That's more faith than science I suspect.
The power supply will respond to the load drawn by the audio.
The audio will be modulating the current waveforms in the power supply.
The only way that's not going to happen is if you have some freakish class A amplifier which draws the same load irrespective of programme content.
There is always the possibility of expectation bias.
Someone brought up an excellent case of this when saying that if they see on a chart that an amplifier has faithfully reproduced the test signal that the amplifier is as good as it gets. I do not see how this could possibly be any less powerful than one expecting better results from a brute force supply versus a regulated one.
I sure wish it was that easy. if it was one would think the market would be flooded with many excellent sounding amplifiers that would all, serendipitously, have the same excellent sound. But we know amplifiers do not sound the same even with specifications with variances
akin to what happens when you run the same test twice.
One would think/hope that this alone would inspire skepticism in using nothing more than measurements. To me it shows how an amplifier performs with that waveform. A waveform that has nothing to do with how music is processed by an amplifier. Sure these measurements are useful, mainly for finding gross errors but I find it hard to believe that anyone who actually listens to music cannot hear the differences between amplifiers.
I agree noise is an important thing to remove/improve. Many of the things we do just move the noise somewhere else. As in the case of Sid Smith electing to use a half wave rectifier in the 7 preamp. Certainly he did not have to worry about the cost of another rectifier. I have been listening to the SALAS with an CLCLCLC supply with a half wave rectifier since you cannot very effectively use a choke with one. I know that PSUD is a modeler like any other model - who knows how accurate their noise figures are absolutely? But one can find (and of course it is pure chance if these values really work in the real world) configurations that yield microsonic noise figures even with the half wave.
I have using this for the last six months. I hear no more noise through the (JBL 2441) than i did before. The bass is every bit as present. What is missing this noise I found akin to digital noise - not quite as prevalent but the same kind of mechanism - it is not a click or a pop, it is the hazy stuff that clouds the window. But I swear there is less of it with the half wave and I can only figure Mr. Smith heard the same thing.
Worth a try for those so inclined. It would not bother me at all for someone to write they did not find anything good at all about it after trying it. That's what these forums are all about, isn't it?
Someone brought up an excellent case of this when saying that if they see on a chart that an amplifier has faithfully reproduced the test signal that the amplifier is as good as it gets. I do not see how this could possibly be any less powerful than one expecting better results from a brute force supply versus a regulated one.
I sure wish it was that easy. if it was one would think the market would be flooded with many excellent sounding amplifiers that would all, serendipitously, have the same excellent sound. But we know amplifiers do not sound the same even with specifications with variances
akin to what happens when you run the same test twice.
One would think/hope that this alone would inspire skepticism in using nothing more than measurements. To me it shows how an amplifier performs with that waveform. A waveform that has nothing to do with how music is processed by an amplifier. Sure these measurements are useful, mainly for finding gross errors but I find it hard to believe that anyone who actually listens to music cannot hear the differences between amplifiers.
I agree noise is an important thing to remove/improve. Many of the things we do just move the noise somewhere else. As in the case of Sid Smith electing to use a half wave rectifier in the 7 preamp. Certainly he did not have to worry about the cost of another rectifier. I have been listening to the SALAS with an CLCLCLC supply with a half wave rectifier since you cannot very effectively use a choke with one. I know that PSUD is a modeler like any other model - who knows how accurate their noise figures are absolutely? But one can find (and of course it is pure chance if these values really work in the real world) configurations that yield microsonic noise figures even with the half wave.
I have using this for the last six months. I hear no more noise through the (JBL 2441) than i did before. The bass is every bit as present. What is missing this noise I found akin to digital noise - not quite as prevalent but the same kind of mechanism - it is not a click or a pop, it is the hazy stuff that clouds the window. But I swear there is less of it with the half wave and I can only figure Mr. Smith heard the same thing.
Worth a try for those so inclined. It would not bother me at all for someone to write they did not find anything good at all about it after trying it. That's what these forums are all about, isn't it?
when possible/applicable , I always prefer proper active ( read - having error amplifier, read - mine) Shunt Reg
when not applicable , next best thing for me is unregulated brute force (more copper/iron , the merrier) supply
series regs are reserved (in my universe) for non-audio duties ...... pure digital being slight exception
when not applicable , next best thing for me is unregulated brute force (more copper/iron , the merrier) supply
series regs are reserved (in my universe) for non-audio duties ...... pure digital being slight exception
This thread is about regulators or no regulators in small signal circuits. Which tend to draw constant current. Certainly with a power amp the situation is dramatically different and even a "normal" class A amp does not have constant current draw and there will be audio frequencies in the power transformer.
Do you find so many CL stages really essential if you use a regulator? TBH i have never tried more than a single choke + possibly a CM choke in this arrangement. And yes, a choke input filter is not a good idea for half wave rectification
Member
Joined 2009
Paid Member
Ripple is not really an issue. It is easy and cheap to go bonkers with capacitance. Even with good sounding capacitors. Good sounding layered chokes are neither cheap, nor easy to squeeze in a box. I would need some persuasion before considering three of those per preamp rail.
It is interesting how different the sound with half wave rectification is. Even compared to a full wave centre tapped. I wonder if dc magnetization of the transformer core plays any role in this, although in my experiments i make an effort to equalize this with a reversely biased second winding.
No. Amps do not draw audio current from the mains. They may (if Class B) draw mains current whose envelope varies with the signal envelope, but that is quite different.Chris888 said:
Yes. This is why the output impedance of a regulator matters, and why slapping on a big cap can make things worse.
No, it is science.
As I said, it is the signal envelope which gets through - not the signal.
Class A does not draw constant current from the PSU. On the contrary, it imposes the full signal as current draw. Class A draws constant average current, unlike Class B.
With the exception of the consant current sourced SE class A topology.
Perhaps this is also a reason for the sonic advantage of ccs-ed shunt regulators, as these make the current draw of even a class B amp appear constant.
Now, if anyone hopes that these topologies are somehow exempt from the effects of power cables...
Last edited:
They are, as long as you have designed a competent power supply / regulator stage. If you can hear the effects of changing a power cord, your power supply design isn't adequate.
I have experience with this issue, and a friend who believes in power cords making a difference. He also indicates that he can't hear the difference with my power supply designs. Science rules.
-Chris
I said current at audio frequencies. That's not necessarily current which is linearly related to the audio signal, but current at frequencies in the audio band.
The fact that the process is not linear is probably not a helpful thing.
Possibly we are off on a tangent here.
What matters to me is knowing that the rails will never be completely clean, so I want to firstly design to avoid that influencing the signal path, and secondly measure that influence.
Yes, but in some other Universe
Imho a cable invariant mains powered amp, preamp, digital source does not exist. Components and systems with extreme problems, capable of masking such effects sadly do exist.
I have taken existing designs, improved them to a point where the AC cord has no effect when before they did. The other massive improvement has been in noise floor. In other words, a more perfect regulator and power supply.
One of the first things I do is to reduce filter capacitor size to sane values in keeping with the current load. It's not me claiming that I can make equipment power cord invariant, it's my friend and customers. I never assess my own work unless I find fault with it.
-Chris
One of the first things I do is to reduce filter capacitor size to sane values in keeping with the current load. It's not me claiming that I can make equipment power cord invariant, it's my friend and customers. I never assess my own work unless I find fault with it.
-Chris
OK. Amps do not take current at audio frequency from the mains. Is that better?Chris888 said:
It is conceivable that an amp with a poorly designed (but possibly expensive 'audiophile') PSU and poor PSRR in the audio circuitry and perhaps poor grounding arrangements could be sensitive to the mains cable. It is conceivable that some may misinterpret this as a sign of fine discrimination in the equipment, when in fact it is a sign of poor design and construction.
I can see that an amplifier can put a signal on its mains cable. I can see that a signal on a mains cable can get back into the signal path of the amplifier. So, I suppose it is possible for the characteristic of a mains cable to influence the output of the amplifer.
But in a normal house, the mains is polluted with all sorts of dross anyway. It would be a bad amplifier that was susceptible to this kind of effect. There are lots of bad amplifiers about. I had one which chirped just before my mobile phone rang. Somehow susceptible to the phone's transmission.
We can change mains leads, but they are only plugged into the wall, where we have less control. There might be a metre of mains lead, then 20 metres of wiring back to the fuse box, with all sorts of stuff connected. Of course in a lab, we might have clean mains. Or not.
Still, we have it easier than in the automotive world. The 12V in cars can be pretty messy. Military vehicles 28V with spikes of 120V or something in either direction! But they are less picky about how their Wagner sounds.
But in a normal house, the mains is polluted with all sorts of dross anyway. It would be a bad amplifier that was susceptible to this kind of effect. There are lots of bad amplifiers about. I had one which chirped just before my mobile phone rang. Somehow susceptible to the phone's transmission.
We can change mains leads, but they are only plugged into the wall, where we have less control. There might be a metre of mains lead, then 20 metres of wiring back to the fuse box, with all sorts of stuff connected. Of course in a lab, we might have clean mains. Or not.
Still, we have it easier than in the automotive world. The 12V in cars can be pretty messy. Military vehicles 28V with spikes of 120V or something in either direction! But they are less picky about how their Wagner sounds.
Not uncommon these days with lots of rf in the average home, better rfi screening/suppression needed
So a decent audio amp should have the same noise performance in it’s power supply as a full on lab power supply?
Yeah right...
And you can’t (or shouldn’t) really assume that the worst that you have experienced on an ac line is the same as what someone else is dealing with.
I wouldn’t say that just because you can pick up noise from the line that a piece of equipment is poorly designed, but that in that instance that something else was needed.
Yeah right...
And you can’t (or shouldn’t) really assume that the worst that you have experienced on an ac line is the same as what someone else is dealing with.
I wouldn’t say that just because you can pick up noise from the line that a piece of equipment is poorly designed, but that in that instance that something else was needed.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.