Speaker Amplifier

Hi Guys, i had a couple of questions regarding Speaker Amplifier,

I was looking at the Hypex Products as some have referred to in the past, and the Speaker thread in Studio A was a reminder,
and while i was doing some reasearch and looking around, there are two things that got me thinking.

One beeing that some builds based on Hypex, have used one Toroidal for each Amp and one PSU for each, what is the benefit of doing so?
One bigger VA and beefier PSU would do the trick no? Or it's a safety practice like another....
EDIT: Forgot the price factor beeing the one... :

And what's the advantages of having a potted vs unpotted Power transformer, i read:

Electromagnetic Shielding          info or Intox?

Mechanical and environemental Protection                seems logical


Anything else? Beside the looks....


Thanks for your toughts and feedbacks.

Two power supplies are theoretically better at reducing crosstalk via PS rejection path. But generally just more expensive if design has good PSRR.

Potted should be quieter with less mechanical vibration, while torroids are not as mechanically noisy as EI transformers. 


JR

the mechanical vibrations and leakage magnetic flux are proportional to the power trasformer's out power that is mainly equal to the amp's out power if this amp is an AB class type.
a good toroidal trasformer has very low mechanical vibrations and low leakage magnetic flux, so for low amp's output power the mechanical vibrations and leakage magnetic flux are pratically near to zero so the amp have pratically no induced noise in its output (when the amp's out power is low the power requested to pwer trasformer is low). When the out power of the amp is big however the small induced noise is well covered by the music even if the trasformer makes much more mechanical vibrations and leakage magnetic flux than in the case the amp out power is low.
Moreover, a toroidal trasformer is smaller than a traditional one of the same VA rate. 

Thanks guys, you made that clear...

So Concerning PSU, there are many talks about Switching PSU versus Linear on the Sound effect In a Power amp,
Saying a Good PSU design will always make the Amplifier restitute better on speakers....
Seems understandable if this one cleans the signal well enough and gives a clean DC on the other end,
well if i'm getting the picture correctly....

A good old linear Supply with a good amount of filtering and protection can't be as good as a Switching one?

And so Linear would be one with poor Crosstalk reduction VS Switching, or vice versa, or depends?

They were some threads here about Switching, and the fact that they are not very reliable,
i actually experienced that quite often, so i can understand that haha...., but i guess that also depends on the
parts quality and on the design itself.....
but also the fact that Switching are used most of the time by manufactures, to be taken out of the picture if something fails
in that domain, as they use Third part Company's for those...
But don't seem to remember one saying that Switching was better in audio quality of Speaker Amps,
but i believe most of the active speaker ranges are all based on Switching PSU's right?


Thanks again for the toughts guys, much appreciated.

it's pratically impossible reduce to zero the mechanical noise of a power switching PSU, since its power transistors and its trasformer make mechanical noise. Very often a fraction of the ripple noise made by the SPSU's, is in the amp's out even in the case there is not present any audio signal in this out. Yes, since the fundamental frequency harmonic generated by a SPSU is in the ultrasonic band,  is impossible to hear it in the speaker, but this noise does intermodulation distortion in the amp so there are products of this noise in the audio band too.   
 

zayance said:

But don't seem to remember one saying that Switching was better in audio quality of Speaker Amps,
but i believe most of the active speaker ranges are all based on Switching PSU's right?

Click to expand...

To my knowledge, most amplified studio monitors use a linear PSU. SMPS are used mostly in amplified PA speakers.
A linear PSU has the advantage of having a higher voltage when idling or at low power, which gives it the possibility to have more transient power (typically a power amp with linear PSU will have an instant peak power close to 2x its rms power (3dB more headroom). If it has an SMPS, the DC voltage is regulated, which makes the peak power about equal to rms.
Obviously, an SMPS has more parts than a linear, so the MTBF is lower, as almost anybody has experienced.
It is true that SMPS produce interference; however, they are not really present in the ouptut (in a well designed product), but they can be transmitted by magnetic radiation. That's the main problem.
A linear PSU has more ripple (essentially 100Hz residue) than SMPS, but this is generally not a problem since the power amplifiers provide quite a good PSRR.
SMPS may have high frequency noise, which the power amp may have difficulties in eliminating.
PSRR typically varies in parallels with open-loop gain, so decreases rapidly with frequency.

Thanks a lot guys, understood, i'm asking these because i''ll maybe build one with the Hypex stuff later this year,
I don't have the opportunity to listen to one of these amps in my situation, but will see....

If i go for it I want to bring the cost a little down concerning the build of it, some will say buy all the stuff from Hypex and plug and play, but....;
For some parts i think it will be the case, they seem to have a good price for their potted transformers, seems i can buy these Stuff
From a Hifi French Webshop so will see.....

Now will i use one or two?, with what was said seems two would be better, because maybe the PSU i'll make will have a poor PSRR?

Actually what are the factors that makes a PSU have a good PSRR, a simple PSU using Bridge rectifier/Caps etc... will have a poor PSRR?

And i read that when using high VA transformers like 500VA, these will have a high Inrush current on start up,
and so it's best to have a slow start circuitry for that, but isn't that even better for a 250VA transformer then?

For their Linear PSU, i think they are asking a little too much, but i may be wrong, as good Caps will cost some money,
And i think it's also more fun to have at least something to do in the process, so i guess it's possible to achieve similar results at home,
designing a Circuit board, home etch it etc...., and use all the extra stuff i have laying around, and try to buy from one supplier only,
Of course time is not counted.

Here are the Schemo's i found of one guy that was building a Monoblock using a UcD400 for driving his Subwoofer,
and while he added Switch circuit that drives a Relay for Start up, i will omit that and go on simple Rocker switch,
and a simple ON Led on the Front Panel, nice and easy, but this gave me some hints and help anyhow, what do you think? Any upgrades? Or nice as is?

So using UltraFast Diode Rectifiers, i was thinking of using only MUR860 ones, or it's better to stick with the HFA25PB60 or equivalent?
For the Caps i was thinking of Cornell Dubilier of 10000uF 80V Caps instead of 22000 63V this guy used, less Filtering yes, the Higher the better in that situation?
40000uF is what is used in Hypex stuff...


Concerning the PSU protection, i was going to add, as stated on the UcD Product sheet, and as seen on this Schematic, fast 4A fuse on each Lines, but on the Hypex Website
they say that their Linear PSU has DC protection, and sorry for my lack of knowledge on that but what is tipically used for DC protection?

And on the Schemo the little network of C7 / R2 / D5 / D6 is used for helping the Caps Discharing on Power off and helps reducing hum while in use, right?
what is the purpose of the diodes here, sorry if it sounds stupid  :-\.

Thanks again for your toughts and knowledge guys.



PSU Circuit found on the Web:




Slow Start Circuit found on the Web, will omit Rel1....

zayance said:

Now will i use one or two?, with what was said seems two would be better, because maybe the PSU i'll make will have a poor PSRR?

Actually what are the factors that makes a PSU have a good PSRR, a simple PSU using Bridge rectifier/Caps etc... will have a poor PSRR?

Click to expand...

PSRR is intrinsic to the power amp. You don't have to worry about your PSU having poor PSRR. But you have to make sure you use properly dimensioned capacitors, which is very easy, in particular if you have Hypex' recommended value. You'll need to make sure you take the v connection from the junction of both smoothing caps, and that the onnection between those two caps is as solid as possible.

And i read that when using high VA transformers like 500VA, these will have a high Inrush current on start up,
and so it's best to have a slow start circuitry for that, but isn't that even better for a 250VA transformer then?

Click to expand...

KISS. Use a slo-blo fuse; you can also use a CTP inrush limiter, like this:
http://www.tubeampdoctor.com/en/shop_Resistors_MOV_special_resitors/Thermistor_like_used_in_Fender_replaces_C60_11_1689

Concerning the PSU protection, i was going to add, as stated on the UcD Product sheet, and as seen on this Schematic, fast 4A fuse on each Lines, but on the Hypex Website
they say that their Linear PSU has DC protection, and sorry for my lack of knowledge on that but what is tipically used for DC protection?

Click to expand...

Fuses?

And on the Schemo the little network of C7 / R2 / D5 / D6 is used for helping the Caps Discharing on Power off right?

Click to expand...

No, its basically an RC circuit that introduces a "soft" path between audio ground and earth, for avoiding ground loops. But for safety reasons, there are back-to-back diodes that limit the ground voltage to less than 0.7V.

PSRR is intrinsic to the power amp. You don't have to worry about your PSU having poor PSRR. But you have to make sure you use properly dimensioned capacitors, which is very easy, in particular if you have Hypex' recommended value. You'll need to make sure you take the v connection from the junction of both smoothing caps, and that the onnection between those two caps is as solid as possible.

Click to expand...

Oh ok, so in my case if i design a PCB for my purpose using Eagle, I'll first make sure the tracks are thick enough, but GND will be GND plane, good enough?
Of course i could build it P2P and so take good care of connecting these caps together as safely as possible as you said....

KISS. Use a slo-blo fuse; you can also use a CTP inrush limiter, like this:
http://www.tubeampdoctor.com/en/shop_Resistors_MOV_special_resitors/Thermistor_like_used_in_Fender_replaces_C60_11_1689

Click to expand...

hmm so i could omit the Slow start circuitry and just use those Thermistors on each line and be done with it? Nice.

No, its basically an RC circuit that introduces a "soft" path between audio ground and earth, for avoiding ground loops. But for safety reasons, there are back-to-back diodes that limit the ground voltage to less than 0.7V.

Click to expand...

Oh ok understood.

Thanks a lot for the clarifications abbey.

zayance said:

KISS. Use a slo-blo fuse; you can also use a CTP inrush limiter, like this:
http://www.tubeampdoctor.com/en/shop_Resistors_MOV_special_resitors/Thermistor_like_used_in_Fender_replaces_C60_11_1689

Click to expand...

hmm so i could omit the Slow start circuitry and just use those Thermistors on each line and be done with it? Nice.

Click to expand...

Just one, in series with the fuse is enough.

Allright cool.
So theorically one Power transformer for each Amplifier channel is better for Crosstalk, but i could just stick with
One Power transfo (500Va) and that for driving the two channels and one PSU for bringing the cost down
and this will still deliver good results if good components are used right?
What you guys would recommend in your experience....


Thanks a lot.

abbey road d enfer said:

It is true that SMPS produce interference; however, they are not really present in the ouptut (in a well designed product), but they can be transmitted by magnetic radiation. That's the main problem.
A linear PSU has more ripple (essentially 100Hz residue) than SMPS, but this is generally not a problem since the power amplifiers provide quite a good PSRR.
SMPS may have high frequency noise, which the power amp may have difficulties in eliminating.
PSRR typically varies in parallels with open-loop gain, so decreases rapidly with frequency.

Click to expand...

Yes, magnetic radiation it's the main problem but the high frequency noise is however a big problem. Infact, often, to reduce the costs several amps with SMPS's are designed likewise they have a linear PSU's because, however, the designers think that high frequency noise is not udible. As I remember (we speak of 12 years ago) , that two amp design manuals say that the SMPS's must have the fundamental frequency over 20 khz to avoid to hear it (and its harmonics) in amps' out. This problem is expecially for the car amps since the SMPS of these amps often handles very hight currents since it is powered with 12-14V.
Yes, there are well designed amps with SMPS with no noise in their outs but the cheap car amps are the most diffused amps with SMPS.