A phono stage with 60 db of gain to line level needs an ultra quiet supply. Nothing less will do. However, because of the small voltages, the dynamic load is very small.
The other end of that spectrum is a push-pull output, where the high signal levels and high PSRR tax the current capabilities of either shunt or series regulation, but noise is not the overriding consideration.
I am saying that there is not one perfect supply for every circuit.
You need to look at a number of variables and choose the supply that makes the most sense to you.
I was trying to decide between the PS I posted on my 6H30 preamp or a Tube shunt regulated PS. After thinking about it a week, I concluded that either had the potential to sound fabulous.
Doug
DougL-
thanks for the reply
For all of You-
thanks for all Your replies in this thread- I will have to, after all, try at least few different designs to find something suitable- first of all the two easiest ones (bipolars/ mosfets) and then some with tubes regulators- I will make two identical prototypes of my line to compare them more easilly.
Thanks again a lot
regards
daniel
thanks for the reply
For all of You-
thanks for all Your replies in this thread- I will have to, after all, try at least few different designs to find something suitable- first of all the two easiest ones (bipolars/ mosfets) and then some with tubes regulators- I will make two identical prototypes of my line to compare them more easilly.
Thanks again a lot
regards
daniel
Frank, Frank... rationalize.... I've said it before and I'll say it again, it's all in your head...
What's that have to do with anything? We aren't looking for a Zo match, we're looking for invisible Zo (zero ohms ideally). Consistency is no problem, add 20uF and a series resistor to tune to taste. At F's below where the 20uF does anything, the regulator can handle it. It's made to.
HWA!?
Face it, it's class A, there always has to be a bias current in the thing. Well, if you unload it, it might be possible to call it AB, depending on how it's wired. But chances are, it's also wired to be "linear" (I put in quotes because the next part, NFB, is true, and we're near cutoff, so I'd rather use the word "responsive") at these levels too. So that also breaks down.
Two things: 1. error amps handle 100kHz easily. And that's worst case. 2. BYPASS CAPACITORS.
Which of course can be squashed by using far more C than gas tubes can handle.
And what's this BS about regulating a class A amp anyway, it doesn't even give a $@*&!!
- With a SS regulator, provided it's wired correctly, you can remove the load: no problem.
You can pull the series pass device: no DC pass, no problem.
You can pull the error amp or VR, ditto.
(Of course, such is not necessary because the devices are more reliable = semi-permanent. If one fails, it doesn't really matter if it takes the house down with it, because it doesn't happen nearly as often. Not to say that's good engineering practice, of course.)
Tim
fdegrove said:
What's that have to do with anything? We aren't looking for a Zo match, we're looking for invisible Zo (zero ohms ideally). Consistency is no problem, add 20uF and a series resistor to tune to taste. At F's below where the 20uF does anything, the regulator can handle it. It's made to.
HWA!?
Face it, it's class A, there always has to be a bias current in the thing. Well, if you unload it, it might be possible to call it AB, depending on how it's wired. But chances are, it's also wired to be "linear" (I put in quotes because the next part, NFB, is true, and we're near cutoff, so I'd rather use the word "responsive") at these levels too. So that also breaks down.
Two things: 1. error amps handle 100kHz easily. And that's worst case. 2. BYPASS CAPACITORS.
Which of course can be squashed by using far more C than gas tubes can handle.
And what's this BS about regulating a class A amp anyway, it doesn't even give a $@*&!!
- With a SS regulator, provided it's wired correctly, you can remove the load: no problem.
You can pull the series pass device: no DC pass, no problem.
You can pull the error amp or VR, ditto.
(Of course, such is not necessary because the devices are more reliable = semi-permanent. If one fails, it doesn't really matter if it takes the house down with it, because it doesn't happen nearly as often. Not to say that's good engineering practice, of course.)
Tim
Hi,
It obviously isn't in yours yet....Maybe if you'd look beyond the basic function of a series re you'd understand it better?
I can only hope, right?
Quite a lot, actually...
Should have said: Act as if they're class B devices.
Response time of the regulator itself is heavily dependent on the AC feedback cap's value, it's a matter of fiddling between ripple suppression (if you want that) and speed.
Which goes to show that you seem incapable of looking beyond the basic function of a regulator?
Cheers,😉
It obviously isn't in yours yet....Maybe if you'd look beyond the basic function of a series re you'd understand it better?
I can only hope, right?
Quite a lot, actually...
Should have said: Act as if they're class B devices.
Response time of the regulator itself is heavily dependent on the AC feedback cap's value, it's a matter of fiddling between ripple suppression (if you want that) and speed.
Which goes to show that you seem incapable of looking beyond the basic function of a regulator?
Cheers,😉
Frank,
The nice thing about using a LM317 is it's simple to use, few parts and it works well and will last along time. I redesigned several old MARANTZ preamps with LM317 what a nice improvement and still working over 20 years later.
I guess the debate is if a tube regulator can perform as well as a transistor regulator. I really can't see that happening, as you know amplifiers circuits like low impedances over their frequency range, low noise, and fast response tubes just don’t perform in that application as well as their transistor counter parts. However, I am not saying that a tube regulators are bad, they can be quite good. 🙂
The nice thing about using a LM317 is it's simple to use, few parts and it works well and will last along time. I redesigned several old MARANTZ preamps with LM317 what a nice improvement and still working over 20 years later.
I guess the debate is if a tube regulator can perform as well as a transistor regulator. I really can't see that happening, as you know amplifiers circuits like low impedances over their frequency range, low noise, and fast response tubes just don’t perform in that application as well as their transistor counter parts. However, I am not saying that a tube regulators are bad, they can be quite good. 🙂
Hi,
Jim,
To me it's not just a matter of what performs best on paper.
I realize quite well that if it's a regulator and nothing else you're looking for, the semiconductor based regulator will quite likely show the better specs.
Audio design however is a little less straightforward and if that regulator has a very distinct sonic fingerprint then I need to look elsewhere or at least find a way to get rid of that aspect of it.
After all, if it's just paper performance we're looking for, we'd probably all be designing with semi-conductors....
Anyway, it's good to know you had excellent results from using a LM317.
I'm not saying this or that is better but from experience I just know that every single regulator I ever designed was optimized for the load and task at hand...
In the end I did always arrive at what I wanted to "hear" from it but it didn't come from just looking at a spec sheet if you catch my drift.
Cheers, 😉
Jim,
To me it's not just a matter of what performs best on paper.
I realize quite well that if it's a regulator and nothing else you're looking for, the semiconductor based regulator will quite likely show the better specs.
Audio design however is a little less straightforward and if that regulator has a very distinct sonic fingerprint then I need to look elsewhere or at least find a way to get rid of that aspect of it.
After all, if it's just paper performance we're looking for, we'd probably all be designing with semi-conductors....
Anyway, it's good to know you had excellent results from using a LM317.
I'm not saying this or that is better but from experience I just know that every single regulator I ever designed was optimized for the load and task at hand...
In the end I did always arrive at what I wanted to "hear" from it but it didn't come from just looking at a spec sheet if you catch my drift.
Cheers, 😉
Alright Mr. Smartypants, HOW does it act class B then.
I've only ever seen it as exactly what it is, a class A cathode (or emitter, or source) follower.
Tim
I've only ever seen it as exactly what it is, a class A cathode (or emitter, or source) follower.
Tim
Hi,
This is what is said:
This is what you make of it?
Ask yourself:
What is a series voltage regulator?
What does it have do to?
Is it an amplifier?
If so, how does it respond to changes at the load?
Cheers, 😉
This is what is said:
This is what you make of it?
Ask yourself:
What is a series voltage regulator?
What does it have do to?
Is it an amplifier?
If so, how does it respond to changes at the load?
Cheers, 😉
Ehm Frank,
A pre amp usually draws an almost constant current. No matter what kind of pass device you use, whether hot glass or cold sand, it passes almost constant current. It only has to cope with small current deviations.
Cheers 😉
A pre amp usually draws an almost constant current. No matter what kind of pass device you use, whether hot glass or cold sand, it passes almost constant current. It only has to cope with small current deviations.
Cheers 😉
Hi,
Pjotr,
The pass device is not the issue anyway...
We don't even know what this regulator is going to be used for, neither do we know what class of operation the load is working only that the current range will be about 30mA tops.
The point is however that the series regulator will only react when the load is causing an error, in that respect it acts on demand just as a class B stage would...Hence the analogy.
Nowhere have I stated that series regulators are biased for class B operation as seems to be the assumption.
Cheers, 😉
Pjotr,
The pass device is not the issue anyway...
We don't even know what this regulator is going to be used for, neither do we know what class of operation the load is working only that the current range will be about 30mA tops.
The point is however that the series regulator will only react when the load is causing an error, in that respect it acts on demand just as a class B stage would...Hence the analogy.
Nowhere have I stated that series regulators are biased for class B operation as seems to be the assumption.
Cheers, 😉
fdegrove said:
OHHHHHH
Then speak clearer. You originally said they're like class B. The definition of Class B states that it is half on, half off, as an amplifier. By definition, however, it is a class A element. Class says nothing of demand, and there's just something faulty with saying "it only operates on demand".
Tim
Oh my....
Didn't meen to start that....
To clear a little bit- attached is the schematic of the preamplifier that needs that regulated power supply- guess I could do that in the first mail of the thread.... Current per ch is 14mA and plate voltage of the tube is 105Vdc. Thanks for Your replies- maybe this will help somebody (fdegrove 🙂 ).
best regards
daniel
Didn't meen to start that....
To clear a little bit- attached is the schematic of the preamplifier that needs that regulated power supply- guess I could do that in the first mail of the thread.... Current per ch is 14mA and plate voltage of the tube is 105Vdc. Thanks for Your replies- maybe this will help somebody (fdegrove 🙂 ).
best regards
daniel
isn't that a buffer instead of a pre-amp? okok...a preamp with unity gain...you could really do with a few chokes and some caps...why do u f33l like using SS regulators?? because they are cheaper??
i posted on RAT, might as well post here -- I have scanned and made a PDF of the "Mother of All Tube Regulators" -- the KEPCO 615 -- 600 V, 300 ma -- uses 3 6550's as the pass elements. anyone who wants a PDF emailed to them just email me:
jack@tech-diy.com
lifting the 615 onto the bench is a two-man job.
jack@tech-diy.com
lifting the 615 onto the bench is a two-man job.
Sch3mat1c said:
Indeed Tim,
Besides that regulated PSU’s usually don’t work in PP but single ended. That is by definition Class-A
But a regulator is not the same thing as an audio amp. It is usually loaded with a big cap. At low frequencies feedback takes care of the output voltage and is responsible for the output impedance but at medium and high frequencies the output cap takes care of the output impedance.
The regulator merely keeps the output voltage constant and acts as an active ripple filter. But if a regulated PSU is necessary is questionable indeed.
Cheers 😉
li_gangyi said:
That "buffer" has a gain around 13-14......🙄
I have tried a few chokes and resistors and I would like to try it with some regulators- to see what is happening..😀
regards
daniel
Pjotr said:
Hi,
The output element in a regulator works in class A indeed but doesn't the error amp work in class AB/B?
I think that Frank is right.
Regards
Milan
No, both need to operate continuous (class A). The error amp may cut off completely if loaded severely, in which case voltage will drop because the pass tube is saturated (entering class 2).
Then at least the error amp can be called class AB, but this is the same situation as overdriving a class A stage into AB2 - it's outside the defined proper operating area and thus doesn't count.
SY's right, class doesn't really apply here anyway.
Tim
Then at least the error amp can be called class AB, but this is the same situation as overdriving a class A stage into AB2 - it's outside the defined proper operating area and thus doesn't count.
SY's right, class doesn't really apply here anyway.
Tim
- Status
- Not open for further replies.