Flatpicker
Well-known member
Haha, it just occured to me that I should put this in the power supply meta. :wink:
Help Support GroupDIY Audio Forum:
Hmmm, maybe I should do a better job, then.
Let me see if I can make some normal copies to scan.
cj
Let me see if I can make some normal copies to scan.
cj
Larry Sheehan
Well-known member
This is the same article where this guy said his boutique power cables "broke in" over the course of the first year, correct?
G
Guest
Guest
Thanks to Gus on this one..
That little snippet of information (current is asked for) was the missing link that I needed to go and find out more info about
The rest of the stuff now makes sense and falls into place
Cheers
That little snippet of information (current is asked for) was the missing link that I needed to go and find out more info about
The rest of the stuff now makes sense and falls into place
Cheers
jtoole
Well-known member
[quote author="Larry Sheehan"]This is the same article where this guy said his boutique power cables "broke in" over the course of the first year, correct?[/quote]
Yup - thats the one :? Is there any physics to support that?
Yup - thats the one :? Is there any physics to support that?
Another interesting thing I read was that a single ended circuit will see greater effects from power supply differences as compared to a push pull circuit.
So cap changes will make more of a difference in an SE amp.
Something about the power supply being "in the circuit" as opposed to a push puller.
That, I do not understand. A push pull amp needs current too, right?
Sounds like an interesting experiment, push pull vs SE pwr csap swap out.
So cap changes will make more of a difference in an SE amp.
Something about the power supply being "in the circuit" as opposed to a push puller.
That, I do not understand. A push pull amp needs current too, right?
Sounds like an interesting experiment, push pull vs SE pwr csap swap out.
da_ardvark
Well-known member
CJ, speaking of caps, I was reading an article that was saying the even non-polar caps have a good way to install them and a bad way.
In other words even tho it "really doesnt matter" which way they're installed the effect on sound is noticable.
Any truth to this or is it just so much BS
Denny
In other words even tho it "really doesnt matter" which way they're installed the effect on sound is noticable.
Any truth to this or is it just so much BS
Denny
I remember reading something about that.....
It was something about the way in wich way you mount the caps "in the circuit" so to speak. The outer foil should be connected with lowest Z to ground or something like that.
Cant remember where i read it and dont now if theres any truth to it.
Andre
Well-known member
[quote author="CJ"]Another interesting thing I read was that a single ended circuit will see greater effects from power supply differences as compared to a push pull circuit.
So cap changes will make more of a difference in an SE amp.
Something about the power supply being "in the circuit" as opposed to a push puller.
That, I do not understand. A push pull amp needs current too, right?
Sounds like an interesting experiment, push pull vs SE pwr csap swap out.[/quote]
I'm thinking guitar amps... Power supply ripple will be seen in opposite polarity in each "half" of the signal path. When the affected signals are recombined, the ripple will cancel to some extent. This is why guitar amps typically have poor filtering on the first leg of the rail, which supplies the power tubes.
So cap changes will make more of a difference in an SE amp.
Something about the power supply being "in the circuit" as opposed to a push puller.
That, I do not understand. A push pull amp needs current too, right?
Sounds like an interesting experiment, push pull vs SE pwr csap swap out.[/quote]
I'm thinking guitar amps... Power supply ripple will be seen in opposite polarity in each "half" of the signal path. When the affected signals are recombined, the ripple will cancel to some extent. This is why guitar amps typically have poor filtering on the first leg of the rail, which supplies the power tubes.
Andre
Well-known member
Correction... They see the same ripple, but are combined at opposite polarities. I think.
da_ardvark
Well-known member
Yes Kit thats what I read. Couldn't remember exactly what foil should go where :roll:
pstamler
Well-known member
Something to further muddy the waters: Different audio circuits respond to noise on the rails in different amounts. This is called "power supply rejection ratio", and in some circuits (e.g. opamps) it varies with frequency.
For example, a simple gain stage with no feedback has a very poor power supply rejection ratio; if it's (say) a triode, crud on the line will only be reduced by rp / (rp + Rp) where rp is the plate resistance and Rp is the parallel resistance of the plate resistor and the load resistor. Whereas a gain stage with a constant-current plate load will have a *huge* power supply rejection ratio.
So the amount that a supply influences the sound is a function both of the behavior of the supply and the behavior of the audio circuits.
Here's where it gets messy, though: the numbers indicate that typical good IC opamps ought to be quite impervious to power supply considerations...but practical experience says they ain't.
Peace,
Paul
For example, a simple gain stage with no feedback has a very poor power supply rejection ratio; if it's (say) a triode, crud on the line will only be reduced by rp / (rp + Rp) where rp is the plate resistance and Rp is the parallel resistance of the plate resistor and the load resistor. Whereas a gain stage with a constant-current plate load will have a *huge* power supply rejection ratio.
So the amount that a supply influences the sound is a function both of the behavior of the supply and the behavior of the audio circuits.
Here's where it gets messy, though: the numbers indicate that typical good IC opamps ought to be quite impervious to power supply considerations...but practical experience says they ain't.
Peace,
Paul
dale116dot7
Well-known member
Ideally in a design everything should be decoupled sufficiently so as to not apply very high frequency stuff to the output of the power supply or back up the supply rails. In theory the power rails should be as solid as a brick outhouse and never allow signals to travel up and down them - at all frequencies.
One other possibility is that many old tube-mic power supplies had many stages of filtering and no 'amplification' where a solid-state regulator typically does the opposite. That solid-state regulator will put out some amount of very high frequency noise, and if you don't do something to get rid of that high frequency stuff, it'll get into your audio circuits too. Even a 'linear' regulator like an LM7815 or LM317 will generate a fair bit of high-frequency hash and hiss. If you want a noise generator, take the noise from a zener diode and amplify it. Funny that voltage regulators do exactly that... Anyways, in this case, proper filtering of the solid-state regulator should get rid of the noise.
Look carefully at the PSRR vs. frequency of an op-amp... and look at the underside of a modern 'Pentium' or similar class motherboard. There's a lot of capacitor there, and for good reason. The instantaneous spikes from switching the transistors in such a processor can be in the 20 to 100 amp region! Enough to upset the power supply rails.
My Soundcraft 2400 has horrible decoupling everywhere. Not enough of it, and some stages even oscillate mildly (only a few millivolts at 6 MHz). Some attention to that and the oscillation goes away - and it sounds much cleaner. The power supply should not make a significant difference on a console's crosstalk unless the channel strip filtering is not sufficient.
One other possibility is that many old tube-mic power supplies had many stages of filtering and no 'amplification' where a solid-state regulator typically does the opposite. That solid-state regulator will put out some amount of very high frequency noise, and if you don't do something to get rid of that high frequency stuff, it'll get into your audio circuits too. Even a 'linear' regulator like an LM7815 or LM317 will generate a fair bit of high-frequency hash and hiss. If you want a noise generator, take the noise from a zener diode and amplify it. Funny that voltage regulators do exactly that... Anyways, in this case, proper filtering of the solid-state regulator should get rid of the noise.
Look carefully at the PSRR vs. frequency of an op-amp... and look at the underside of a modern 'Pentium' or similar class motherboard. There's a lot of capacitor there, and for good reason. The instantaneous spikes from switching the transistors in such a processor can be in the 20 to 100 amp region! Enough to upset the power supply rails.
My Soundcraft 2400 has horrible decoupling everywhere. Not enough of it, and some stages even oscillate mildly (only a few millivolts at 6 MHz). Some attention to that and the oscillation goes away - and it sounds much cleaner. The power supply should not make a significant difference on a console's crosstalk unless the channel strip filtering is not sufficient.
Paul, do you have the great article on mixers and preamps floating around in a pdf or something?
Or is that copywrited thru Audio_X or something?
Or is that copywrited thru Audio_X or something?
pstamler
Well-known member
[quote author="CJ"]Paul, do you have the great article on mixers and preamps floating around in a pdf or something?
Or is that copywrited thru Audio_X or something?[/quote]
It is, I'm afraid. They bought the rights. You can get back issues, though. One of these days they'll put more recent issues on CD-ROMs along with the 1970s ones.
Glad to see that picture is still floating around, though.
Peace,
Paul
Or is that copywrited thru Audio_X or something?[/quote]
It is, I'm afraid. They bought the rights. You can get back issues, though. One of these days they'll put more recent issues on CD-ROMs along with the 1970s ones.
Glad to see that picture is still floating around, though.
Peace,
Paul
