one measure distortion for error signals down in low microvolt region and often in the noise floor the noise floor for actual signal currents?
It might be intesting to think about how low distortion is acheived in typical circuits. A power supply is swinging very low voltages so the voltage dependent distortions like modulation of device capacitances are orders of magnitude lower than audio signal swings at the typical 1 V and above levels.
Larger bias current levels are one of the
standard ways to increase linearity in a circuit. They tend to decrease open loop output impedance. and increase bandwidth. These factors will result in lower impedance levels and/or the requirement for less negative feedback for a given output impedance. They also often increase in the bandwidth over the impedance is flat.
PSRR is also extremely important and even power supplies with what appear to be very good numbers still let one hear differences in the design and component quality for the raw supply.
Rejection of frequencies above the audioband and especially at RF frequencies are crucial and may even more important than output impedance.
It seems to me to be the only advantage of the shunt over series topology for the same type regulator with the same bias current.
Linearity is a great goal and I have no problem with it but it is such a blanket term that it is almost useless without
some measurements and specific design techniques to achieve it.
The things that make a good sounding audio circuits seem to make good sounding power supplies as well, they being; parts quality, RF immunity, intellegent amounts (not extemely large) of negative feedback, and good transient response in and above the audio bandwith.
Lumping all this under the lable of "more linear circuits" seems to lead right back into the loop of not knowing the actual measureable design parameters that make for good sonics. The endless arguments about what makes a good amplifier circuit can now be expanded to include power supplies and the subjective morass that leads to.
You have raised the blind about a quarter of an inch, but still have not shed enough light on the subject for useful design goals and measurement
techniques. You are an engineer after all, and not just a subjective "lets try everything" ....... right?
Andy,
Not a disagreement but, some might now ask how one achieves very good linearity in a power supply, and more importantly, how does one measure distortion for error signals down in low microvolt region and often in the noise floor the noise floor for actual signal currents?
It might be intesting to think about how low distortion is acheived in typical circuits. A power supply is swinging very low voltages so the voltage dependent distortions like modulation of device capacitances are orders of magnitude lower than audio signal swings at the typical 1 V and above levels.
Larger bias current levels are one of the
standard ways to increase linearity in a circuit. They tend to decrease open loop output impedance. and increase bandwidth. These factors will result in lower impedance levels and/or the requirement for less negative feedback for a given output impedance. They also often increase in the bandwidth over the impedance is flat.
PSRR is also extremely important and even power supplies with what appear to be very good numbers still let one hear differences in the design and component quality for the raw supply.
Rejection of frequencies above the audioband and especially at RF frequencies are crucial and may even more important than output impedance.
It seems to me to be the only advantage of the shunt over series topology for the same type regulator with the same bias current.
Linearity is a great goal and I have no problem with it but it is such a blanket term that it is almost useless without
some measurements and specific design techniques to achieve it.
The things that make a good sounding audio circuits seem to make good sounding power supplies as well, they being; parts quality, RF immunity, intellegent amounts (not extemely large) of negative feedback, and good transient response in and above the audio bandwith.
Lumping all this under the lable of "more linear circuits" seems to lead right back into the loop of not knowing the actual measureable design parameters that make for good sonics. The endless arguments about what makes a good amplifier circuit can now be expanded to include power supplies and the subjective morass that leads to.
You have raised the blind about a quarter of an inch, but still have not shed enough light on the subject for useful design goals and measurement
techniques. You are an engineer after all, and not just a subjective "lets try everything" ....... right?
Re: I'm certain many will disagree too!
After reading all those regulator threads (and there was quite a number of them), I still feel that above suggestion is probably the only choice left for me to learn which supply is good for preamp. All this talk about PSRR, distortion, linearity, open loop, bandwith, output impedance, negative feedback, RF immunity, shunt and series topology, transient response and so on, may be entertaining for all you engineers in a house, but for somebody looking for "good preamp supply" I believe is just another useless info.
I still have to see a thread, where the issues of PS are described in more friendly way and some useful examples are presented (together with subjective comments, "what sounds better"). Otherwise we will always end up in an "endless arguments about what makes a good power supply".
I guess my first comparison will be between the supply Andy offers on his site and a simple one transistor version Nelson used in his preamps. I guess I'll try batteries as well😉
Did somebody say, I'm not doing my own R&D's? 😉
Fred Dieckmann said:
After reading all those regulator threads (and there was quite a number of them), I still feel that above suggestion is probably the only choice left for me to learn which supply is good for preamp. All this talk about PSRR, distortion, linearity, open loop, bandwith, output impedance, negative feedback, RF immunity, shunt and series topology, transient response and so on, may be entertaining for all you engineers in a house, but for somebody looking for "good preamp supply" I believe is just another useless info.
I still have to see a thread, where the issues of PS are described in more friendly way and some useful examples are presented (together with subjective comments, "what sounds better"). Otherwise we will always end up in an "endless arguments about what makes a good power supply".
I guess my first comparison will be between the supply Andy offers on his site and a simple one transistor version Nelson used in his preamps. I guess I'll try batteries as well😉
Did somebody say, I'm not doing my own R&D's? 😉
linearity
It's curious to me that linearity is considered the most important feature of a DC power supply but not the most important attribute of an amplifier (presumably driven by a power supply...).
It's curious to me that linearity is considered the most important feature of a DC power supply but not the most important attribute of an amplifier (presumably driven by a power supply...).
- Status
- Not open for further replies.