System warm-up
 

On Thu, 15 Apr 2004 10:13:39 -0400, "Arny Krueger"
wrote:

Nick Gorham wrote:
Dave Plowman wrote:
In article ,
Mike Gilmour wrote:

What constitutes 'dirty' mains anyway? What about utilities
signalling systems, mains borne baby alarms, intercom systems etc,
not as invasive as Stewarts suggestion of a neighbouring welding
shop but pollution never the less. I'm way out the in country but
'scoping my mains showed evenly spaced spikes... tracked down to an
electric fence a third of a mile away. ;-)


A well designed mains transformer will attenuate anything outside
about 50 Hz by a considerable amount - if it didn't, decent audio
transformers wouldn't be so expensive. ;-)

I have seen normal mains torroids used as valve output transformers.
Thay do a adiquate job, so they must let through (in that situation
anyway) 50hz.

The transformer is the secondary or tertiary line of defense. The power
supply itself carries the major burden.

A typical power supply for audio gear has about 15 VAC on the secondary of
the transformer, which includes a lot of power line and trash. The
rectifier actually ups the trash content by quite a bit so we're still
talking something like 15 volts of power line and other trash at the input
to the filter cap(s). Most audio gear filters the mess with a simple
capacitive filter, which brings the trash content down to a volt or two.
Then, the DC+trash goes through a commodity voltage regulator chip that
costs $0.50 or less. The trash is now under 1 millivolt. Total attenuation
gets us from 15 volts or more of trash to less than one millivolt or trash
or more than 80 dB of attenuation.

On top of the good the power supply does, most audio gear has additional
power supply rejection built into the circuitry itself. There is another 20
or more dB of trash attenuation there, so total trash attenuation is 100 dB
or more.

In every case the power line frequency is the predominate source of trash.
Most of the means used to deal with the trash are very broad band, so the
basic process of keeping power line hum away from the output terminals also
nails the other kinds of trash.

Nice explanation. Keep it up.

System warm-up
 

Arny Krueger wrote:
Nick Gorham wrote:

Dave Plowman wrote:

In article ,
Mike Gilmour wrote:


What constitutes 'dirty' mains anyway? What about utilities
signalling systems, mains borne baby alarms, intercom systems etc,
not as invasive as Stewarts suggestion of a neighbouring welding
shop but pollution never the less. I'm way out the in country but
'scoping my mains showed evenly spaced spikes... tracked down to an
electric fence a third of a mile away. ;-)


A well designed mains transformer will attenuate anything outside
about 50 Hz by a considerable amount - if it didn't, decent audio
transformers wouldn't be so expensive. ;-)

I have seen normal mains torroids used as valve output transformers.
Thay do a adiquate job, so they must let through (in that situation
anyway) 50hz.


The transformer is the secondary or tertiary line of defense. The power
supply itself carries the major burden.

Agreed, I was just pointing out that transformers on their own did
little, and as you say a cap input PSU with cheap rectifiers may well
create move crud than gets in from he mains.

Not sure however just how many power output stages are driven from a
regulated supply, I would have expected the pre stages to be fed from a
regulated supply.

--
Nick

System warm-up
 

Nick Gorham wrote:
Arny Krueger wrote:
Nick Gorham wrote:

Dave Plowman wrote:

In article ,
Mike Gilmour wrote:


What constitutes 'dirty' mains anyway? What about utilities
signalling systems, mains borne baby alarms, intercom systems
etc, not as invasive as Stewarts suggestion of a neighbouring
welding shop but pollution never the less. I'm way out the in
country but 'scoping my mains showed evenly spaced spikes...
tracked down to an electric fence a third of a mile away. ;-)

A well designed mains transformer will attenuate anything outside
about 50 Hz by a considerable amount - if it didn't, decent audio
transformers wouldn't be so expensive. ;-)

I have seen normal mains torroids used as valve output transformers.
Thay do a adiquate job, so they must let through (in that situation
anyway) 50hz.

The transformer is the secondary or tertiary line of defense. The
power supply itself carries the major burden.

Agreed, I was just pointing out that transformers on their own did
little, and as you say a cap input PSU with cheap rectifiers may well
create move crud than gets in from he mains.

It's almost a certainty! Basically the rectifiers take a sine wave with a
few percent crud or less, and turn it into something that has tens of
percent of THD when meausred.

Not sure however just how many power output stages are driven from a
regulated supply,

Typically there is no formal regulation in the power supply related to the
feed for the output stages..

Under full load, there are usually several volts of AC across the power
supply caps. It's easy enough to measure in most power amps.

The first line of defense is the power supply noise rejection of the power
amp itself. That that this is sufficient to build power amps with all noise
and distortion 90 or more dB down should be noted.

I would have expected the pre stages to be fed from a regulated supply.

That is usually the case. The early stages of the main part of the power amp
circuit often run off the same source as the output devices. In other cases
there is some built-in extra filtering.

http://www.citycom.gr/electronics/pr...l/100ampl5.gif shows a
fairly typical design. Everthing runs off the same voltage source as the
output devices.

BTW, RIP Doub Self's "Amplifier Institute". A sad loss.

System warm-up
 

In article , Arny Krueger
writes
Nick Gorham wrote:
Dave Plowman wrote:
In article ,
Mike Gilmour wrote:

What constitutes 'dirty' mains anyway? What about utilities
signalling systems, mains borne baby alarms, intercom systems etc,
not as invasive as Stewarts suggestion of a neighbouring welding
shop but pollution never the less. I'm way out the in country but
'scoping my mains showed evenly spaced spikes... tracked down to an
electric fence a third of a mile away. ;-)


A well designed mains transformer will attenuate anything outside
about 50 Hz by a considerable amount - if it didn't, decent audio
transformers wouldn't be so expensive. ;-)

I have seen normal mains torroids used as valve output transformers.
Thay do a adiquate job, so they must let through (in that situation
anyway) 50hz.

The transformer is the secondary or tertiary line of defense. The power
supply itself carries the major burden.

A typical power supply for audio gear has about 15 VAC on the secondary of
the transformer, which includes a lot of power line and trash. The
rectifier actually ups the trash content by quite a bit so we're still
talking something like 15 volts of power line and other trash at the input
to the filter cap(s). Most audio gear filters the mess with a simple
capacitive filter, which brings the trash content down to a volt or two.
Then, the DC+trash goes through a commodity voltage regulator chip that
costs $0.50 or less. The trash is now under 1 millivolt. Total attenuation
gets us from 15 volts or more of trash to less than one millivolt or trash
or more than 80 dB of attenuation.

On top of the good the power supply does, most audio gear has additional
power supply rejection built into the circuitry itself. There is another 20
or more dB of trash attenuation there, so total trash attenuation is 100 dB
or more.

In every case the power line frequency is the predominate source of trash.
Most of the means used to deal with the trash are very broad band, so the
basic process of keeping power line hum away from the output terminals also
nails the other kinds of trash.



15 volts seems a tad low when I last looked the Audiolab amps had about
twice that. Not that I can say I've ever had a problem with power line
supplies....
--
Tony Sayer

System warm-up
 

"Arny Krueger" wrote in message
...
Nick Gorham wrote:
Dave Plowman wrote:
In article ,
Mike Gilmour wrote:

What constitutes 'dirty' mains anyway? What about utilities
signalling systems, mains borne baby alarms, intercom systems etc,
not as invasive as Stewarts suggestion of a neighbouring welding
shop but pollution never the less. I'm way out the in country but
'scoping my mains showed evenly spaced spikes... tracked down to an
electric fence a third of a mile away. ;-)


A well designed mains transformer will attenuate anything outside
about 50 Hz by a considerable amount - if it didn't, decent audio
transformers wouldn't be so expensive. ;-)

I have seen normal mains torroids used as valve output transformers.
Thay do a adiquate job, so they must let through (in that situation
anyway) 50hz.

The transformer is the secondary or tertiary line of defense. The power
supply itself carries the major burden.

A typical power supply for audio gear has about 15 VAC on the secondary
of
the transformer, which includes a lot of power line and trash. The
rectifier actually ups the trash content by quite a bit so we're still
talking something like 15 volts of power line and other trash at the input
to the filter cap(s). Most audio gear filters the mess with a simple
capacitive filter, which brings the trash content down to a volt or two.
Then, the DC+trash goes through a commodity voltage regulator chip that
costs $0.50 or less. The trash is now under 1 millivolt. Total
attenuation
gets us from 15 volts or more of trash to less than one millivolt or
trash
or more than 80 dB of attenuation.

On top of the good the power supply does, most audio gear has additional
power supply rejection built into the circuitry itself. There is another
20
or more dB of trash attenuation there, so total trash attenuation is 100
dB
or more.

In every case the power line frequency is the predominate source of trash.
Most of the means used to deal with the trash are very broad band, so the
basic process of keeping power line hum away from the output terminals
also
nails the other kinds of trash.


Replacing rectifiers with Schottky barrier ones that supposedly don't
generate reverse recovery transients from stored charge & minority carrier
injection - does that change the situation much even if you don't use a
regulator chip?

System warm-up
 

On Thu, 15 Apr 2004 12:36:58 -0400, "Arny Krueger"
wrote:

That is usually the case. The early stages of the main part of the power amp
circuit often run off the same source as the output devices. In other cases
there is some built-in extra filtering.

http://www.citycom.gr/electronics/pr...l/100ampl5.gif shows a
fairly typical design. Everthing runs off the same voltage source as the
output devices.

If by 'extra filtering' you're talking about C9 and 1, then it's not
really very useful--those capacitors will follow the main smoothing
capacitors and discharge through the output stage. A schottky diode to
separate the main supply capacitors from local smoothing on an
isolated small signal supply would work wonders to reject rail sags
and grunge caused by the output stage.

--
td

System warm-up
 

The Devil wrote:
On Thu, 15 Apr 2004 12:36:58 -0400, "Arny Krueger"
wrote:

That is usually the case. The early stages of the main part of the
power amp circuit often run off the same source as the output
devices. In other cases there is some built-in extra filtering.

http://www.citycom.gr/electronics/pr...l/100ampl5.gif
shows a fairly typical design. Everthing runs off the same voltage
source as the output devices.

If by 'extra filtering' you're talking about C9 and 1, then it's not
really very useful-

I'm not talking about that at all. Sorry for any possible lack of clarity. I
intended this to be an example of a power amp where everything runs off the
same basic voltage source as the output devices. In fact, that's what I
said - " Everything runs off the same voltage
source as the output devices."


-those capacitors will follow the main smoothing
capacitors and discharge through the output stage.

Agreed.

A schottky diode to
separate the main supply capacitors from local smoothing on an
isolated small signal supply would work wonders to reject rail sags
and grunge caused by the output stage.

As a rule, tweaks such as this just don't have any practical technical
advantages. Wire 'em in and there's usually no visible effect on the display
of residuals from distortion analyzer.

System warm-up
 

In article ,
Nick Gorham wrote:
A well designed mains transformer will attenuate anything outside
about 50 Hz by a considerable amount - if it didn't, decent audio
transformers wouldn't be so expensive. ;-)

I have seen normal mains torroids used as valve output transformers.
Thay do a adiquate job, so they must let through (in that situation
anyway) 50hz.

I must admit to not having tested a toroidal - it's ages since I looked at
this. But if it's letting through significant amounts of 'interference'
that the normal smoothing and decoupling doesn't stop within the audio
band, I'd assume you'd hear it.

--
*The e-mail of the species is more deadly than the mail *

Dave Plowman London SW 12
RIP Acorn

System warm-up
 

Arny Krueger wrote:



BTW, RIP Doub Self's "Amplifier Institute". A sad loss.

Ouch. What happened. Any idea why it went? We have already lost Linsley
Hood, Doug Self hasn't gone too has he?

Ian

System warm-up
 

In message , Dave Plowman
writes
In article ,
Nick Gorham wrote:
A well designed mains transformer will attenuate anything outside
about 50 Hz by a considerable amount - if it didn't, decent audio
transformers wouldn't be so expensive. ;-)

I have seen normal mains torroids used as valve output transformers.
Thay do a adiquate job, so they must let through (in that situation
anyway) 50hz.

I must admit to not having tested a toroidal - it's ages since I looked at
this. But if it's letting through significant amounts of 'interference'
that the normal smoothing and decoupling doesn't stop within the audio
band, I'd assume you'd hear it.


Don't forget that the toroids used in audio amps tend not to have
inter-winding screens. The inter-winding capacitance also tends to be
higher than that of E-I core transformers. They therefore let through
pretty much all the common-mode noise on the mains.

I know that neutral is bonded to ground at the sub-station, but in many
places the common-mode on the mains is horrible (not helped by all the
CM crap coming out of SMPSUs, even those with power-factor correction).
A good common-mode filter will clean things up a lot.


--
Chris Morriss