Hi ! 
i have a question related to a story.
A friend of mine owns two small but nice mono amps, Albarry m408.
Then he tried the integrated from the same brand that is exactly the same design in one box but with just one transformer and a pair of ps caps, the same parts that can be found in each monos.
Driving a pair of full range speakers there was absolutely no story.
The two monos sounded so much more convincing that he was even surprised given the same design.
Of course the ps power was doubled.
So my question is ... if i take a small but nice integrated of which i like the sound a lot and i double both transformer VA and ps caps uF can i expect a similarly remarkable improvement in sound ?
Moreover in the case above, a transformer double the size and 4 caps would have given to the integrated the same sound of the two monos ?
thanks and regards, gino
i have a question related to a story.
A friend of mine owns two small but nice mono amps, Albarry m408.
Then he tried the integrated from the same brand that is exactly the same design in one box but with just one transformer and a pair of ps caps, the same parts that can be found in each monos.
Driving a pair of full range speakers there was absolutely no story.
The two monos sounded so much more convincing that he was even surprised given the same design.
Of course the ps power was doubled.
So my question is ... if i take a small but nice integrated of which i like the sound a lot and i double both transformer VA and ps caps uF can i expect a similarly remarkable improvement in sound ?
Moreover in the case above, a transformer double the size and 4 caps would have given to the integrated the same sound of the two monos ?
thanks and regards, gino
Last edited:
You can expect an improvement in sound yeah, how 'remarkable' remains to be seen. But when adding capacitance why not add some series L too to get improved ripple filtering? And instead of merely doubling the number of existing capacitors, design the new supply to give the lowest possible impedance to the output stage.
Hi and thanks a lot for the very valuable reply.
My feeling is that there is a huge attention and debate about circuits design and of course this is fine, but much less attention to power supplies.
But it seems that power supplies are not trendy ...
No differences about mains transformers, one cap is like the other ...
For instance i see caps same uF and WV but very different sizes ... do they have really the same performance ? i do not think so.
The best circuit without a strong power supply that supports it has no chance to deliver the sound that it should.
especially with difficult loads like more recent speakers.
On the contrary even just decent circuits but with a powerful power supply can sing nicely.
And also all this discussion about oversized power supplies ... why ? what is the problem of having a big power supply aside weight and cost ?
sorry but series L do not increase impedance ?
this gets complicated. Replace a transformer and a pair of caps can be done with some practice. Re-design the power supply is much more challenging.
To end i wanted to say that i truly think that very often a weak power supply is indeed the Achille's heel of many commercial amps, that instead sound quite nice and clean.
I had even a Denon entry level receiver .. the sound was extremely clean but flat and with a weak bass.
Then i look inside and i saw 2 small ps caps of doubtful quality. And the mains transformer did not impress me either.
How would have sounded that amp with a "decent" power supply ? like 600VA of transformer and 40.000uF of high grade caps so to speak ?
Moreover i see that amps power consumption in amps is going down lately. Even some TOTL models !!!
No surprise that they fail to convince when tested.
They are extremely nice to see ... not impressive at all for sound, especially with demanding speakers.
Thanks again, gino
Last edited:
'Oversized' in relation to what? If a large power supply is required to get sufficient SQ, then a large PSU is not 'oversized', its sized for the job in hand, which is customer satisfaction. So seems to me some doublethink going on in the use of the word 'oversized' - oversized relative to the power output spec is not 'oversized' in the subjective SQ sense.
Sure it does - mains power supplies though have two requirements which to some extent conflict - low impedance to the OPS and good ripple rejection. A series L does increase the impedance but improves the ripple filtering.
'Weak' in relation to being able to deliver the specified power? I think not, 'weak' only in the sense of not delivering all the dynamics that the active components might be capable of. But then I'd suggest the weakness isn't really in the power supply, rather in inadequate PSRR of the electronics. I'm of the view that if the amp's based on discretes its probably cheaper (and hence more prudent engineering) to improve the PSRR by 6dB than double the trafo and cap sizes.
I think i see your point and i agree. If it works and sounds fine is well sized.
This i am not sure i understand.
I have never had any issue of noise at all, unless there was a fault in the amp.
The amps are all usually very silent.
But if i use them with a demanding full range speakers they drive the speakers differently.
Drums are an acid test for what i mean but also organ music like "Toccata e fuga" by Bach and piano.
Some higher W amps pushes the woofers less than lower wattage amps !!!
I got more music from a 50W than from a 100W
The 50W had a very big power supply ... someone would call it oversized
Thanks again, gino
'Getting more music' presumably means you get greater amounts of sound that sounds 'musical' from the 50W than from the 100W amp. Which to me says the 100W one was introducing audible artifacts (noise modulation) at a lower volume level than the 50W one.
I'm not clear on the significance of your bolded text above, seeing as I wasn't writing about noise. But nevertheless the problem of 'weak' amps is indeed a noise problem (not in the sense of the usual signal/noise measurements where the noise is quantified in the absence of signal) which can be addressed in either of the two ways I outlined - a bigger PSU or better PSRR.
I'm not clear on the significance of your bolded text above, seeing as I wasn't writing about noise. But nevertheless the problem of 'weak' amps is indeed a noise problem (not in the sense of the usual signal/noise measurements where the noise is quantified in the absence of signal) which can be addressed in either of the two ways I outlined - a bigger PSU or better PSRR.
You're right, and for example the HK "twin power" receivers do have very good bass.
They have two power transformers, one per channel, with 4700uF on each rail, per channel.
Hi again and sorry i think i have expressed myself badly.
My friend (a good and experienced listener) told me that with the 50W amp the bass was more kicking, powerful and controlled.
It seemed like the less wattage amp was driving better the woofers.
The speakers were an old pair of AR 9, mighty speakers indeed.
But they ask for current.
With the 100W the woofer just grumbled ... moaned.
Clear evidence that the real power was not arriving to the coils.
I have another very rude experiment to tell.
Once I made a dummy load with power resistors ... 200W/5 ohm.
Than i connect the load to some amps i had at hand.
With one amp the load remained at ambient temperature but the fins of the amp got extremely hot
Another amp warmed up the load and became warm.
The last one warmed up but made the load very hot.
Incidentally the last amp was the one with the better bass and also sound, without any doubt.
The first amp instead had a joke of a bass.
What i mean is that an ampermeter in series with the load could tell many many things and indeed i have some experiments in mind to assess the behaviour in current of an amp
given that you were talking about Power Supply Rejection Ratio i thought you were addressing noise. I mistook the issue.
here is where i do not agree.
The circuit for me is like a water tap. If you do not have pressure upstream you can fully open the tap but the outcome will be weak.
The "pressure" is the current coming from the power supply and the transformer, very little talked about, can be the first and decisive bottle-neck for current.
I strongly need an high current amp. Not many watts instead.
20-30 can be enough ... something like the First Watt or similar.
And i guess it will be DIY .. in the end.
Thanks a lot again.
Kind regards, gino
Hi and thanks a lot.
This is exactly what i am looking for. Maybe not a receiver but just the amp.
I will look for them. Maybe the sound also nice.
Thanks again, gino
Attachments
I think you may well be conflating how you perceive the sound with the engineering principles. The power was still arriving at the coils (or the amp wasn't in reality a 100W amp at all, let's take it as read that it did fulfil its spec) but the music was corrupted by noise from the power supply (due in turn to too-weak PSRR). Hence in the presence of noise you lost clarity in the music so the perceived effect was weakness (lack of kick, lack of control).
I take 'pressure' to be voltage, but it drops when the tap opens as there's flow (current). So the PSU caps do sag when current is demanded but not enough to make the amp clip (or its power rating would be a fraudulent one). So while you do need current, you need it in the absence of power supply noise. I rather suspect that you don't in fact need any difference in output current.
burst test for audio amplifiers - will reveal the weeknes of power supply
Hi,
For a better understanding of the phenomenon I think we need to read and understud the test methods from here: http://www.tubecad.com/2007/11/blog0124.htm
Take a look at Figure 4 and 5. Even if it is about testing a tube amplifier, the results and the method are accurate method for testing transistors amplifiers.
Hi,
For a better understanding of the phenomenon I think we need to read and understud the test methods from here: http://www.tubecad.com/2007/11/blog0124.htm
Take a look at Figure 4 and 5. Even if it is about testing a tube amplifier, the results and the method are accurate method for testing transistors amplifiers.
Before one tries to modify PS, one ought better to understand what is
going on in the amplifier. The most important thing is that the electrical
design of the amp is of high quality. The second is that the power supply
is sufficient for the intended application. Get a good book on amps.
Hi and thanks for the explanations.
But then an upgrade of mains transformer and ps caps can reduce the noise ?
I understand that a better designed circuit will have a higher PSRR.
But i am more inclined to restoration/upgrading/modifications that to build from the scratch. I like the modding thing a lot.
I still remember when i replaced just one cap in series with the tweeters in a pair of speakers and was amazed by the better transparency. Very very better.
I see. But i have another question.
Let's say that i connect an analog voltemeter across the ps caps.
I am sure the needle will move differently in different amps.
In an amp like the ML2 here below it did not move at all in many domestic situations.
and this will be a sign of very steady power supply to the output devices.
Instead in an amp with "weak" power supply the needle will bounce backwards a lot indeed with the "energic" music.
Am i wrong ?
Thanks again, gino
Sure - lower impedance of the supply means lower noise.
I prefer that too, mods can help improve the PSRR and there may not be any need to start from scratch. For just one example - often there are decoupling (RC) networks (or regulators) for the signal stages fed from the main rails, the caps in these can be beefed up to make a gain in PSRR. In my own integrated power amp (Xindak) I found that improving the regulation to the opamp input stage (fed from LM317/337s initially) made a significant improvement to the PRaT.
Depends on the degree of PSU weakness, but yes I'd expect when the 'weak' amp is played loud you'd see some movement.
Hi and thanks a lot for the valuable advice.
how can i assess this ? i have no clue even with a schematic at hand.
There are amps that are sufficient with some speakers and catch fire with very demanding speakers (i remember Apogee Scintilla speakers, almost a short circuit).
To have an amp able to drive any speaker around will be nice.
I mean ... it is difficult to define the intended application. Loads can vary enourmosly from speaker to speaker.
Do you have any suggestion ?
Thanks and regards, gino
Thanks ! finally one point in favour of ps upgrade
this is exactly my attitude. In the end i see a huge number of Modders around some of them got very famous.
The fact is that i have the feeling that some equipment are basically "sane" designs but cost-cutting decisions have undermined their high potential.
And power supply is where cost-cutting can do most harm ... maybe ?
I have noticed that at the audio fair lately they tend to demo even small bookshelf speakers with huge mono amps.
This seems like overpowering them but instead that take out the problem from the amp.
For me the real only problem of very big linear supplies is their size and weight.
But i think i will end with some big unit.
Thanks a alot again.
Regards, gino
Thanks. I sometimes find something interesting that I want to share with him about electronics, that is the tough part. It was 2010 Christmas day and he was aged 48. It was the Flu that killed him. Like what I heard about Ebola it was when he seemed to be improving that he lost the fight. It goes to prove Flu injections are worth having as they will mostly prevent what is coming. His email name was Valvedabbler. Someone who tries valve circuits. The National Valve Museum was kind enough to publish a letter from the school where he was their technician. I often write things here as it is a small reminder of him. I did see him two hours before the end I am pleased to say. I had to battle the snow to do it and had no idea he was so ill. Sorry to use this as an excuss to say. The Flu warning makes it allowable I feel.
Simon Pearson
Search books by G.Randy Slone and Douglas Self.
Any correctly designed amp should be able to drive the usual
4 ohm loads and have an implemented short circuit and dc
protection.
Abnormal impedance loudspeaker should be treated as a malfunctioning
product in need of a modification or exchange for a normal one.
Let's make an analogy in that respect. You don't make a sports car for
everyday use to weigh 3 tons with an engine of 2000 hp when you can
make a Porsche.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.