advertisement


Some questions about power supplies

wulbert

wulbert

pfm Member
A few things I don't understand about power supplies for Hi Fi stuff:

1) Why does a "good" power supply make such a difference to audio gear? What is a good power supply doing?
2) Why are they so expensive? It's just a transformer right? A really old, well-understood, basic bit of electrical engineering. What is in them can can cost £700 and upwards to supply 12 volts, at a really low current?
3) Is there a source of commercial, bench-top power supplies e.g. for industry, laboratory use, schools etc that would provide something just as good (sounding) without the Hi Fi price tag?
4) Does a 13.8V or 15 volt power supply work OK on a Hi Fi component rated at 12volts DC? Does the supply's voltage rating need to exactly match the device's rating?
 
wulbert said:
A few things I don't understand about power supplies for Hi Fi stuff:

1) Why does a "good" power supply make such a difference to audio gear? What is a good power supply doing?
2) Why are they so expensive? It's just a transformer right? A really old, well-understood, basic bit of electrical engineering. What is in them can can cost £700 and upwards to supply 12 volts, at a really low current?
3) Is there a source of commercial, bench-top power supplies e.g. for industry, laboratory use, schools etc that would provide something just as good (sounding) without the Hi Fi price tag?
4) Does a 13.8V or 15 volt power supply work OK on a Hi Fi component rated at 12volts DC? Does the supply's voltage rating need to exactly match the device's rating?
Click to expand...
1) So long as it is a competent design, that doesn't spew noise into the equipment, nothing else
2) Marketing rip offs
3) Commercial supplies are sometimes electrically and audibly noisy. You have to pick and chose
4) Best to match to avoid over heating and linear regulators inside or triggering protection components. In particular "13.8V" supplies are emulating car power and under light load may give a LOT more.
 
Been doing a bit of digging and there's clearly quite a lot to this topic. My questions above are probably lazy. I'll do more reading and ask specific questions once I know more.

I'm looking for a decent power supply for a Aries Mini streamer and was a bit gutted to have missed out on an MRCU one on Ebay yesterday, when the train I was on went through a tunnel just as the auction was ending and my phone froze.
I don't want to waste my money so would like to understand a bit more about power supplies. However, I can see that just about any electronics, in a metal case, made in a democracy, conforming to safety laws, is going to cost £150 upwards just to cover time, materials and a small profit.
 
There are AC and DC power supplies. AC for things like filaments, DC for audio signals. AC is straightforward - the supply comes straight off the transformer. The voltage needs to be close to specifications, like 12.5V would be the max for a 12V supply. Certainly not 13.5V or 15V.

DC supplies are what feeds the audio circuit. The DC has to be as pure as possible - no AC in it, or only tiny amounts. AC causes hum and compromises the DC supply. So elaborate means are used to filter out the AC component
1. The rectifier comes first, can be valve or solid state. This gives you the "raw" DC supply
2. Next you have a filter which can be a choke, symbol L, or usually a capacitor, symbol C. There may be multiple capacitors and these have resistors between them to further filter the DC, symbol R. So a power supply cn be referred to as e.g. CLCLCRC. That would be 4 capacitors, 2 chokes and one resistor.
3. Valve amps have high voltage DC supplies, like 250v to 600v typically. So capacitors are typically a few hundred microfarads (uF). Solid state amps have low voltage DC supplies like 25V to 100V and have much larger capacitors - thousands of uFs.
4. There may also be a "voltage regulator" which further filters the DC and also gives a fixed voltage. This can be quite expensive in itself since it contains several parts on a PCB.

As you can see from the above, DC supplies are an art in themselves. The capacitors are usually cheap enough but the transformers and chokes and voltage regs can be very costly. All this to get the purest DC supply possible. Yes, it makes a difference to the sound.
 
What do you actually need? A 12V supply for a streamer? There are quite a few cheap(ish) supplies on AliExpress that can be modded to give very good results, usually they just need genuine caps and opamps. I did a BRZ one with Pana FM caps, now noiseless.
 
I think it can be hard to justify the cost of some power supplies.
There are a lot of simple design amendments that can make a big difference such as hard wiring rather than using PCBs.
A big one is introducing a little asymmetry to the connections between the main reservoir caps and the star point for connection to decoupling caps, feedback networks and the 0v line.
Doug Self measured a 100 fold increase in THD when all the earths join at the same point between the main reservoir caps and the capacitor bank 0v line.
Rod Elliot on his ESP website talks a fair bit of sense in his articles but not everyone would agree with some of his rather firm assertions.
Cheers Andy.
 
Dan K said:
What do you actually need? A 12V supply for a streamer? There are quite a few cheap(ish) supplies on AliExpress that can be modded to give very good results, usually they just need genuine caps and opamps. I did a BRZ one with Pana FM caps, now noiseless.
Click to expand...
On checking I see the Auralic Aries Mini streamer needs a 15 volt supply. (Apologies, I thought it was 12v). Electronic modifications are beyond my abilities.
 
andrewsutton said:
I think it can be hard to justify the cost of some power supplies.
There are a lot of simple design amendments that can make a big difference such as hard wiring rather than using PCBs.
A big one is introducing a little asymmetry to the connections between the main reservoir caps and the star point for connection to decoupling caps, feedback networks and the 0v line.
Doug Self measured a 100 fold increase in THD when all the earths join at the same point between the main reservoir caps and the capacitor bank 0v line.
Rod Elliot on his ESP website talks a fair bit of sense in his articles but not everyone would agree with some of his rather firm assertions.
Cheers Andy.
Click to expand...
If you have a link to an article (or a few articles) on those ideas, I'd be every-so-grateful!

Do they mainly apply to low current "precision" PSUs (streamers, amplifier front end), high current PSUs (power amplifier driver circuits), low-voltage (transistor) or high voltage (valve).

So does "it vary".
 
I’m not of the digerati but Google Rod Elliott ESP and there is a great resource. Doug Self did articles in audio/ electronic journals and gathered some of them into books. Again, Google will find them.
Rod is sometimes a bit pedantic, sometimes overly opinionated but some evidence based stuff that is really helpful.
If I said be selective I would go too far. Just weigh up the information. You won’t go far wrong.
The bits I have built based using his pcbs has all been very good.
I think I only question his position on earthing power supplies, as above and perhaps rectifiers. But he has had a few 35 amp bridge rectifiers with good heat sinking blow……
Cheers Andy.
 
"1) Why does a "good" power supply make such a difference to audio gear? What is a good power supply doing?"

If you look at the circuit of a power amp, the only thing between the output of the power supply (the terminals of the big smoothing capacitors) and the loudspeaker, is an output transistor. That transistor is controlled by everything else that goes in front of it and modulates the current coming from the smoothing caps to create the output signal. And everything else is also fed from those power rails - they need to be clean and stable.
 
Good evening all,

I'm watching this thread out of personal interest - I'm certainly far from qualified to comment from a technical viewpoint.

I comment based on what (I believe) my ears tell me.

The change from toroidal to EI transformers was, for me, clearly audible - there were detail changes. You know them when you hear them because they immediately strike you when listening to familiar pieces and something comes at you that is different.

When I conducted my simple test with the CAP6 fitted with the SMD FWBR I only tried the test with the modified PSU's in the one amplifier. The two amplifiers were fitted with identical SE amplifier boards, BUT, and it may be a big but, one amplifier had 700VA of transformer and the other had 1000VA of transformer - so was it a completely fair test?

I note Doug Self comments on the use of snubbing capacitors to help suppress RF emissions from diode switching in amplifier power supplies, it was suggested to me that I should avoid doing this. He also mentions fitting a capacitor across the live/ neutral lines as near to mains inlet as possible.

Which is right or wrong? I haven't a clue.

Regards

Richard
 
andrewsutton said:
I’m not of the digerati but Google Rod Elliott ESP and there is a great resource. Doug Self did articles in audio/ electronic journals and gathered some of them into books. Again, Google will find them.
Rod is sometimes a bit pedantic, sometimes overly opinionated but some evidence based stuff that is really helpful.
If I said be selective I would go too far. Just weigh up the information. You won’t go far wrong.
The bits I have built based using his pcbs has all been very good.
I think I only question his position on earthing power supplies, as above and perhaps rectifiers. But he has had a few 35 amp bridge rectifiers with good heat sinking blow……
Cheers Andy.
Click to expand...
Thank you. That's majestically unhelpful.
 
Richard Lines said:
BUT, and it may be a big but, one amplifier had 700VA of transformer and the other had 1000VA of transformer - so was it a completely fair test?
Click to expand...
The question that should be asked - is it a 230V transformer or a real 240V transformer? Many "230V" examples struggle at high mains in the UK of up to 253V. If the transformer goes into saturation, you get buzz, large input currents and heat.
 
wulbert said:
A few things I don't understand about power supplies for Hi Fi stuff:

1) Why does a "good" power supply make such a difference to audio gear? What is a good power supply doing?
2) Why are they so expensive? It's just a transformer right? A really old, well-understood, basic bit of electrical engineering. What is in them can can cost £700 and upwards to supply 12 volts, at a really low current?
3) Is there a source of commercial, bench-top power supplies e.g. for industry, laboratory use, schools etc that would provide something just as good (sounding) without the Hi Fi price tag?
4) Does a 13.8V or 15 volt power supply work OK on a Hi Fi component rated at 12volts DC? Does the supply's voltage rating need to exactly match the device's rating?
Click to expand...
2. absolutely! you can make it at home for nothing, all you need is a small aubergine!
3. good regulated bench power supplies are usually more expensive than hifi ones
 
It's not what a good power supply does so much as what a poor one fails to do. Poor ones vary under load and the no load voltage can be too high and fry 12V stuff.
however there's no reason why they should cost thousands.
 
wulbert said:
A few things I don't understand about power supplies for Hi Fi stuff:

1) Why does a "good" power supply make such a difference to audio gear? What is a good power supply doing?
2) Why are they so expensive? It's just a transformer right? A really old, well-understood, basic bit of electrical engineering. What is in them can can cost £700 and upwards to supply 12 volts, at a really low current?
3) Is there a source of commercial, bench-top power supplies e.g. for industry, laboratory use, schools etc that would provide something just as good (sounding) without the Hi Fi price tag?
4) Does a 13.8V or 15 volt power supply work OK on a Hi Fi component rated at 12volts DC? Does the supply's voltage rating need to exactly match the device's rating?
Click to expand...

1) Assuming the audio gear is well designed i.e. has good PSRR then the PS on the front end shouldn't theoretically make any difference. However it seems that they do actually make a difference to the sound. There are possible reasons why PSUs with similar specs might sound different - too complex and system dependent to discuss here.

2) Because people will pay!!
No it's not just a transformer. Most DC supplies will include regulation and maybe some filtering.

3) Plenty of commercial supplies out there, usually SMPS. How they affect the sound is a lottery.

4) Depends on the hifi component. If it has internal regulation (almost always) with sufficiently rated parts and can dissipate the extra heat due to the extra power loss, then - no problem.
 
You must log in or register to reply here.


advertisement


Back
Top