How important is the PSU in your amp?

[Thompsonuxb]

Let me just say if I'm winding people up I'm not doing it intentionally......

So how important is the Power supply unit in you amp, its not a tec question as such - but of all the things you can change within your amp would it have the most dramatic affect on the actual sound an amp could produce if changed.

I know Cryus are one of thwe few companys that offer this as an 'upgrade' path but changing the PSU does it technically make your amp a 'different' amp or not?

This is a sort of follow on from the 'Do amps sound the same' thread, after what Bob did, just curious ref your opinions.

 

[simonlewis]

Well i thought the answer would be obvious, i'm guessing extremly important, you don't want some cheap tat from china that would burn out after a couple of weeks use.

 

[abacus]

It is important for stability and performance.

Upgrade:

On its own the differences would not be that great as the rest of the components were chosen and built around the existing power supply.

If the rest of the components had been designed to accept a better power supply (Such as Cyrus, Naim etc.) then yes, as it would be like moving up to the next model (Or higher) in the range.

Hope this helps

Bill

 

[Overdose]

If you need to upgrade the power supply in whatever it is you have, then I would suggest that you bought the wrong piece of equipment in the first place.

 

[Sospri]

If you need to upgrade the power supply in whatever it is you have, then I would suggest that you bought the wrong piece of equipment in the first place.

What a silly post,

Or was your first car an Aston Martin then?

 

[Overdose]

If you need to upgrade the power supply in whatever it is you have, then I would suggest that you bought the wrong piece of equipment in the first place.

What a silly post,

Or was your first car an Aston Martin then?

My first car was a MK III Ford Cortina 2000e with vinyl roof. No amount of power plant upgrade was ever going to make it an Aston Martin.

Your analogy has illustrated my point precisely.

 

[davedotco]

I recall talking to a designer of very well regarded british amplifiers about the importance of the power supply. He said it was absolutely pivotal, and described it something like this (I'm relying on memory here, he was probably more succinct):

"Most people think that an amplifier takes in a small signal at the input and progressively makes that signal bigger, stage by stage until it appears ready to drive your steakers at the amps output terminals."

"Nothing could be further than the truth", he said, "it would be closer to the truth to think of an amplifier being a device that switches the mains (power) on and off in time with the music".

 

[davedotco]

I recall talking to a designer of very well regarded british amplifiers about the importance of the power supply. He said it was absolutely pivotal, and described it something like this (I'm relying on memory here, he was probably more succinct):

"Most people think that an amplifier takes in a small signal at the input and progressively makes that signal bigger, stage by stage until it appears ready to drive your speakers at the amps output terminals."

"Nothing could be further than the truth", he said, "it would be closer to the truth to think of an amplifier being a device that switches the mains (power) on and off in time with the music".

 

[Electro]

In my opinion the PSU of an amplifier is the single most important parts of an amplifier, especially a power amplifier and even more so with direct coupled amps like mine .

In fact bad or under spec power supplies are one of the main reason that so many amps can never reach their full potential IMO .

 

[DocG]

How can you read the quality of the PSU in the amp's specs? Is it in doubling the wattage when impedance is halved?

To put it differently: when driving current-hungry speakers (say Magnepans, with a flat 4 Ohm impedance), are you better off with an amp that delivers 150 W into 8 Ohm, 200 W into 4 Ohm and 240 W into 2 Ohm or with one that pumps 30 W into 8, 60 W into 4 and 120 W into 2 Ohm?

(PS: I'm no electrical engineer, as you might have guessed...)

 

[abacus]

I recall talking to a designer of very well regarded british amplifiers about the importance of the power supply. He said it was absolutely pivotal, and described it something like this (I'm relying on memory here, he was probably more succinct):

"Most people think that an amplifier takes in a small signal at the input and progressively makes that signal bigger, stage by stage until it appears ready to drive your steakers at the amps output terminals."

"Nothing could be further than the truth", he said, "it would be closer to the truth to think of an amplifier being a device that switches the mains (power) on and off in time with the music".

You have misunderstood what he was saying there; an amplifier increases a low level signal (Music for example) to a high level signal, and the power supply supplies the means for the amplifier to achieve it.

Bill

 

[davedotco]

How can you read the quality of the PSU in the amp's specs? Is it in doubling the wattage when impedance is halved?

To put it differently: when driving current-hungry speakers (say Magnepans, with a flat 4 Ohm impedance), are you better off with an amp that delivers 150 W into 8 Ohm, 200 W into 4 Ohm and 240 W into 2 Ohm or with one that pumps 30 W into 8, 60 W into 4 and 120 W into 2 Ohm?

(PS: I'm no electrical engineer, as you might have guessed...)

But are you a real doctor, I think we should be told....... :?

The Magnepans are unusual in that there impedence is largely resistive with minimal capacitance or inductance, so in this case the answer is simple, more power is better all other things being equal.

Some amplifiers, early Krell KSA50s for example, were rated at 50w, 100w and 200w into 8, 4 and 2 ohms respectively. In reality it produced nearly 100w into 8 ohm load so the real power did not double as the spec suggested.

The way an amplifier drives a real loudpeaker is complex too, some get upset at a capacitive load (electrstatics for example) others don't like highly inductive loads, some take it all in their stride but even they find it difficult to double the measured power in the real world.

 

[abacus]

How can you read the quality of the PSU in the amp's specs? Is it in doubling the wattage when impedance is halved?

To put it differently: when driving current-hungry speakers (say Magnepans, with a flat 4 Ohm impedance), are you better off with an amp that delivers 150 W into 8 Ohm, 200 W into 4 Ohm and 240 W into 2 Ohm or with one that pumps 30 W into 8, 60 W into 4 and 120 W into 2 Ohm?

(PS: I'm no electrical engineer, as you might have guessed...)

There are many factors that affect how an amplifier does its job, and so cannot be split into an easy and simple formula.

The formula that you mention would only be applicable if the speakers was purely resistive, which most speakers aren’t thus adding another layer of complexity.

Hope this helps

Bill

 

[davedotco]

I recall talking to a designer of very well regarded british amplifiers about the importance of the power supply. He said it was absolutely pivotal, and described it something like this (I'm relying on memory here, he was probably more succinct):

"Most people think that an amplifier takes in a small signal at the input and progressively makes that signal bigger, stage by stage until it appears ready to drive your steakers at the amps output terminals."

"Nothing could be further than the truth", he said, "it would be closer to the truth to think of an amplifier being a device that switches the mains (power) on and off in time with the music".

You have misunderstood what he was saying there; an amplifier increases a low level signal (Music for example) to a high level signal, and the power supply supplies the means for the amplifier to achieve it.

Bill

No.

He may not have been correct, I am not an electronic engineer so I don't know, but in effect what he was saying was that the power supply produces the +ve and -ve output rail voltage and the input voltage is used, virtually as a control voltage, to modulate the output between these rails.

This may be complete b*****ks in engineering terms but that is what he was saying.

 

[davedotco]

How can you read the quality of the PSU in the amp's specs? Is it in doubling the wattage when impedance is halved?

To put it differently: when driving current-hungry speakers (say Magnepans, with a flat 4 Ohm impedance), are you better off with an amp that delivers 150 W into 8 Ohm, 200 W into 4 Ohm and 240 W into 2 Ohm or with one that pumps 30 W into 8, 60 W into 4 and 120 W into 2 Ohm?

(PS: I'm no electrical engineer, as you might have guessed...)

There are many factors that affect how an amplifier does its job, and so cannot be split into an easy and simple formula.

The formula that you mention would only be applicable if the speakers was purely resistive, which most speakers aren’t thus adding another layer of complexity.

Hope this helps

Bill

Full range Magnepans are almost entirely resistive, which is what made the post interesting.

 

[DocG]

How can you read the quality of the PSU in the amp's specs? Is it in doubling the wattage when impedance is halved?

To put it differently: when driving current-hungry speakers (say Magnepans, with a flat 4 Ohm impedance), are you better off with an amp that delivers 150 W into 8 Ohm, 200 W into 4 Ohm and 240 W into 2 Ohm or with one that pumps 30 W into 8, 60 W into 4 and 120 W into 2 Ohm?

The Magnepans are unusual in that there impedence is largely resistive with minimal capacitance or inductance, so in this case the answer is simple, more power is better all other things being equal.

In my example, do I go for the first or the second amplifier? (Is doubling watts better? Or just the more watts into 4 Ohm, the better?)

 

[Frank Harvey]

It all starts with the power supply...

 

[Al ears]

It does indeed. It does not really matter if you have tubes or solid state stuff after that in the chain. Neither are going to make a significant difference if the power supply section is rubbish.

This is why most of the well respected manufacturers of amplifiers wind their own power supply transformers.

Tom Evans Audio being just one of those.

 

[busb]

How can you read the quality of the PSU in the amp's specs? Is it in doubling the wattage when impedance is halved?

To put it differently: when driving current-hungry speakers (say Magnepans, with a flat 4 Ohm impedance), are you better off with an amp that delivers 150 W into 8 Ohm, 200 W into 4 Ohm and 240 W into 2 Ohm or with one that pumps 30 W into 8, 60 W into 4 and 120 W into 2 Ohm?

The Magnepans are unusual in that there impedence is largely resistive with minimal capacitance or inductance, so in this case the answer is simple, more power is better all other things being equal.

In my example, do I go for the first or the second amplifier? (Is doubling watts better? Or just the more watts into 4 Ohm, the better?)

The easier (more resisitive) the load that is also close to 8Ohms at all frequencies - it will be the former where its poor(er) damping factor (higher output Z) won't matter too much.

 

[proffski]

Let me just say if I'm winding people up I'm not doing it intentionally...... So how important is the Power supply unit in you amp, its not a tec question as such - but of all the things you can change within your amp would it have the most dramatic affect on the actual sound an amp could produce if changed. I know Cryus are one of thwe few companys that offer this as an 'upgrade' path but changing the PSU does it technically make your amp a 'different' amp or not? This is a sort of follow on from the 'Do amps sound the same' thread, after what Bob did, just curious ref your opinions.

Extremely important! It does not take much to figure out when you look at the topology of an output stage that the power supply is all important. I have seen AV amplifiers rated at 100WPC and I would not have put some of the power supplies in a stereo amplifier rated at 35WPC.

 

[davedotco]

How can you read the quality of the PSU in the amp's specs? Is it in doubling the wattage when impedance is halved?

To put it differently: when driving current-hungry speakers (say Magnepans, with a flat 4 Ohm impedance), are you better off with an amp that delivers 150 W into 8 Ohm, 200 W into 4 Ohm and 240 W into 2 Ohm or with one that pumps 30 W into 8, 60 W into 4 and 120 W into 2 Ohm?

The Magnepans are unusual in that there impedence is largely resistive with minimal capacitance or inductance, so in this case the answer is simple, more power is better all other things being equal.

In my example, do I go for the first or the second amplifier? (Is doubling watts better? Or just the more watts into 4 Ohm, the better?)

The easier (more resisitive) the load that is also close to 8Ohms at all frequencies - it will be the former where its poor(er) damping factor (higher output Z) won't matter too much.

How did you deduce that the 'former' amplifier had a poorer damping factor?

 

[andyjm]

Ignoring the comments about winding your own mains transformers (why on earth would you?), the PSU question is quite interesting.

For low power sections (pre amp, DAC, early stages of a power amp etc) a stabilised supply makes very good sense. There are many ways to do this and one or two designs which have a cult following for low noise and high stabilty, google 'Jung super regulator' for a popular design.

Strangely, none involve winding your own transformer (or braiding your own mains cable)...

For the power stages of an amp, it gets more complicated. Ideally a stabilsed supply would make obvious sense, but it needs to react quickly to transients (it almost needs to be an amp in itself), and high wattage linear stabilised supplies are a performance to cool. Most (I believe) amps use simple bridge rectifier / large storage caps for their output stage power supplies. This is actually where a switch mode supply, or class D with feedback come into their own (interestingly, an amp with switch mode supply isn't that different from class D when you think about it)

So yes, PSUs do matter, but a big transformer, big diodes and big caps get you to where you need to get for output stages, and a decent regulated supply for the low level stages. No magic, just basic engineering.

 

[Nelly]

Now this interest me because im going to get a tripath thats rated at 160w+160w but that is only if you can get a 36v 5amp power supply because it come with a 26v 4amp supply now is this rue for all amps if you up the power supply do you up the watts of the amp ?

 

[busb]

How can you read the quality of the PSU in the amp's specs? Is it in doubling the wattage when impedance is halved?

To put it differently: when driving current-hungry speakers (say Magnepans, with a flat 4 Ohm impedance), are you better off with an amp that delivers 150 W into 8 Ohm, 200 W into 4 Ohm and 240 W into 2 Ohm or with one that pumps 30 W into 8, 60 W into 4 and 120 W into 2 Ohm?

The Magnepans are unusual in that there impedence is largely resistive with minimal capacitance or inductance, so in this case the answer is simple, more power is better all other things being equal.

In my example, do I go for the first or the second amplifier? (Is doubling watts better? Or just the more watts into 4 Ohm, the better?)

The easier (more resisitive) the load that is also close to 8Ohms at all frequencies - it will be the former where its poor(er) damping factor (higher output Z) won't matter too much.

How did you deduce that the 'former' amplifier had a poorer damping factor?

150W into 8Ohms

200W into 4Ohms

Without resorting to calculating the output Z looks pretty average but probably uses less than average negative feedback. IMO these figures would suit the mentioned speakers better at higher volumes

Whereas:

30W into 8Ohms

60W into 4Ohms

Implies extremely low output Z if you get a virtual doubling of power with half the load impedance with likely high levels of negative feedback to keep the damping factor so high. DF is the ratio of Z load divided Z source where Z is the impedance. The impedance "seen" by the amplifier will be a complex one which will vary with frequency, cable chraracteristics including length, Xover components & voice coil temperature so will vary from low to high power. An amp with such characteristics would suit difficult loads if not to particularly high volume.

 

[davedotco]

How can you read the quality of the PSU in the amp's specs? Is it in doubling the wattage when impedance is halved?

To put it differently: when driving current-hungry speakers (say Magnepans, with a flat 4 Ohm impedance), are you better off with an amp that delivers 150 W into 8 Ohm, 200 W into 4 Ohm and 240 W into 2 Ohm or with one that pumps 30 W into 8, 60 W into 4 and 120 W into 2 Ohm?

The Magnepans are unusual in that there impedence is largely resistive with minimal capacitance or inductance, so in this case the answer is simple, more power is better all other things being equal.

In my example, do I go for the first or the second amplifier? (Is doubling watts better? Or just the more watts into 4 Ohm, the better?)

The easier (more resisitive) the load that is also close to 8Ohms at all frequencies - it will be the former where its poor(er) damping factor (higher output Z) won't matter too much.

How did you deduce that the 'former' amplifier had a poorer damping factor?

150W into 8Ohms

200W into 4Ohms

Without resorting to calculating the output Z looks pretty average but probably uses less than average negative feedback. IMO these figures would suit the mentioned speakers better at higher volumes

Whereas:

30W into 8Ohms

60W into 4Ohms

Implies extremely low output Z if you get a virtual doubling of power with half the load impedance with likely high levels of negative feedback to keep the damping factor so high. DF is the ratio of Z load divided Z source where Z is the impedance. The impedance "seen" by the amplifier will be a complex one which will vary with frequency, cable chraracteristics including length, Xover components & voice coil temperature so will vary from low to high power. An amp with such characteristics would suit difficult loads if not to particularly high volume.

Interesting, whilst I totally understand what you are saying, I would have thought that these variations would have been pretty small compared to the more obvious limitations of power supplies, output devices and the rest. You live and learn...... :?