HY60
 

"Jim Lesurf" wrote in message


That isn't really a matter of changing to an actively
regulated PSU. Just one of ensuring the PSU has a low
output impedance and ripple even when the current demand
is high. if the "aural difference was dramatic" then I'd
suspect something was wrong with the unregulated PSU
you'd used.

Agreed.


Unregulated supplies can give the advantage
of providing higher peak transient powers. Given that
music tends to have higher crest factors than test
sinewaves that can be quite useful. However the amp has
to be built to cope.

I've done a number of experiments powering power amps via power supplies
that have current limiting. As long as a reasonable energy storage
reservoir is available, amplifying music does not require a great deal of
average current.

HY60
 

"Eeyore" wrote in
message
Eiron wrote:

Eeyore wrote:

Don't be silly. Even Metallica isn't square waves with
no dynamic range.

or twice that using 4 ohm speakers.

Who mentioned 4 ohm speakers. Other than you ?

Serge and Woody mentioned them.

But not the OP. Whom I hope is aware of the issue.


Haven't you heard, or heard about, the new Metallica CD?
It was in the news recently that it was so compressed
that even heavy metal fans were complaining, so pretty
close to square waves with no dynamic range.

I heard about the poor dynamic range.

It won't be square waves though, so no worse than
full-power sinewave. And no matter how hard you compress
it'll probably still have 3 dB of dynamic range. I speak
as someone who has run such tests.

The fact that there are concurrent significant signals at different
frequencies literally decimates the peak-to-average ratio.

HY60
 

"Rob" wrote in message


Do you think most off the shelf amplifiers are designed
to be driven hard continuously, and if not, how is the
sound affected?

Most amplifiers are designed to be driven to full power for an extended
period of time, into a typical speaker load.

BTW, I downloaded a segment of Metallica's "That was just your life". Its
average power is about 4 dB less than that of a full-scale square wave, and
about 1 dB less than a full-scale sine wave. IOW a power amp that can
handle a full power sine wave should have no problems amplifying it.

I still don't understand this 'all modern amplifiers
sound the same' thing - hence the question!

The chore that a power amplifier addresses is pretty simple and the means
for engineering equipment to perform that chore is pretty well understood
and documented.

Why would you expect a amplifier to sound different, other than for obvious
reasons like one power amp is rated at 10 watts, and the next is rated at
100 watts, and your listening habits run at around 20 watts.

HY60
 

Woody wrote:
[sni[]
Interestingly, one of the amp modules still has some 10nF ceramics
soldered to the power rails.

[snip]


Does that suggest that there may have been a slight problem with
instability in the unit design if it needs to be decoupled right at
the input.


Not the input, the supply rails.

I rescued them from a PA system that was being scrapped at a telephone
exchange many years ago. A Strowger exchange. Perhaps the original builder
was attempting to suppress some of the electrical noise that must have been
around. For all I know he may even have been running the PA system from the
50V
DC supply that supplied the Strowger switches.


For those that are not old enough to remember them have a look at
http://datasheet.digchip.com/000/000-0-HY60.pdf
which shows that the heatsink is actually quite large.


Yes, they are like small motorcycle cylinder heads :-)

And if it gets too hot you can always add a small fan?


There is a lot of mass and surface area in those heatsinks. As much as most
30W amps I have come across. I'll deal with a heating problem if it arises.
But an issue in this thread seems to be dissipation in the supply rail
bypass transistors. I'll just suck it and see :-)

TonyL

HY60
 

"TonyL"
Phil Allison wrote:

It depends very much on the type of regulator topology.

A zener controlled emitter follower is not very low Z


Phil,

I don't understand that....explain ??


** What bit needs explaining?



...... Phil

HY60
 

"David Looney"
"Phil Allison"
"Eeysore" wrote:


The *thickness* of the case is
almost irrelevant.


** No it is NOT you LYING ****wit.

**** OFF !!

The thermal conductivity matters too, of course, but if it's aluminium
then that isn't likely to be a problem unless the heatsink was so thin as
to be mechanically unsound.

** ********.


So thickness of the metal is, as Graham said, almost irrelevant.


** And what thickness did I suggest as a safe minimum ?

You ASD ****ed MORON !!



...... Phil

HY60
 

Phil Allison wrote:

"Eeysore Lying Criminal ****wit "

Even pro amps with SMPSs don't make any serious attempt
to regulate the rails.

** Shame how so many of them do - you know nothing ASS !!

Like all the ones with PFC corrected PSUs !!!!!!!!!!!!!!

That's merely the effect of the more stable DC bus. It's NOT 'regulation'.

** It most certainly IS regulation of the DC supply rails

No, PFC correction simply provides a fairly stable live side DC bus. The output
will still sag a bit under load but somewhat less than it would otherwise.


- YOU ASININE, ****WIT, LYING, ****ING ASSHOLE

Provide an example.

Graham

HY60
 

"Arny Krueger"


I've done a number of experiments powering power amps via power supplies
that have current limiting. As long as a reasonable energy storage
reservoir is available, amplifying music does not require a great deal of
average current.


** Correct.

The important thing is to HAVE a large electro cap at the output of the
regulator - so the series pass device will not "see" the current peaks being
drawn by the amplifier.

In the OP's example of a 30 watt @ 8 ohms amplifier with +/- 25 volt DC
rails - even full sine wave power requires under 1 amp of DC current per
rail while normal music programme with peaks to full power will require no
more than 1/3 of that.



..... Phil

HY60
 

TonyL wrote:

Eeyore wrote:

Not since I raided my heatsink store for a previous project...a
13.8V 30A transceiver supply. For this little project the metal
enclosure will make an adequate heatsink.

How do you KNOW that ? I'm a heatsink expert too btw.

I don't KNOW beyond all doubt.

You may get a surprise.


But that doesn't matter.

No ?


This isn't a
commercial design that has to be right first time.

Yeah .... but if you break even one of the modules it won't be a design at all.



If my fingers tell me
that the enclosure is not an adequate heatsink during testing I will simply
add a larger extruded/finned version.

You have calibrated fingers ? I use thermocouples.

Graham

HY60
 

Jim Lesurf wrote:

Can also help suppress snap noise from the bridge diodes if you haven't
already shunted the diodes with suitable RF caps.

Which is an infinitely better place to put them. I remember reading a review in
Studio Sound in the early days of the imminent EMC legislation and 2 reviewers
including I think Doug Self were totally perplexed by mains noise emissions and
came up with an audiophool-like 'voodoo explanation' for it. It was of course a
lack of said diodes.

Graham