Asen
Agreed, active devices are one of the main causes of distortion in amplifiers. However, in an amplifier they are providing voltage or current gain and the requirement is to preseve the input waveform as much as possible (ideally the output should match the input with only a change of level).
In the capacitance multiplier, we are not trying to preserve an input waveform. In fact, exactly the opposite. We are trying to modify a voltage comprising a dc element with a superimposed, (nearly) triangular ripple voltage. The aim is to get as close to pure dc as possible by reducing the rms ripple voltage to a level at which it will not be audible. If the pass transistor distorts the ripple waveform whilst reducing its magnitude, this is of little importance provided high frequency harmonics are not introduced (as they sometimes are in voltage regulators).
It does not really matter whether the ripple voltage waveform is near triangular or a little less near triangular. However, if you look at the waveform at the output of the capacitance multiplier, you will find that it is much smoother and more rounded than the triangular waveform produced by capacitors alone. It is not a pure sine wave, but being smoother, any audible ripple effects should be much less objectionable than those for a triangular waveform.
This is another benefit of the capacitance multiplier. As well as reducing the cost by needing smaller capacitors, the resultant ripple waveform is more benign.
Timo
Thanks for your observations on the audible effects of choke input filters. I have read of similar results elsewhere. I must get round to trying it myself one of these days.
Three-phase is not a practical option here in the UK as three-phase supplies are not permitted in residential properties and I have no intention of going to the trouble of building a single-phase to three-phase converter. Taking things a stage further though, even better results could be obtained from a three-phase 400Hz supply
Like you, I don't like ic voltage regulators. Have you tried using an audio power ic (eg LM1875 or similar) in a regulator circuit instead? I believe that this could give some sonic improvements but I have yet to try it out.
Geoff
[Edited by Geoff on 04-22-2001 at 11:09 AM]
Agreed, active devices are one of the main causes of distortion in amplifiers. However, in an amplifier they are providing voltage or current gain and the requirement is to preseve the input waveform as much as possible (ideally the output should match the input with only a change of level).
In the capacitance multiplier, we are not trying to preserve an input waveform. In fact, exactly the opposite. We are trying to modify a voltage comprising a dc element with a superimposed, (nearly) triangular ripple voltage. The aim is to get as close to pure dc as possible by reducing the rms ripple voltage to a level at which it will not be audible. If the pass transistor distorts the ripple waveform whilst reducing its magnitude, this is of little importance provided high frequency harmonics are not introduced (as they sometimes are in voltage regulators).
It does not really matter whether the ripple voltage waveform is near triangular or a little less near triangular. However, if you look at the waveform at the output of the capacitance multiplier, you will find that it is much smoother and more rounded than the triangular waveform produced by capacitors alone. It is not a pure sine wave, but being smoother, any audible ripple effects should be much less objectionable than those for a triangular waveform.
This is another benefit of the capacitance multiplier. As well as reducing the cost by needing smaller capacitors, the resultant ripple waveform is more benign.
Timo
Thanks for your observations on the audible effects of choke input filters. I have read of similar results elsewhere. I must get round to trying it myself one of these days.
Three-phase is not a practical option here in the UK as three-phase supplies are not permitted in residential properties and I have no intention of going to the trouble of building a single-phase to three-phase converter. Taking things a stage further though, even better results could be obtained from a three-phase 400Hz supply
Like you, I don't like ic voltage regulators. Have you tried using an audio power ic (eg LM1875 or similar) in a regulator circuit instead? I believe that this could give some sonic improvements but I have yet to try it out.
Geoff
[Edited by Geoff on 04-22-2001 at 11:09 AM]
Geoff
When I said "distortion" I didn't mean the ripple form or the reproduction of a sine wave, but the ability of the PSU transistors to provide a sufficient current when necessary.
Geoff and Timo
You are talking about three-phase supplies, which reminds me - has s.o. tried a balanced power supply? I heard this kind of supplies are widely used in record studios for noise reduction and better dynamics. This is connected with future DIY plans of mine, but any experience is welcome.
Regards
Asen
When I said "distortion" I didn't mean the ripple form or the reproduction of a sine wave, but the ability of the PSU transistors to provide a sufficient current when necessary.
Geoff and Timo
You are talking about three-phase supplies, which reminds me - has s.o. tried a balanced power supply? I heard this kind of supplies are widely used in record studios for noise reduction and better dynamics. This is connected with future DIY plans of mine, but any experience is welcome.
Regards
Asen
Asen
I now see what you are getting at. However, I don't see a problem. The output capacitor on the capacitance multiplier will, if adequately sized, meet any short term peak current requirments even though I doubt that the pass transistor will have any difficulty in providing the necessary current since it is operating well within its capabilities.
You have completely lost me with the balanced supply. I have heard of balanced signal inputs and outputs (used to reduce noise etc)but never a balanced power supply. Can you supply more details about what you are thinking of?
Geoff
I now see what you are getting at. However, I don't see a problem. The output capacitor on the capacitance multiplier will, if adequately sized, meet any short term peak current requirments even though I doubt that the pass transistor will have any difficulty in providing the necessary current since it is operating well within its capabilities.
You have completely lost me with the balanced supply. I have heard of balanced signal inputs and outputs (used to reduce noise etc)but never a balanced power supply. Can you supply more details about what you are thinking of?
Geoff
Re: Balanced Power
I follow the discussion with great interest.
But I don´t see the problem too.
There are only two solutions ( for my at my level of knowledge) - cheaper and more expensive. Demands on power supply are: sufficient current and as low as possible ripple.
First solution uses capacitor multiplier (i combination with 10mF at the output). Second one only capacitors. The first one offers ripple around 5mV, in simulation current more than 4A. Second one (with capacity more than 6OmF) has ripple not less than 1 V. Of course it is able to get more dynamic current. Even using 100mF I can not get better values than 500mV. The costs you can compare yourself.
My question - what can be heared more and what bothers more - ripple or worse ( a little) dynamic performace /at sure cost/?
I am for capacitor multiplier (like ESP) with 10mF.
I follow the discussion with great interest.
But I don´t see the problem too.
There are only two solutions ( for my at my level of knowledge) - cheaper and more expensive. Demands on power supply are: sufficient current and as low as possible ripple.
First solution uses capacitor multiplier (i combination with 10mF at the output). Second one only capacitors. The first one offers ripple around 5mV, in simulation current more than 4A. Second one (with capacity more than 6OmF) has ripple not less than 1 V. Of course it is able to get more dynamic current. Even using 100mF I can not get better values than 500mV. The costs you can compare yourself.
My question - what can be heared more and what bothers more - ripple or worse ( a little) dynamic performace /at sure cost/?
I am for capacitor multiplier (like ESP) with 10mF.
Asen
Thanks for the link. An interesting concept, but not really a practical proposition in a domestic situation. I think I would find it rather difficult to source a quality transformer with a 240V primary and two 120V secondaries, even if I could afford the cost of rewiring the house.
Koy
Thanks for your comments and support. I was beginning to think I was a lone voice on this one.
Geoff
Thanks for the link. An interesting concept, but not really a practical proposition in a domestic situation. I think I would find it rather difficult to source a quality transformer with a 240V primary and two 120V secondaries, even if I could afford the cost of rewiring the house.
Koy
Thanks for your comments and support. I was beginning to think I was a lone voice on this one.
Geoff
Asen
OK, let's exchange some more ideas!
Why should the pass transistor in a capacitance multiplier be any 'slower' than the diodes the rectifier? In a normal rectifier/capacitor supply the diodes supply the current to the amp during the capacitor charging period, so will the capactitance multiplier be any worse?
The majority of amps perform well with relatively low values of smoothing capacitor (4700uF to 10000uF) in a traditional rectifier/capacitor arrangement. They are generally Class-AB designs with widely varing current requirements, yet the low value of capacitance does not cause a problem. One of my commercial amps has 6700uF smoothing capacitors yet will deliver 40A to the load for short periods.
With the amp designs we are considering (Aleph, JLH and similar) the peak current is limited by the design, and by the quiescent current, and is going to be less than 5A or so. The only reason we want a large capacitance is to reduce the supply rail ripple because a) the simple circuits we are using do not have as good a PSRR as more complicated ones with by-passed current sources, current mirrors et al, and b) because a high ripple is present at low amplifier output levels and is therefore more audible than in a Class-AB amp where the ripple only becomes significant at high output levels and is then masked by the signal.
Provided the output capacitor on the capacitance multiplier is adequately sized (though, I must stress, not excessively so) the 'speed' of the supply should be no worse than that for a conventional rectifier capacitor arrangement.
Over to you to continue the debate (I hope we are not getting too far away from hifi's original topic for this thread - it's a long time since I read it!).
Geoff
OK, let's exchange some more ideas!
Why should the pass transistor in a capacitance multiplier be any 'slower' than the diodes the rectifier? In a normal rectifier/capacitor supply the diodes supply the current to the amp during the capacitor charging period, so will the capactitance multiplier be any worse?
The majority of amps perform well with relatively low values of smoothing capacitor (4700uF to 10000uF) in a traditional rectifier/capacitor arrangement. They are generally Class-AB designs with widely varing current requirements, yet the low value of capacitance does not cause a problem. One of my commercial amps has 6700uF smoothing capacitors yet will deliver 40A to the load for short periods.
With the amp designs we are considering (Aleph, JLH and similar) the peak current is limited by the design, and by the quiescent current, and is going to be less than 5A or so. The only reason we want a large capacitance is to reduce the supply rail ripple because a) the simple circuits we are using do not have as good a PSRR as more complicated ones with by-passed current sources, current mirrors et al, and b) because a high ripple is present at low amplifier output levels and is therefore more audible than in a Class-AB amp where the ripple only becomes significant at high output levels and is then masked by the signal.
Provided the output capacitor on the capacitance multiplier is adequately sized (though, I must stress, not excessively so) the 'speed' of the supply should be no worse than that for a conventional rectifier capacitor arrangement.
Over to you to continue the debate (I hope we are not getting too far away from hifi's original topic for this thread - it's a long time since I read it!).
Geoff
Geoff
First, I’d like to draw your attention to the following link:
http://tnt-audio.com/clinica/ssps1_e.html
May be it is not fair to debate with you using someone other’s words, but I think this article is rather well written.
You are right, there are still other semiconductors in the PSU – the rectifier diodes. I usually use super fast diodes, where it is possible, because these devices have better recovering characteristics. I used such devices in my last SRPP tube preamp project. The result was wonderful, but I haven’t tried the same PSU with regular diodes (1N4007). Who knows – may be the picture won’t be very different. The circuit draws 20mA and I have 3 RC filters with 4 X 100 uF in the supply. The result is 2mV noise of the whole preamp circuit (the output is circa 25V). The preamp has very good dynamic performance.
Once I tried to upgrade my CD player (second transformer for the analog circuitry, 1N4007 bridge, 4400 uF caps per rail, 78XX – 79XX regulators, and another 2 X 440 uF caps after the regulators). All this was done for a dual OP amp, which draws 10 mA. The result – as if fog had fallen down in my listening room. The noise was less, but the sound was lifeless. When I showed a friend of mine (an engineer) what I had done he couldn’t believe that the original supply performed better. But that was the bitter truth. So I left my simple supply consisting of 2 Elko 2200 uF caps and a RC filter (1 Ohm, 47 uF + 22 nF) per rail. We’re talking negligible currents here – 10-20 mA. In power amps problems should rise together with the voltages and the currents. I myself have seen even worse cases than you, Geoff – a Yamaha receiver with 5600 uF per rail. But commercial manufacturers put money in color displays, knobs, switches etc. to make the clients happier and neglect the parts that should reproduce music.
I think that the most important thing in the supply are the caps. The bigger the better, and the faster the better. I own a hand-made class AB amp that performs pretty well for such a simple circuitry. It does so for 30-40 seconds even unplugged from the wall. During this period there are no diodes involved, just VERY big caps. That makes me to think that a big capacitance could compensate even slow recovering in the PSU due to slow diodes or bad transformers.
It’s difficult for me to compare PSUs for class AB and class A amps. The process of supplying and recovering in PSUs for class A puts the transformers, bridges and caps in different dynamic conditions.
In general Koy is right – if one wants low hum, one should use capacitance multiplier; if the aim is more dynamic performance – one should use big caps.
I’ll add: if one wants the best of both worlds – should use a battery PSU.
Just take a look at Hiraga’s Monster amp. There are schematics of the amp somewhere in the Internet. It’s a 8W class A project. Pay attention to the PSU – lots of caps and motor-lorry batteries. This PSU can handle with currents that could kill a horse with ease. I do not want to think that names like Hiraga, Pass, etc. would put such big supplies in their amps just like that – for nothing. After all they are world-wide known for the performance of their products, not for the supplies they put in them.
OK – lets put the final stop here. Theory and practice sometimes show differences. I’ll try both types of PSUs for the JLH and the Andrea’s Follower and will inform you for the results. This will take time though.
Thanks for the patience of everyone.
Best regards
Asen
First, I’d like to draw your attention to the following link:
http://tnt-audio.com/clinica/ssps1_e.html
May be it is not fair to debate with you using someone other’s words, but I think this article is rather well written.
You are right, there are still other semiconductors in the PSU – the rectifier diodes. I usually use super fast diodes, where it is possible, because these devices have better recovering characteristics. I used such devices in my last SRPP tube preamp project. The result was wonderful, but I haven’t tried the same PSU with regular diodes (1N4007). Who knows – may be the picture won’t be very different. The circuit draws 20mA and I have 3 RC filters with 4 X 100 uF in the supply. The result is 2mV noise of the whole preamp circuit (the output is circa 25V). The preamp has very good dynamic performance.
Once I tried to upgrade my CD player (second transformer for the analog circuitry, 1N4007 bridge, 4400 uF caps per rail, 78XX – 79XX regulators, and another 2 X 440 uF caps after the regulators). All this was done for a dual OP amp, which draws 10 mA. The result – as if fog had fallen down in my listening room. The noise was less, but the sound was lifeless. When I showed a friend of mine (an engineer) what I had done he couldn’t believe that the original supply performed better. But that was the bitter truth. So I left my simple supply consisting of 2 Elko 2200 uF caps and a RC filter (1 Ohm, 47 uF + 22 nF) per rail. We’re talking negligible currents here – 10-20 mA. In power amps problems should rise together with the voltages and the currents. I myself have seen even worse cases than you, Geoff – a Yamaha receiver with 5600 uF per rail. But commercial manufacturers put money in color displays, knobs, switches etc. to make the clients happier and neglect the parts that should reproduce music.
I think that the most important thing in the supply are the caps. The bigger the better, and the faster the better. I own a hand-made class AB amp that performs pretty well for such a simple circuitry. It does so for 30-40 seconds even unplugged from the wall. During this period there are no diodes involved, just VERY big caps. That makes me to think that a big capacitance could compensate even slow recovering in the PSU due to slow diodes or bad transformers.
It’s difficult for me to compare PSUs for class AB and class A amps. The process of supplying and recovering in PSUs for class A puts the transformers, bridges and caps in different dynamic conditions.
In general Koy is right – if one wants low hum, one should use capacitance multiplier; if the aim is more dynamic performance – one should use big caps.
I’ll add: if one wants the best of both worlds – should use a battery PSU.
Just take a look at Hiraga’s Monster amp. There are schematics of the amp somewhere in the Internet. It’s a 8W class A project. Pay attention to the PSU – lots of caps and motor-lorry batteries. This PSU can handle with currents that could kill a horse with ease. I do not want to think that names like Hiraga, Pass, etc. would put such big supplies in their amps just like that – for nothing. After all they are world-wide known for the performance of their products, not for the supplies they put in them.
OK – lets put the final stop here. Theory and practice sometimes show differences. I’ll try both types of PSUs for the JLH and the Andrea’s Follower and will inform you for the results. This will take time though.
Thanks for the patience of everyone.
Best regards
Asen
Asen
I'll keep this short (I hope).
Agreed, battery supply is best (but not really practical for Class-A power amps).
I'm not surprised you had problems with your CD player. 78xx/79xx regulators are somewhat lacking (understatement).
To use one of your own references, if ultrafast diodes are so good why does Nelson Pass suggest that ordinary diodes are too fast and should be slowed down by using capacitors in parallel with them? ("The Importance of the Power Supply" - May, 1997)
I'm looking forward to hearing the results of your trials.
Geoff
I'll keep this short (I hope).
Agreed, battery supply is best (but not really practical for Class-A power amps).
I'm not surprised you had problems with your CD player. 78xx/79xx regulators are somewhat lacking (understatement).
To use one of your own references, if ultrafast diodes are so good why does Nelson Pass suggest that ordinary diodes are too fast and should be slowed down by using capacitors in parallel with them? ("The Importance of the Power Supply" - May, 1997)
I'm looking forward to hearing the results of your trials.
Geoff
Geoff,
Caps in parallel with diodes are to shunt switching transients in much the same way that folks frequently put a .1 uF cap (or thereabouts) across a switch to supress arcs between the contacts at turn on/off. It's not so much to 'slow the diode down' as to provide an alternate path for high frequencies so as to avoid having them zip down the line into the rail.
Soft recovery diodes don't generate as brutal a spike and do not benefit much (if any) from caps placed in parallel with them. I haven't done a direct cost comparison, but I imagine that a cheap diode with a parallel cap is probably less expensive than a soft recovery diode (IRF FRED or the equivalent from Harris, et. al.). In any event, think of a diode w/cap as a poor man's soft recovery diode. One of you fellows can simulate the situation with software, but I'd imagine that the diode/cap combination probably provides 90% of the performance at half the price. Nelson isn't notably a tweak kind of guy; more bang for the buck. I imagine that he chooses diode/cap for the price/performance ratio.
In my case, I spent a few extra dollars and bypassed soft recovery diodes (IRF FREDs, in my case) with caps. Redundant? Probably. Audible? Frankly, I don't know. But Sprague Orange Drops are relatively cheap, so I put 'em in. Even if they're inaudible, I'm only out a few dollars. It's not as though I have to answer to an accounting department.
And I agree, 78/79XX regulators sound just plain awful.
Grey
Caps in parallel with diodes are to shunt switching transients in much the same way that folks frequently put a .1 uF cap (or thereabouts) across a switch to supress arcs between the contacts at turn on/off. It's not so much to 'slow the diode down' as to provide an alternate path for high frequencies so as to avoid having them zip down the line into the rail.
Soft recovery diodes don't generate as brutal a spike and do not benefit much (if any) from caps placed in parallel with them. I haven't done a direct cost comparison, but I imagine that a cheap diode with a parallel cap is probably less expensive than a soft recovery diode (IRF FRED or the equivalent from Harris, et. al.). In any event, think of a diode w/cap as a poor man's soft recovery diode. One of you fellows can simulate the situation with software, but I'd imagine that the diode/cap combination probably provides 90% of the performance at half the price. Nelson isn't notably a tweak kind of guy; more bang for the buck. I imagine that he chooses diode/cap for the price/performance ratio.
In my case, I spent a few extra dollars and bypassed soft recovery diodes (IRF FREDs, in my case) with caps. Redundant? Probably. Audible? Frankly, I don't know. But Sprague Orange Drops are relatively cheap, so I put 'em in. Even if they're inaudible, I'm only out a few dollars. It's not as though I have to answer to an accounting department.
And I agree, 78/79XX regulators sound just plain awful.
Grey
Grey, Asen et al
With regard to the debate on diodes, I can do no better than to quote from the Nelson Pass' article.
"Rectifiers.
Yeah, sure, rectifiers are important, after all, the AC has to get converted to DC, but I don’t like the fast recovery types that some audiophiles have raved about. Fast recovery means that they withstand many amps and volts in a tenth of a few nano-seconds, something we don’t see very often on the old 60 Hz AC line. They are essential element in switching power supplies, but for regular "linear" power supplies, I much prefer SLOW diodes, and we create them by placing small capacitor circuits across the diodes, which greatly reduces radiated noise."
As usual in audio, we have two conflicting schools of thought, fast recovery diodes or slow diodes. Which is correct? I don't know. Is there an audible difference? Again I don't know. Basically, it comes down to personal preference and to which pundit we choose to believe.
Geoff
With regard to the debate on diodes, I can do no better than to quote from the Nelson Pass' article.
"Rectifiers.
Yeah, sure, rectifiers are important, after all, the AC has to get converted to DC, but I don’t like the fast recovery types that some audiophiles have raved about. Fast recovery means that they withstand many amps and volts in a tenth of a few nano-seconds, something we don’t see very often on the old 60 Hz AC line. They are essential element in switching power supplies, but for regular "linear" power supplies, I much prefer SLOW diodes, and we create them by placing small capacitor circuits across the diodes, which greatly reduces radiated noise."
As usual in audio, we have two conflicting schools of thought, fast recovery diodes or slow diodes. Which is correct? I don't know. Is there an audible difference? Again I don't know. Basically, it comes down to personal preference and to which pundit we choose to believe.
Geoff
Diodes
As to me I agree - with slow diodes (diode + cap).
As I can remember, Mr. Dudek (one of the Czech leading constructors, he is called "Czech Burmester") preferres the slow type too. Why fast?? On one hand we use filters to remove high frequency noise on the other we put in the fast diodes. Why?
As to me I agree - with slow diodes (diode + cap).
As I can remember, Mr. Dudek (one of the Czech leading constructors, he is called "Czech Burmester") preferres the slow type too. Why fast?? On one hand we use filters to remove high frequency noise on the other we put in the fast diodes. Why?
Geoff,
I think the key is in the last line of your quote from Nelson: "...which greatly reduces radiated noise." Which is the same thing I was saying. Caps across diodes don't 'slow them down,' per se, they supress spikes.
My apologies...I believe I'm guilty here of being mushy in my terminology--"soft" recovery vs. "fast" recovery--they're two separate things. The diodes I used in the power supplies for my tube amp are soft recovery, meaning that an oscilloscope shows less of a glitch as they go from conducting to non-conducting, and vice versa. I seem vaguely to recall seeing a 'scope shot that showed a fast recovery diode with a pretty sharp switching spike. I can see how one would want to avoid that.
At any rate, I routinely run caps across all my bridges, whether fast, slow, or soft recovery. Reduces high frequency hash, and provides protection for the diodes from high frequency spikes coming in from the outside.
Does it actually sound better? I dunno. But it's cheap insurance. One of these days when I have nothing better to do, I'll pull the caps off some of these bridges and see if I can hear a difference.
Grey
P.S.: Okay, maybe I'm not so lost in the woods as I feared. I just took a second to review my previous post. I did say "soft" recovery. You had me worried, there, thinking I'd put "fast" when I meant to say "soft." *Whew* Okay, back to you guys.
[Edited by GRollins on 04-25-2001 at 09:54 AM]
I think the key is in the last line of your quote from Nelson: "...which greatly reduces radiated noise." Which is the same thing I was saying. Caps across diodes don't 'slow them down,' per se, they supress spikes.
My apologies...I believe I'm guilty here of being mushy in my terminology--"soft" recovery vs. "fast" recovery--they're two separate things. The diodes I used in the power supplies for my tube amp are soft recovery, meaning that an oscilloscope shows less of a glitch as they go from conducting to non-conducting, and vice versa. I seem vaguely to recall seeing a 'scope shot that showed a fast recovery diode with a pretty sharp switching spike. I can see how one would want to avoid that.
At any rate, I routinely run caps across all my bridges, whether fast, slow, or soft recovery. Reduces high frequency hash, and provides protection for the diodes from high frequency spikes coming in from the outside.
Does it actually sound better? I dunno. But it's cheap insurance. One of these days when I have nothing better to do, I'll pull the caps off some of these bridges and see if I can hear a difference.
Grey
P.S.: Okay, maybe I'm not so lost in the woods as I feared. I just took a second to review my previous post. I did say "soft" recovery. You had me worried, there, thinking I'd put "fast" when I meant to say "soft." *Whew* Okay, back to you guys.
[Edited by GRollins on 04-25-2001 at 09:54 AM]
Grey
It's OK, I didn't misunderstand you. We are in agreement here. Even the most basic rectifier diode is fast enough for a power supply. What we want is the device or circuit arrangement that generates the least spuriae, be it soft recovery diodes, parallel capacitors or both. I only wish I had the test gear (other than my ears) to measure the alternatives.
Geoff
It's OK, I didn't misunderstand you. We are in agreement here. Even the most basic rectifier diode is fast enough for a power supply. What we want is the device or circuit arrangement that generates the least spuriae, be it soft recovery diodes, parallel capacitors or both. I only wish I had the test gear (other than my ears) to measure the alternatives.
Geoff
true cost
Maybe I don't understand the circuit well enough, but don't the transistors end up supplying the full current and voltage. This being the case then the power supply would need to be of similar size to a class b output of equal wattage. I.e. if you had a 100 watt class A(yes I know this is large) amp with 200 watts+input end draw then the power supply transistors would need to sized like a 200+ watt class b output stage with heatsinks to match. In addition the power supply transistors would have inefficiencies that would require a significant increase in the transformer size.
For small amps like the 10watt Hood amp this was designed for 1 pair of to3's a small heat sink and would be about all that is required in addition to some more transformer. For a larger amp it will still get expensive. Obviously you will never get the same regulation from capacitors that will fit in the same amount of space, but the efficiency and heat trade off is not insignificant.
Wade
Maybe I don't understand the circuit well enough, but don't the transistors end up supplying the full current and voltage. This being the case then the power supply would need to be of similar size to a class b output of equal wattage. I.e. if you had a 100 watt class A(yes I know this is large) amp with 200 watts+input end draw then the power supply transistors would need to sized like a 200+ watt class b output stage with heatsinks to match. In addition the power supply transistors would have inefficiencies that would require a significant increase in the transformer size.
For small amps like the 10watt Hood amp this was designed for 1 pair of to3's a small heat sink and would be about all that is required in addition to some more transformer. For a larger amp it will still get expensive. Obviously you will never get the same regulation from capacitors that will fit in the same amount of space, but the efficiency and heat trade off is not insignificant.
Wade
Wade
No, the power dissipation in the capacitance multiplier is not that great (one of its advantages is that it dissipates far less power than a voltage regulator). Though the current through the pass transistor is the same as that for the amp, the volt drop across it is in the order of 3V, so for an amp drawing say 4A (64Wrms into 8ohm)the power dissipated (as heat) in the pass transistor is only 12W. It will still need a heatsink, but not one of an extravagant size.
Geoff
No, the power dissipation in the capacitance multiplier is not that great (one of its advantages is that it dissipates far less power than a voltage regulator). Though the current through the pass transistor is the same as that for the amp, the volt drop across it is in the order of 3V, so for an amp drawing say 4A (64Wrms into 8ohm)the power dissipated (as heat) in the pass transistor is only 12W. It will still need a heatsink, but not one of an extravagant size.
Geoff
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.