Voltage filter, low-pass, yes. But the voltage is relatively un-interesting, as long as it stays high-enough to not crash into the signal. The current is where the important action is. The PSU needs to be able to shoot current out, always exactly as much and as fast as the power transistors try to allow. However, whenever current is supplied, the voltage is pulled down. For both a) a peak current amplitude that we can either decide on or let the load impedance decide for us, and b) long-enough time periods of peak or less than peak current demand, and any combination of a and b, the capacitor (i.e. rail) voltages need to stay high-enough to ensure that there is always enough current available to satisfy what the power transistors try to allow to flow. Whenever there is not enough charge in the reservoir, at a high-enough voltage relative to a requested signal swing and the current needed to swing it, then the output signal will get distorted.
As has been pointed out too many times, it is an extremely easy problem to solve, by using simple overkill.
But as has also been pointed out, the only reason for this thread is to investigate and characterize the region BETWEEN "overkill" and "insufficient".
Even if we imagined that overkill was the only solution, or even just the best solution, we would still want to ask how much overkill might be good-enough, and would want to provide the analysis to enable specific cost/size/weight/etc vs performance tradeoffs to be calculated.
In engineering anything, there is always a multi-dimensional trade-space, which must be characterized. And we always want to find out how to at least be ABLE to optimize performance, cost, and schedule, and properly evaluate and mitigate the risks associated with each step we might want or need to take.
And as far as I know, everyone here already believes, or at one time believed, that the bigger the reservoir capacitance was made, the better the PSU and the power amplifier would be. We are all way beyond that, or trying to go beyond that.
So, in the context of the later portion of this thread, it is unhelpful, and tiresome, for anyone to continue to simply repeat that assertion.
This is Nico's thread. He started it so he owns it. The only valid purpose of the discussion here is to be helpful in trying to find answers for what Nico asked about.
Personally, I have always tended to like overkill. But I believe that for any power amplifier linear PSU there is a large-enough but less-than-infinite rail-capacitance value such that increasing it will provide zero additional benefit, or a benefit that is so small that we would never consider increasing the capacitance beyond that point. How do we calculate that capacitance value, for a given amplifier and PSU circuit with given performance specifications?
(Hint for dunces: Saying "make it as big as possible" is NOT the type of answer we are seeking.)
Similarly, for any power amplifier with a linear PSU there is a lowest reservoir capacitance, below which the amplifier will not meet its specifications. How can we calculate the value of that capacitance?
Anywhere between those two capacitance values, the capacitance value used will affect the performance (and cost, etc) of the PSU and the amplifier, in multiple dimensions. How could we calculate those effects? After we can calculate them, then how could we calculate the minimum capacitance required to meet a given set of performance specifications? And how do we set the performance specifications?
Some of that is beyond what Nico asked for but I just wanted to mention a few things that any engineer, or engineering student, should be thinking about, in this context.
Finally, I have to say that it is a wonderful priviledge to be having this discussion, with so many intelligent people on board. Intelligence is interesting, eh? For example, in my life so far, I have noticed that many quite-intelligent people seem to think that they know more and know better than everyone else, about most things, or even about everything, for at least a part of their lives. Depending on their personality, their attitude can sometimes make them obnoxious and undesirable to have around. Sometimes, other people will not even want to bother with trying to set them straight, which just prolongs the problem. The most-intelligent ones eventually realize that almost everyone feels that way, either when they are young or at different times in their life, and then, having realized that, maybe they can learn better ways to view others, and see themselves in a more-proper perspective, and learn to communicate with and respect others at an acceptable level, and hopefully they find that their life is enriched and the direction they should take is more clear and they are more effective at whatever they try to do in life. I have found that even the stupidest people usually know something better than I do, and can often teach me something that I appreciate learning.
Engineers and scientists, in particular, have a tendency to not be very good at interpersonal communication and social interaction, maybe because they believe it is less important than being great at science and mathematics. We are dead wrong, about that, since it results in us letting others, who are probably not as smart, manage and lead and decide and implement everything, plus take most of the money and the hottest babes! To "win", but mainly to also be able to make our world the best possible world to be in, being masters of communication and persuasion and politics and advertising is MORE important than being technically proficient. After all, knowing all of the answers is completely useless if you can't persuade others to acknowledge and implement them!
I hope it is clear to someone that their attitude, thinking, and communication style and content are at odds with their apparent goals and all need to be radically altered.
Cheers,
Tom
As has been pointed out too many times, it is an extremely easy problem to solve, by using simple overkill.
But as has also been pointed out, the only reason for this thread is to investigate and characterize the region BETWEEN "overkill" and "insufficient".
Even if we imagined that overkill was the only solution, or even just the best solution, we would still want to ask how much overkill might be good-enough, and would want to provide the analysis to enable specific cost/size/weight/etc vs performance tradeoffs to be calculated.
In engineering anything, there is always a multi-dimensional trade-space, which must be characterized. And we always want to find out how to at least be ABLE to optimize performance, cost, and schedule, and properly evaluate and mitigate the risks associated with each step we might want or need to take.
And as far as I know, everyone here already believes, or at one time believed, that the bigger the reservoir capacitance was made, the better the PSU and the power amplifier would be. We are all way beyond that, or trying to go beyond that.
So, in the context of the later portion of this thread, it is unhelpful, and tiresome, for anyone to continue to simply repeat that assertion.
This is Nico's thread. He started it so he owns it. The only valid purpose of the discussion here is to be helpful in trying to find answers for what Nico asked about.
Personally, I have always tended to like overkill. But I believe that for any power amplifier linear PSU there is a large-enough but less-than-infinite rail-capacitance value such that increasing it will provide zero additional benefit, or a benefit that is so small that we would never consider increasing the capacitance beyond that point. How do we calculate that capacitance value, for a given amplifier and PSU circuit with given performance specifications?
(Hint for dunces: Saying "make it as big as possible" is NOT the type of answer we are seeking.)
Similarly, for any power amplifier with a linear PSU there is a lowest reservoir capacitance, below which the amplifier will not meet its specifications. How can we calculate the value of that capacitance?
Anywhere between those two capacitance values, the capacitance value used will affect the performance (and cost, etc) of the PSU and the amplifier, in multiple dimensions. How could we calculate those effects? After we can calculate them, then how could we calculate the minimum capacitance required to meet a given set of performance specifications? And how do we set the performance specifications?
Some of that is beyond what Nico asked for but I just wanted to mention a few things that any engineer, or engineering student, should be thinking about, in this context.
Finally, I have to say that it is a wonderful priviledge to be having this discussion, with so many intelligent people on board. Intelligence is interesting, eh? For example, in my life so far, I have noticed that many quite-intelligent people seem to think that they know more and know better than everyone else, about most things, or even about everything, for at least a part of their lives. Depending on their personality, their attitude can sometimes make them obnoxious and undesirable to have around. Sometimes, other people will not even want to bother with trying to set them straight, which just prolongs the problem. The most-intelligent ones eventually realize that almost everyone feels that way, either when they are young or at different times in their life, and then, having realized that, maybe they can learn better ways to view others, and see themselves in a more-proper perspective, and learn to communicate with and respect others at an acceptable level, and hopefully they find that their life is enriched and the direction they should take is more clear and they are more effective at whatever they try to do in life. I have found that even the stupidest people usually know something better than I do, and can often teach me something that I appreciate learning.
Engineers and scientists, in particular, have a tendency to not be very good at interpersonal communication and social interaction, maybe because they believe it is less important than being great at science and mathematics. We are dead wrong, about that, since it results in us letting others, who are probably not as smart, manage and lead and decide and implement everything, plus take most of the money and the hottest babes! To "win", but mainly to also be able to make our world the best possible world to be in, being masters of communication and persuasion and politics and advertising is MORE important than being technically proficient. After all, knowing all of the answers is completely useless if you can't persuade others to acknowledge and implement them!
I hope it is clear to someone that their attitude, thinking, and communication style and content are at odds with their apparent goals and all need to be radically altered.
Cheers,
Tom
Last edited:
knowing the limitations and capabilities of your power power supplies, there should not be issues...
many of these these so-called issues were borne out by the ignorance of the end users i would like to think....
no matter how good one engineers a product, at the hands of abusive users all bets are off...
many of these these so-called issues were borne out by the ignorance of the end users i would like to think....
no matter how good one engineers a product, at the hands of abusive users all bets are off...
Haha. Reminds me of an old saying, something like, "If you make it idiot-proof they'll invent a better idiot.".
you can say that again Tom.....
have we come up with rule of thumb?
sometime in the 80's Walter Jung came out with an article in Audio magazine about "building an energy storage bank", i do not know if a scan of that is available in the net or not but i still have those articles scanned on my hardrive.....the article extolled the virtue of having such capacitor bank...... i asked about it here: http://www.diyaudio.com/forums/solid-state/17220-walter-g-jungs-energy-storage-bank.html
about a few years ago i was able to pm him in another forum about his article, and he said that he has changed his opinion....
so yes, people do change opinions....
people seem to covet things that they do not have....and the search for the amp with the best specs goes on.....
have we come up with rule of thumb?
sometime in the 80's Walter Jung came out with an article in Audio magazine about "building an energy storage bank", i do not know if a scan of that is available in the net or not but i still have those articles scanned on my hardrive.....the article extolled the virtue of having such capacitor bank...... i asked about it here: http://www.diyaudio.com/forums/solid-state/17220-walter-g-jungs-energy-storage-bank.html
about a few years ago i was able to pm him in another forum about his article, and he said that he has changed his opinion....
so yes, people do change opinions....
people seem to covet things that they do not have....and the search for the amp with the best specs goes on.....
Better then simulation is reading a good book about this matter. Every basic electronic book describe this with advantage that there is a change that you understand what really happens, simulation doesnt.
Instead simulation reading a book is better. Every good basic electronic book describe this matter. And no not an Audio book. It gives the change you understand what really happens, simulations doesnt.
Power supply is not simple but needs good basic knowledge.
If HF is important for good Lf, why does a valve amp sounds better? The transformer hardly goes further then the audio band.
.
You are making generalist assumptions. Most valve amps I've heard sound distinctly worse than a good SS amp. But that is my preference/bias. You need to clarify yours.
Lot of people aggreed that valves sounds better so its a good idea that you clarify instead of me.
The only good sounding SS amps are the Stasis amps by Threshold and Nakamichi.
By the way these amps use simple power supplies and may be you can advice them.
Okay, you like all valve amps and only Stasis SS amps. That means we now understand your preconceptions a little better.
liching1952 that last comment is a typical trolling comment. Though I'm sure you already knew that.No I didn said I like all valve amps, there are also bad ones like Philips which is really junk. I like good valve amps like Quad, Beard, Conrad Johnson etc.
And yes I hear a lot of expensive SS amps, all sounds bad exept the Stasis.
Last edited:
May be thats the reason why your valve amp didnt sounds well. Its very hard and needs lot of knowledge to build a good audio transformer for a valve amp.I don't buy amps, expensive or otherwise, I build them.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.