Dear Bob,
I stripped an old Sony Esprit that I got from my dad and I saw that they did very much as you suggest. They run all the pre amplification and VAS from a stabilised PSU. The output stage they use 22000uF per rail then throught an inductor (round core) followed by a capacitance multiplier.
Their output devices are 2SC3856 and 2SA1402 (eight per channel) I have not established if they are BJT.
Besides that they make use of star on star. They also have two 4700uF 100V caps directly across the output transistors from rail to rail.
The capacitors used are Great Supply by Nichicon all rated at 105 degree centigrade.
Although all caps are soldered to the boards, they are also clamped together with a rubber shock absorbing block between them.
All wiring are stranded pre-tinned and connections are wire wrap not soldered.
PCB was double sided with earthplane and gold plated, really looked like they took trouble with this baby.
The pre-amplifier is completely passive but still boasts bass treble balance, by-pass and a magnitude of knobs and switches for dubbing etc.
I stripped an old Sony Esprit that I got from my dad and I saw that they did very much as you suggest. They run all the pre amplification and VAS from a stabilised PSU. The output stage they use 22000uF per rail then throught an inductor (round core) followed by a capacitance multiplier.
Their output devices are 2SC3856 and 2SA1402 (eight per channel) I have not established if they are BJT.
Besides that they make use of star on star. They also have two 4700uF 100V caps directly across the output transistors from rail to rail.
The capacitors used are Great Supply by Nichicon all rated at 105 degree centigrade.
Although all caps are soldered to the boards, they are also clamped together with a rubber shock absorbing block between them.
All wiring are stranded pre-tinned and connections are wire wrap not soldered.
PCB was double sided with earthplane and gold plated, really looked like they took trouble with this baby.
The pre-amplifier is completely passive but still boasts bass treble balance, by-pass and a magnitude of knobs and switches for dubbing etc.
Rodent said:Bob, I never said amplifiers sound the same, the actual question was how large should a capacitor be and what is the raiting of the transformer for a given power and load.
At the least I am providing the reader with a possible solution and a practical implemetation.
You on the other hand offer nothing exept that whomever offers some solution must obviously be wrong..
So how about it then, offer this forum your rule of thumb for discussion
You need to re-read my post #216 and some of the others. Your point about how large the capacitors should be and what the rating of the transformer should be is well-taken, but is only the starting point, barely scratching the surface - that I what I have pointed out.
It is always dangerous to give out a simple one-size-fits-all rule of thumb, so I usually try to focus on the considerations governing choices, rather than the choices themselves. But I'll make an exception for you. As a minimum, each rail capacitor should be 20,000 uF. The transformer should be sized, as a minimum, as follows: If the amplifier clips at X watts into 8 ohms resistive, it should clip at no less than 1.8X into 4 ohms resistive. An amplifier that clips at 100 watts into 8 ohms will then clip at 180 watts into 4 ohms. This means that the amplifier, which ideally would provide 2X power into 4 ohms, will be shy of this mark by only 10% at 4 ohms. This is pretty stringent, I admit.
Bob
Thank you for your contribution. Would you say that regulation may be an answer, transistors are much ceaper than monster caps, what about capacitance multipliers what are your views on this issue. Class A draws constant current (sort of) would a capacitance multiplier do the job. Lets say resevior = 10 000 uF followed by capcitance multiplier to give another "100 000 uF" this would also get rid of anu ripple as well as harmonics on the power line.
Kindest regards
Kindest regards
Hi Bob,
I have a view that the PSU ability should be designed to suit the same load as the amplifier is supposed to meet.
Your minimum of 20mF suggests that load impedance may not be a significant factor in designing your PSU. Would you care to elaborate?
I have a similar power criteria.
Design for Xwatts into Yohm(non reactive). Test the prototype at 2Yohms and measure the continuous power delivered (Zwatts). To be Yohm (reactive) capable the amp should be able to deliver 3Zwatts into Y/2ohms(non reactive).
example. 60W into 8r and >=180W into 2r (on a nearly continuous basis, heatsink permitting/time limiting) is one criteria for a 4ohm capable amplifier. These numbers imply about 100W to 110W into 4r and bring them VERY close to your 1.8factor.
I have a view that the PSU ability should be designed to suit the same load as the amplifier is supposed to meet.
Your minimum of 20mF suggests that load impedance may not be a significant factor in designing your PSU. Would you care to elaborate?
is this a design criteria for an 8ohm (reactive load) amplifier, or for a 4ohm capable amplifier?
I have a similar power criteria.
Design for Xwatts into Yohm(non reactive). Test the prototype at 2Yohms and measure the continuous power delivered (Zwatts). To be Yohm (reactive) capable the amp should be able to deliver 3Zwatts into Y/2ohms(non reactive).
example. 60W into 8r and >=180W into 2r (on a nearly continuous basis, heatsink permitting/time limiting) is one criteria for a 4ohm capable amplifier. These numbers imply about 100W to 110W into 4r and bring them VERY close to your 1.8factor.
Rodent said:
Regulation of the power supply to the output stage is understandably an issue of debate, and it is not done very often. There are at least three reasons. The first is that if you use a conventional regulator that regulates to a fixed voltage, then you give up dynamic headroom. The second is that conventional regulators have an inductive output and may not be good at supplying very high short-term current. The third is that a conventional regulator will dissipate considerable heat, especially under high line voltage conditions if it regulates to a fixed voltage. There is a middle ground where most of these limitations can be avoided, and it is probably best described as being in the category of a capacitance multiplier rather than a regulator.
Bob
AndrewT said:
I generally design my amplifiers to be at least 4-ohm full-power capable on a continuous basis, and 2 ohm capable (or better) on a limited-term basis. If I design a "200-Watt" amplifier, it will be rated at 200 Watts into 8 ohms and 360 Watts into 4 ohms. Eight ohms will always be a walk in the park for these amplifiers. The main thing about 8 ohms that I design for is that it be capable of 1/3 power into 8 ohms, both channels driven, continuously. This mainly has to do with heat sink sizing.
Bob
ingrast said:
An interesting thing happened. I was sitting listening to music the other night, pondering what I was hearing and a little bit frustrated at my lack of communication skills. As I was thinking things through a little bit of clarity infused itself and I realized that the point that I'd been so insistently presenting was, simply put, wrong. Once the rectifiers stop conducting there is no path through the secondary. I wiggled around it and tried to invoke the gods to create a path and see me out of my quandary, with no luck. I flashed briefly on the month long debate that I perpetuated, easily trashed by a fresh from school EE. Pondered my embarrassment and considered moving to Montana under an assumed alias. Nah, too much effort involved.
Then I pondered the hows and whys of I could have missed such a simple and elementary point, in the end defending it like a crazed zealot in the face of calm reason and intelligent counterpoint. Realizing I had no defense or excuse to protect me, I guessed that one would be left to the psychiatrists. Although I remember reading Andrew Carnegie years ago discussing the human ego and the lengths it will go to protect itself.
All this aside. I have no answers. I know better, but the pressures of life are released in many strange ways. I would like to thank the members of this thread for their patience and reserve (when I’m sure there was many an urge to speak bluntly). For this I’m lucky, because more stress would not have helped.
There is validity to my more coherent thoughts, although I doubt I’ll be attempting to discuss anything audio related for a while.
2006 has been a long year for me and I look forward to the new year.
Enjoy the Christmas Holidays.
Regards, Mike
Mike,
As much as anything, you were nailed by the net. We all get stuck in blind alleys from time to time. But if our discussion partners are physically present, they can immediately smack some sense into us and the issue dissolves in to a trivial diversion. Here, we can get out on a limb and hang there for a long while.
Take heart in this: Many people on this forum use fictional names, and some use them to hide behind, while they sling arrows and insults. Those people, if they were ever able to recognize being in error, would simply slink away, or just avoid the issue. You faced it head on. As our Aussie compatriats would say "good on ya mate".
All that aside, I found the discussion useful - even though I only understand some of it. The detailed theory of the ideal grounding methods is useful for those trying to match an ideal solution to a specific application. For many of the rest of us, a close approximation will due - along with some general principles, so we can improve our ability to predict practical outcomes. Your practical experience is very helpful in this regard.
Sheldon
As much as anything, you were nailed by the net. We all get stuck in blind alleys from time to time. But if our discussion partners are physically present, they can immediately smack some sense into us and the issue dissolves in to a trivial diversion. Here, we can get out on a limb and hang there for a long while.
Take heart in this: Many people on this forum use fictional names, and some use them to hide behind, while they sling arrows and insults. Those people, if they were ever able to recognize being in error, would simply slink away, or just avoid the issue. You faced it head on. As our Aussie compatriats would say "good on ya mate".
All that aside, I found the discussion useful - even though I only understand some of it. The detailed theory of the ideal grounding methods is useful for those trying to match an ideal solution to a specific application. For many of the rest of us, a close approximation will due - along with some general principles, so we can improve our ability to predict practical outcomes. Your practical experience is very helpful in this regard.
Sheldon
MikeBettinger said:
Mike, don't let this little episode dishearten you. We all make mistakes but not all are men enough to admit them, like you have done. I admire you for your courage and honesty and look forward to your contributions in the future.
All, merry Christmas and a happy New Year!
Regards,
Milan
Hi Mike,
Don't worry. I am sure we all make that mistake once in a while, managing too fool ourselves into believing something that is obviously wrong. You were just very persistent, but probably our protests made you believe even more firmly in the claim. I try to live after the principle that I do not care about who is right and who is wrong, but about what is right and what is wrong. So I am happy we have resolved this controversy and found that do after all agree. That makes it one mystery less to understand in the world.
I also think Sheldon is right that it is easier to maintain ones erroneous ideas in a forum conversation, than when having a face to face conversation. I have often wished in the midst of a discussion that we could all just assemble in front of a whiteboard and continue our discussion there.
All's well that ends well. Merry Christmas to you too, and to all the others on the forum
Don't worry. I am sure we all make that mistake once in a while, managing too fool ourselves into believing something that is obviously wrong. You were just very persistent, but probably our protests made you believe even more firmly in the claim. I try to live after the principle that I do not care about who is right and who is wrong, but about what is right and what is wrong. So I am happy we have resolved this controversy and found that do after all agree. That makes it one mystery less to understand in the world.
I also think Sheldon is right that it is easier to maintain ones erroneous ideas in a forum conversation, than when having a face to face conversation. I have often wished in the midst of a discussion that we could all just assemble in front of a whiteboard and continue our discussion there.
All's well that ends well. Merry Christmas to you too, and to all the others on the forum
MikeBettinger said:
Mike,
You could have taken the easy way of leaving all this to fade away, given the net offers no face to face relations.
You opted instead to be honest and show courage to admit publicly an error.
We have all made funny blunders in our past, and they are sculptured in rock, no way to unwind and redo. We learn from them and try to do better next time.
Please keep the good work, we are with you, and the resolution of this incident only makes you a more reliable professional and better person in our eyes.
Best wishes, Rodolfo
Bob
In Class A amplifiers, the use of inductor input filters in the PSU seem a good idea. Perhaps expensive, but with the high standing current it seems that the stress on the rectifiers would be significantly reduced and the inductor would reduce the need for huge capacitors. Alternatively, the use of inductors would permit even larger capacitors, because the current is smoothed out over a wider part of the input cycle (ideally all of it).
Cheers
John
In Class A amplifiers, the use of inductor input filters in the PSU seem a good idea. Perhaps expensive, but with the high standing current it seems that the stress on the rectifiers would be significantly reduced and the inductor would reduce the need for huge capacitors. Alternatively, the use of inductors would permit even larger capacitors, because the current is smoothed out over a wider part of the input cycle (ideally all of it).
Cheers
John
Hello john,
Thanks to introduce inductor in this discussion. More than slicing current and reducing the need of big capacitors, this component introduce modification in subjective sonic audio performance.
Using big inductor was greatly used in old schematic valves amplifiers. Could our specialist say why this solution progressively disappear in modern amplifiers ?
Eric
Thanks to introduce inductor in this discussion. More than slicing current and reducing the need of big capacitors, this component introduce modification in subjective sonic audio performance.
Using big inductor was greatly used in old schematic valves amplifiers. Could our specialist say why this solution progressively disappear in modern amplifiers ?
Eric
Eric Juaneda said:
I'm not an expert, but I can hazard some pretty good guesses:
In the tube days, large value capacitors were very large and expensive - early electrolytics weren't very small, good or reliable. And, of course, neither were foil types small or cheap, though I'd guess more reliable. Chokes are not small or cheap, but they were small and cheap by comparison. Also, in the early tube days, tube diodes (rectifier tubes) were used. They cannot tolerate a very large value cap right at the output (cap input), as they cannot handle the large current spikes. So inductors were used even on most cap input power supplies.
Since then, cap technology has come a long way, and now large value caps are fairly small, light and cheap. And we now have SS diodes, which can easily tolerate large current spikes.
So why not still use inductors? Some tube amp manufacturers do. But for SS, it's much easier to build circuits with very good PSRR, not as easy for tubes - especially SE triode amps. So an adequate power supply can be made with a few caps and resistors. And chokes are as large and heavy as they ever were. Also tube amps run at higher voltages and lower power supply current. SS amps run at much lower PS voltages and much higher current, making the needed choke that much larger and heaver and more expensive.
Sheldon
Hi Eric
Well ... I think two reasons for starters are expense, and expense.
Most choke coils also need air gaps which aren't easy to accomplish in toroidal cores (maybe a fine hacksaw woudl do it) so probably we would be stuck with ordinary EI types. These are bulkier than toriods would be. THe nearest equivalent to toroids would be the old C-core but these seem to have disappeared, as toriods rule!
But it seems to me with some people wanting high power Class A this would be beneficial.
John
Well ... I think two reasons for starters are expense, and expense.
Most choke coils also need air gaps which aren't easy to accomplish in toroidal cores (maybe a fine hacksaw woudl do it) so probably we would be stuck with ordinary EI types. These are bulkier than toriods would be. THe nearest equivalent to toroids would be the old C-core but these seem to have disappeared, as toriods rule!
But it seems to me with some people wanting high power Class A this would be beneficial.
John
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.