Gentlemen:

After measurement on circuits and simulation, I have decided to make
no change to the power supplies of my circuits offered for comment:
http://members.lycos.co.uk/fiultra/T...trafi-crct.jpg
http://members.lycos.co.uk/fiultra/T44bis-'Populaire'-crct.jpg
http://members.lycos.co.uk/fiultra/t...17acircuit.jpg
I am of course still, and always, open to reasoned argument to the
contrary.

Here are some conclusions on common mode chokes, and more specifically
on common mode chokes in choke input power supply filters, offered for
further comment, clarification and kibbitzing:

1. A common mode choke is two closely coupled coils on the same core,
intended to be inserted in the power and ground leads of a supply (or
in signal-carrying and ground leads) for the rejection of common mode
noise.

2. Two separate chokes, one on each line, does not have the same effect
as a common mode choke of the same value. A simple test is to observe
on an oscilloscope a square wave through a common mode choke. The
signal is not visibly degraded except marginally at the very highest
magnification. With two separate similar chokes, one to each line, the
squarewave comes out looking like a mud hut after a monsoon, with all
the corners completely knocked off. Electrically, two separate chokes,
one on each line, are equivalent to two separate chokes in parallel on
one line (that is, a reduction in the total inductance), whereas a
common mode choke, one winding to each line, is equivalent to the two
coils in series on one line (that is, an increase in the total
inductance).

3. Common mode chokes are an economic method for suppression of noise
under various codes; Australian Standard AS 2279 'Harmonic
disturbance' is an example, or try the similar European VDE
Specification 0160 if that is more conveniently accessible. Such common
mode chokes are *commonly* inserted in the choke input position. This,
for instance, is from the circuit description of Zener Electric's
VSC2000 power train for their widely used variable speed motor
controller: "The DC choke (or inductor) is included in the input
circuit of the DC bus between the rectifier and the DC bus
capacitor." A clearer description of a choke-input common-mode choke
one cannot wish for.

4. A common mode choke is effective in reducing power supply noise,
among other parameters, in proportion to the closeness of the coupling
of the twin coils about the single core. The fact that coupling is
never 100% in practice in audio work blocks radio frequency
interference (RFI). The common mode impedance created by less than 100%
coupling also blocks electromagnetic emissions (EMI). Serendipity
strikes again.

5. With or without a common mode choke, parasitic capacitances from the
primary side of the transformer and from the load couple through the
ground reference. Common mode chokes also reduce the noise transmitted
by these stray capacitances. (1)

6. I have used common mode chokes in choke input power supplies on my
amplifiers for over a decade and have never observed any adverse
effects. Quite the contrary. The use of a choke of any kind, including
a common mode choke, on the ground line appears to act as a barrier to
various manners of noise. I have neither the knowledge nor the skill
with test equipment, nor for that matter the inclination or the time,
to separate the "slowing" effects of a choke, any choke, from the
"common core" effects of a common mode choke, but am happy to
accept any additional reduction of noise in my amps whatever the cause.

7. The Ryder materials offered by John Stewart are an unsupported,
overblown scare story. Ryder claims that a choke in the ground lead of
a choke input filter through the secondary to primary capacitance of
the power transformer puts at least half the transformer secondary (in
the case of a centretapped secondary) AC in series with choke through
the ground line. I have been unable to measure this on real-life amps,
have never observed it in more than a decade of using common mode
chokes in choke input filters and, while I don't doubt that there is
a parasitic capacitance as mentioned by Ryder, I doubt the magnitude of
its effect as described by him: hundreds of volts of AC ripple in the
filtered supply. I suppose it is possible in the sense that anything is
possible, but then I don't build a barrier against the *possibility*
that The Titanic will surface, sail down the Irish Channel, cross six
miles of land (uphill), and crash into my house to smash all my
amplifiers. I don't live in that parallel universe - and in this
one I feel quite secure, as apparently does tens of thousands of
professional engineers with liability lawyers looking over their
shoulders, in using common mode chokes in choke input supplies.
Significantly, those who appeal to Ryder's authority cannot explain
how his effects are achieved; clumsy sneering and jeering cuts no ice
with me; the only result is publicly to embarrass those too thick to
explain what they claim by faith in their priesthood.

8. I shall return to the noise hunt in the KISS amps by other means in
another thread.

Thanks you to those who contributed worthwhile thoughts and questions
so far and I look forward to your further contributions.

Andre Jute

(1) There are other means than common mode chokes, or in addition to
common mode chokes, of keeping power supply noise out of the load by
going well beyond the KISS principles my audio amps normally adhere to.
For instance, in a design intended for production, by connecting a
small capacitor from the signal line to the case and another between
ground line and case, it is possible to return noise to its source:
****
noise on signal or ground line to capacitor to metallic
casing to stray capacitance to noise source
****
The problem for the DIYer with this scheme is that, at least in the EU,
you may then have to cover the metal case so that it is nowhere
connected to anything the user can touch. (Iain may know more about
this. I just don't use this method any more.) Control knob shafts,
switch bezels, switch toggles, screws, vent grilles... The list of
metal components you must replace is pretty long.

Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review

"Andre Jute" wrote in message ups.com...
[Deleted]

You are hopelessly, hopelessly confused, Andre.

-Henry

Henry Pasternack wrote:
"Andre Jute" wrote in message ups.com...
[Deleted]

You are hopelessly, hopelessly confused, Andre.

-Henry

Confusion assumes initial good intent.

Peter Wieck
Wyncote, PA

Andre,

As I alluded to in an earlier post, a common mode choke connected between
a rectifier and a reservoir capacitor does not make a choke input filter,
and on the other hand if you phase the two windings, one in the negative
lead and the other in the positive lead, such that you have a choke input
filter, then you loose the common mode rejection capabilities of the
common mode choke connection. You can't have both effects at the same
time, and which you get depends on the relative phasing of the two
windings of the choke.


Regards,

John Byrns


In article . com, "Andre
Jute" wrote:

Gentlemen:

After measurement on circuits and simulation, I have decided to make
no change to the power supplies of my circuits offered for comment:
http://members.lycos.co.uk/fiultra/T...trafi-crct.jpg
http://members.lycos.co.uk/fiultra/T44bis-'Populaire'-crct.jpg
http://members.lycos.co.uk/fiultra/t...17acircuit.jpg
I am of course still, and always, open to reasoned argument to the
contrary.

Here are some conclusions on common mode chokes, and more specifically
on common mode chokes in choke input power supply filters, offered for
further comment, clarification and kibbitzing:

1. A common mode choke is two closely coupled coils on the same core,
intended to be inserted in the power and ground leads of a supply (or
in signal-carrying and ground leads) for the rejection of common mode
noise.

2. Two separate chokes, one on each line, does not have the same effect
as a common mode choke of the same value. A simple test is to observe
on an oscilloscope a square wave through a common mode choke. The
signal is not visibly degraded except marginally at the very highest
magnification. With two separate similar chokes, one to each line, the
squarewave comes out looking like a mud hut after a monsoon, with all
the corners completely knocked off. Electrically, two separate chokes,
one on each line, are equivalent to two separate chokes in parallel on
one line (that is, a reduction in the total inductance), whereas a
common mode choke, one winding to each line, is equivalent to the two
coils in series on one line (that is, an increase in the total
inductance).

3. Common mode chokes are an economic method for suppression of noise
under various codes; Australian Standard AS 2279 'Harmonic
disturbance' is an example, or try the similar European VDE
Specification 0160 if that is more conveniently accessible. Such common
mode chokes are *commonly* inserted in the choke input position. This,
for instance, is from the circuit description of Zener Electric's
VSC2000 power train for their widely used variable speed motor
controller: "The DC choke (or inductor) is included in the input
circuit of the DC bus between the rectifier and the DC bus
capacitor." A clearer description of a choke-input common-mode choke
one cannot wish for.

4. A common mode choke is effective in reducing power supply noise,
among other parameters, in proportion to the closeness of the coupling
of the twin coils about the single core. The fact that coupling is
never 100% in practice in audio work blocks radio frequency
interference (RFI). The common mode impedance created by less than 100%
coupling also blocks electromagnetic emissions (EMI). Serendipity
strikes again.

5. With or without a common mode choke, parasitic capacitances from the
primary side of the transformer and from the load couple through the
ground reference. Common mode chokes also reduce the noise transmitted
by these stray capacitances. (1)

6. I have used common mode chokes in choke input power supplies on my
amplifiers for over a decade and have never observed any adverse
effects. Quite the contrary. The use of a choke of any kind, including
a common mode choke, on the ground line appears to act as a barrier to
various manners of noise. I have neither the knowledge nor the skill
with test equipment, nor for that matter the inclination or the time,
to separate the "slowing" effects of a choke, any choke, from the
"common core" effects of a common mode choke, but am happy to
accept any additional reduction of noise in my amps whatever the cause.

7. The Ryder materials offered by John Stewart are an unsupported,
overblown scare story. Ryder claims that a choke in the ground lead of
a choke input filter through the secondary to primary capacitance of
the power transformer puts at least half the transformer secondary (in
the case of a centretapped secondary) AC in series with choke through
the ground line. I have been unable to measure this on real-life amps,
have never observed it in more than a decade of using common mode
chokes in choke input filters and, while I don't doubt that there is
a parasitic capacitance as mentioned by Ryder, I doubt the magnitude of
its effect as described by him: hundreds of volts of AC ripple in the
filtered supply. I suppose it is possible in the sense that anything is
possible, but then I don't build a barrier against the *possibility*
that The Titanic will surface, sail down the Irish Channel, cross six
miles of land (uphill), and crash into my house to smash all my
amplifiers. I don't live in that parallel universe - and in this
one I feel quite secure, as apparently does tens of thousands of
professional engineers with liability lawyers looking over their
shoulders, in using common mode chokes in choke input supplies.
Significantly, those who appeal to Ryder's authority cannot explain
how his effects are achieved; clumsy sneering and jeering cuts no ice
with me; the only result is publicly to embarrass those too thick to
explain what they claim by faith in their priesthood.

8. I shall return to the noise hunt in the KISS amps by other means in
another thread.

Thanks you to those who contributed worthwhile thoughts and questions
so far and I look forward to your further contributions.

Andre Jute

(1) There are other means than common mode chokes, or in addition to
common mode chokes, of keeping power supply noise out of the load by
going well beyond the KISS principles my audio amps normally adhere to.
For instance, in a design intended for production, by connecting a
small capacitor from the signal line to the case and another between
ground line and case, it is possible to return noise to its source:
****
noise on signal or ground line to capacitor to metallic
casing to stray capacitance to noise source
****
The problem for the DIYer with this scheme is that, at least in the EU,
you may then have to cover the metal case so that it is nowhere
connected to anything the user can touch. (Iain may know more about
this. I just don't use this method any more.) Control knob shafts,
switch bezels, switch toggles, screws, vent grilles... The list of
metal components you must replace is pretty long.

Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review


Surf my web pages at, http://users.rcn.com/jbyrns/

John Byrns wrote:
Andre,

As I alluded to in an earlier post, a common mode choke connected between
a rectifier and a reservoir capacitor does not make a choke input filter,
and on the other hand if you phase the two windings, one in the negative
lead and the other in the positive lead, such that you have a choke input
filter, then you loose the common mode rejection capabilities of the
common mode choke connection. You can't have both effects at the same
time, and which you get depends on the relative phasing of the two
windings of the choke.

Regards,

John Byrns


I don't get his, John. My LL1638 have four terminals. Connected one way
they are a single choke, two coils in series on a single core.
Connected the other way they are a common mode choke, two coils in
different leads on a single core.

I do not see the reason why in the first connection, inserted on the
signal line, that cannot be a choke input filter.

Nor do I understand, if the two coils are connected one to each line
around the common core, why connection as a common mode choke in the
choke input position should preclude operation as a common mode choke.

In each case you appear to be saying that what the thing is a connected
as is precisely what the thing won't do. If there are subtleties here,
they are presently beyond my poor brain.

I should be grateful if you would provide some reasoning or a source I
can look up.

I've found only one audio-specific common-mode choke input design
(alleged) on the net at:
http://www.musicalsurroundings.com/aesthx2.html
where it is claimed "The primary high voltage supply is composed of a
similar massive 170 VA transformer which feeds a common-mode choke
input filtering network" -- but who can tell what they think is "a
common-mode choke input" without the schematic?

Thorsten Loesch, who is well known to me, here (scroll down a long way
to the circuit):
http://www.fortunecity.com/rivendell...da/vasfda.html
shows a circuit schematic with separate chokes in the two lines and
says specifically "the HT Supply circuit operates in choke input mode".
He adds, "Using one choke in each 'leg' offers improved common mode
rejection and doubles the Inductance."

I'm rather enjoying this thread, in the hope of learning something new.

Andre Jute

In article . com, "Andre
Jute" wrote:

Gentlemen:

After measurement on circuits and simulation, I have decided to make
no change to the power supplies of my circuits offered for comment:
http://members.lycos.co.uk/fiultra/T...trafi-crct.jpg
http://members.lycos.co.uk/fiultra/T44bis-'Populaire'-crct.jpg
http://members.lycos.co.uk/fiultra/t...17acircuit.jpg
I am of course still, and always, open to reasoned argument to the
contrary.

Here are some conclusions on common mode chokes, and more specifically
on common mode chokes in choke input power supply filters, offered for
further comment, clarification and kibbitzing:

1. A common mode choke is two closely coupled coils on the same core,
intended to be inserted in the power and ground leads of a supply (or
in signal-carrying and ground leads) for the rejection of common mode
noise.

2. Two separate chokes, one on each line, does not have the same effect
as a common mode choke of the same value. A simple test is to observe
on an oscilloscope a square wave through a common mode choke. The
signal is not visibly degraded except marginally at the very highest
magnification. With two separate similar chokes, one to each line, the
squarewave comes out looking like a mud hut after a monsoon, with all
the corners completely knocked off. Electrically, two separate chokes,
one on each line, are equivalent to two separate chokes in parallel on
one line (that is, a reduction in the total inductance), whereas a
common mode choke, one winding to each line, is equivalent to the two
coils in series on one line (that is, an increase in the total
inductance).

3. Common mode chokes are an economic method for suppression of noise
under various codes; Australian Standard AS 2279 'Harmonic
disturbance' is an example, or try the similar European VDE
Specification 0160 if that is more conveniently accessible. Such common
mode chokes are *commonly* inserted in the choke input position. This,
for instance, is from the circuit description of Zener Electric's
VSC2000 power train for their widely used variable speed motor
controller: "The DC choke (or inductor) is included in the input
circuit of the DC bus between the rectifier and the DC bus
capacitor." A clearer description of a choke-input common-mode choke
one cannot wish for.

4. A common mode choke is effective in reducing power supply noise,
among other parameters, in proportion to the closeness of the coupling
of the twin coils about the single core. The fact that coupling is
never 100% in practice in audio work blocks radio frequency
interference (RFI). The common mode impedance created by less than 100%
coupling also blocks electromagnetic emissions (EMI). Serendipity
strikes again.

5. With or without a common mode choke, parasitic capacitances from the
primary side of the transformer and from the load couple through the
ground reference. Common mode chokes also reduce the noise transmitted
by these stray capacitances. (1)

6. I have used common mode chokes in choke input power supplies on my
amplifiers for over a decade and have never observed any adverse
effects. Quite the contrary. The use of a choke of any kind, including
a common mode choke, on the ground line appears to act as a barrier to
various manners of noise. I have neither the knowledge nor the skill
with test equipment, nor for that matter the inclination or the time,
to separate the "slowing" effects of a choke, any choke, from the
"common core" effects of a common mode choke, but am happy to
accept any additional reduction of noise in my amps whatever the cause.

7. The Ryder materials offered by John Stewart are an unsupported,
overblown scare story. Ryder claims that a choke in the ground lead of
a choke input filter through the secondary to primary capacitance of
the power transformer puts at least half the transformer secondary (in
the case of a centretapped secondary) AC in series with choke through
the ground line. I have been unable to measure this on real-life amps,
have never observed it in more than a decade of using common mode
chokes in choke input filters and, while I don't doubt that there is
a parasitic capacitance as mentioned by Ryder, I doubt the magnitude of
its effect as described by him: hundreds of volts of AC ripple in the
filtered supply. I suppose it is possible in the sense that anything is
possible, but then I don't build a barrier against the *possibility*
that The Titanic will surface, sail down the Irish Channel, cross six
miles of land (uphill), and crash into my house to smash all my
amplifiers. I don't live in that parallel universe - and in this
one I feel quite secure, as apparently does tens of thousands of
professional engineers with liability lawyers looking over their
shoulders, in using common mode chokes in choke input supplies.
Significantly, those who appeal to Ryder's authority cannot explain
how his effects are achieved; clumsy sneering and jeering cuts no ice
with me; the only result is publicly to embarrass those too thick to
explain what they claim by faith in their priesthood.

8. I shall return to the noise hunt in the KISS amps by other means in
another thread.

Thanks you to those who contributed worthwhile thoughts and questions
so far and I look forward to your further contributions.

Andre Jute

(1) There are other means than common mode chokes, or in addition to
common mode chokes, of keeping power supply noise out of the load by
going well beyond the KISS principles my audio amps normally adhere to.
For instance, in a design intended for production, by connecting a
small capacitor from the signal line to the case and another between
ground line and case, it is possible to return noise to its source:
****
noise on signal or ground line to capacitor to metallic
casing to stray capacitance to noise source
****
The problem for the DIYer with this scheme is that, at least in the EU,
you may then have to cover the metal case so that it is nowhere
connected to anything the user can touch. (Iain may know more about
this. I just don't use this method any more.) Control knob shafts,
switch bezels, switch toggles, screws, vent grilles... The list of
metal components you must replace is pretty long.

Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review


Surf my web pages at, http://users.rcn.com/jbyrns/

In article . com, "Andre
Jute" wrote:

John Byrns wrote:
Andre,

As I alluded to in an earlier post, a common mode choke connected between
a rectifier and a reservoir capacitor does not make a choke input filter,
and on the other hand if you phase the two windings, one in the negative
lead and the other in the positive lead, such that you have a choke input
filter, then you loose the common mode rejection capabilities of the
common mode choke connection. You can't have both effects at the same
time, and which you get depends on the relative phasing of the two
windings of the choke.

Regards,

John Byrns


I don't get his, John. My LL1638 have four terminals. Connected one way
they are a single choke, two coils in series on a single core.
Connected the other way they are a common mode choke, two coils in
different leads on a single core.

I do not see the reason why in the first connection, inserted on the
signal line, that cannot be a choke input filter.

Nor do I understand, if the two coils are connected one to each line
around the common core, why connection as a common mode choke in the
choke input position should preclude operation as a common mode choke.

In each case you appear to be saying that what the thing is a connected
as is precisely what the thing won't do. If there are subtleties here,
they are presently beyond my poor brain.

I should be grateful if you would provide some reasoning or a source I
can look up.

I've found only one audio-specific common-mode choke input design
(alleged) on the net at:
http://www.musicalsurroundings.com/aesthx2.html
where it is claimed "The primary high voltage supply is composed of a
similar massive 170 VA transformer which feeds a common-mode choke
input filtering network" -- but who can tell what they think is "a
common-mode choke input" without the schematic?

Thorsten Loesch, who is well known to me, here (scroll down a long way
to the circuit):
http://www.fortunecity.com/rivendell...da/vasfda.html
shows a circuit schematic with separate chokes in the two lines and
says specifically "the HT Supply circuit operates in choke input mode".
He adds, "Using one choke in each 'leg' offers improved common mode
rejection and doubles the Inductance."

I'm rather enjoying this thread, in the hope of learning something new.

Andre Jute


Hi Andre,

I will try to give you a little simple insight into what is going on with
the two different connections of the Lundahl two winding choke. But first
for Plodnick and all the other kibitzers and hangers on let me note that I
am not a professional writer like Ryder, so please don't complain if my
efforts are somewhat lacking, note that even Ryder's prose was as opaque
as a sheet of steel, and he was a professional writer.

Also note that for ease of understanding the discussion below ignores all
"stray" capacitances, although they will exist in a real circuit.

First consider a choke input filter of the ordinary sort where the purpose
of the choke between the rectifier and the first capacitor is to smooth
out the current delivery from the rectifier to the first filer capacitor.
We can take this choke and cut the winding into two parts as in the
Lundahl choke, and if we then reconnect the two cut ends with a jumper we
will once again have a working choke input filter. Next observing that
the current is the same everywhere in a series circuit, and ignoring the
effects of stray capacitance, we can rearrange the series circuit
consisting of the transformer/rectifier, two choke windings, and filter
capacitor/load so as to place one of the two choke windings in the
negative lead of the power supply, as long as we maintain the same
direction of current flow through all the series components as it was in
the original arrangement. So we have moved one winding of our two winding
choke to the negative lead and we still have a working choke input power
supply, with half the total winding in the positive lead, and half in the
negative lead.

Next consider what happens if we go back to the original connection of our
two winding choke where the two windings were directly connected with a
jumper. What happens if we reverse the connections to one of the two
windings? What we end up with is no longer a choke with a large value of
inductance, the choke now has only a small inductance dependent on the
leakage inductance of the two windings, or how tightly they are coupled,
in series with the resistance of the wire in the windings. With this
negligible value of inductance we no longer have a choke input filter, we
essentially have a capacitor input filter. I don't have one of these
Lundahl chokes to try it with, but when you reverse the connection of one
winding like this the DC output voltage under load should increase
considerably, assuming tight coupling between the two windings. So now
all we have is a capacitor input filter and we might just as well throw
out the expensive choke and save some money.

But now consider what happens if we keep the choke anyway, and using our
series circuit rule slide one of the windings to the negative end of the
rectifier, keeping the series current flowing the same direction through
the translated winding as it was with the two windings directly connected
in the positive lead. While we still have a capacitor input filter, the
choke has now become a "common mode" choke. What does a "common mode"
choke do for us? It allows a "common mode" voltage to be impressed on the
transformer/rectifier block without affecting the output of the supply, in
other we can grab hold of the transformer and rectifier and drive it,
including both the positive and negative terminals together, up and down
without affecting the output of the supply.

The bottom line is that with half the winding in the positive lead from
the rectifier, and the other half in the negative lead, whether you have a
choke input filter, or a capacitor input filter with a common mode choke,
depends on the relative phasing of the two windings on the choke.

The key to all this was given in this quote you posted earlier.

I have a private letter from another qualified engineer, who has kindly
given me permission to quote it as long as his name is not dragged into
the inevitable RAT flamewar.

Then he offers an explanation: "Allow me to explain. The two chokes as
connected have their two windings effectively connected in series
aiding. In other words, the DC flux from both windings adds in the
core. Because of this connection the only common mode inductance that
the choke will provide is that due to the leakage inductance between
its two windings. I looked on Lundahl's website for this spec, and
couldn't find it, but I'll bet that the leakage inductance is pretty
low due to the C-Core construction and long low windings. Without the
specifications for both inter and intrawinding capacitance as well as
capacitance to frame for the [LL]1638s, it's impossible to analyze, but
it's certainly possible that your filter could be passing certain bands
of common-mode noise very nicely."


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

John Byrns wrote:

In article . com, "Andre
Jute" wrote:

John Byrns wrote:
Andre,

As I alluded to in an earlier post, a common mode choke connected between
a rectifier and a reservoir capacitor does not make a choke input filter,
and on the other hand if you phase the two windings, one in the negative
lead and the other in the positive lead, such that you have a choke input
filter, then you loose the common mode rejection capabilities of the
common mode choke connection. You can't have both effects at the same
time, and which you get depends on the relative phasing of the two
windings of the choke.

Regards,

John Byrns


I don't get his, John. My LL1638 have four terminals. Connected one way
they are a single choke, two coils in series on a single core.
Connected the other way they are a common mode choke, two coils in
different leads on a single core.

I do not see the reason why in the first connection, inserted on the
signal line, that cannot be a choke input filter.

Nor do I understand, if the two coils are connected one to each line
around the common core, why connection as a common mode choke in the
choke input position should preclude operation as a common mode choke.

In each case you appear to be saying that what the thing is a connected
as is precisely what the thing won't do. If there are subtleties here,
they are presently beyond my poor brain.

I should be grateful if you would provide some reasoning or a source I
can look up.

I've found only one audio-specific common-mode choke input design
(alleged) on the net at:
http://www.musicalsurroundings.com/aesthx2.html
where it is claimed "The primary high voltage supply is composed of a
similar massive 170 VA transformer which feeds a common-mode choke
input filtering network" -- but who can tell what they think is "a
common-mode choke input" without the schematic?

Thorsten Loesch, who is well known to me, here (scroll down a long way
to the circuit):
http://www.fortunecity.com/rivendell...da/vasfda.html
shows a circuit schematic with separate chokes in the two lines and
says specifically "the HT Supply circuit operates in choke input mode".
He adds, "Using one choke in each 'leg' offers improved common mode
rejection and doubles the Inductance."

I'm rather enjoying this thread, in the hope of learning something new.

Andre Jute

Hi Andre,

I will try to give you a little simple insight into what is going on with
the two different connections of the Lundahl two winding choke. But first
for Plodnick and all the other kibitzers and hangers on let me note that I
am not a professional writer like Ryder, so please don't complain if my
efforts are somewhat lacking, note that even Ryder's prose was as opaque
as a sheet of steel, and he was a professional writer.

Also note that for ease of understanding the discussion below ignores all
"stray" capacitances, although they will exist in a real circuit.

First consider a choke input filter of the ordinary sort where the purpose
of the choke between the rectifier and the first capacitor is to smooth
out the current delivery from the rectifier to the first filer capacitor.
We can take this choke and cut the winding into two parts as in the
Lundahl choke, and if we then reconnect the two cut ends with a jumper we
will once again have a working choke input filter. Next observing that
the current is the same everywhere in a series circuit, and ignoring the
effects of stray capacitance, we can rearrange the series circuit
consisting of the transformer/rectifier, two choke windings, and filter
capacitor/load so as to place one of the two choke windings in the
negative lead of the power supply, as long as we maintain the same
direction of current flow through all the series components as it was in
the original arrangement. So we have moved one winding of our two winding
choke to the negative lead and we still have a working choke input power
supply, with half the total winding in the positive lead, and half in the
negative lead.

Next consider what happens if we go back to the original connection of our
two winding choke where the two windings were directly connected with a
jumper. What happens if we reverse the connections to one of the two
windings? What we end up with is no longer a choke with a large value of
inductance, the choke now has only a small inductance dependent on the
leakage inductance of the two windings, or how tightly they are coupled,
in series with the resistance of the wire in the windings. With this
negligible value of inductance we no longer have a choke input filter, we
essentially have a capacitor input filter. I don't have one of these
Lundahl chokes to try it with, but when you reverse the connection of one
winding like this the DC output voltage under load should increase
considerably, assuming tight coupling between the two windings. So now
all we have is a capacitor input filter and we might just as well throw
out the expensive choke and save some money.

But now consider what happens if we keep the choke anyway, and using our
series circuit rule slide one of the windings to the negative end of the
rectifier, keeping the series current flowing the same direction through
the translated winding as it was with the two windings directly connected
in the positive lead. While we still have a capacitor input filter, the
choke has now become a "common mode" choke. What does a "common mode"
choke do for us? It allows a "common mode" voltage to be impressed on the
transformer/rectifier block without affecting the output of the supply, in
other we can grab hold of the transformer and rectifier and drive it,
including both the positive and negative terminals together, up and down
without affecting the output of the supply.

The bottom line is that with half the winding in the positive lead from
the rectifier, and the other half in the negative lead, whether you have a
choke input filter, or a capacitor input filter with a common mode choke,
depends on the relative phasing of the two windings on the choke.

Well, you have covered some ground to illustrate there ain't a free lunch to be
had with any PS.

I just cannot see a reason to use a dual winding choke.

Patrick Turner.



The key to all this was given in this quote you posted earlier.

I have a private letter from another qualified engineer, who has kindly
given me permission to quote it as long as his name is not dragged into
the inevitable RAT flamewar.

Then he offers an explanation: "Allow me to explain. The two chokes as
connected have their two windings effectively connected in series
aiding. In other words, the DC flux from both windings adds in the
core. Because of this connection the only common mode inductance that
the choke will provide is that due to the leakage inductance between
its two windings. I looked on Lundahl's website for this spec, and
couldn't find it, but I'll bet that the leakage inductance is pretty
low due to the C-Core construction and long low windings. Without the
specifications for both inter and intrawinding capacitance as well as
capacitance to frame for the [LL]1638s, it's impossible to analyze, but
it's certainly possible that your filter could be passing certain bands
of common-mode noise very nicely."

Regards,

John Byrns

Surf my web pages at, http://users.rcn.com/jbyrns/

John Byrns wrote:
In article . com, "Andre
Jute" wrote:

John Byrns wrote:
Andre,

As I alluded to in an earlier post, a common mode choke connected between
a rectifier and a reservoir capacitor does not make a choke input filter,
and on the other hand if you phase the two windings, one in the negative
lead and the other in the positive lead, such that you have a choke input
filter, then you loose the common mode rejection capabilities of the
common mode choke connection. You can't have both effects at the same
time, and which you get depends on the relative phasing of the two
windings of the choke.

Regards,

John Byrns

[Material snipped because what John replies to intervening
correspondence flows straight on from what he said in the first
instance:]

Hi Andre,

I will try to give you a little simple insight into what is going on with
the two different connections of the Lundahl two winding choke. But first
for Plodnick and all the other kibitzers and hangers on let me note that I
am not a professional writer like Ryder, so please don't complain if my
efforts are somewhat lacking, note that even Ryder's prose was as opaque
as a sheet of steel, and he was a professional writer.

Also note that for ease of understanding the discussion below ignores all
"stray" capacitances, although they will exist in a real circuit.

First consider a choke input filter of the ordinary sort where the purpose
of the choke between the rectifier and the first capacitor is to smooth
out the current delivery from the rectifier to the first filer capacitor.
We can take this choke and cut the winding into two parts as in the
Lundahl choke, and if we then reconnect the two cut ends with a jumper we
will once again have a working choke input filter. Next observing that
the current is the same everywhere in a series circuit, and ignoring the
effects of stray capacitance, we can rearrange the series circuit
consisting of the transformer/rectifier, two choke windings, and filter
capacitor/load so as to place one of the two choke windings in the
negative lead of the power supply, as long as we maintain the same
direction of current flow through all the series components as it was in
the original arrangement. So we have moved one winding of our two winding
choke to the negative lead and we still have a working choke input power
supply, with half the total winding in the positive lead, and half in the
negative lead.

Next consider what happens if we go back to the original connection of our
two winding choke where the two windings were directly connected with a
jumper. What happens if we reverse the connections to one of the two
windings? What we end up with is no longer a choke with a large value of
inductance, the choke now has only a small inductance dependent on the
leakage inductance of the two windings, or how tightly they are coupled,
in series with the resistance of the wire in the windings. With this
negligible value of inductance we no longer have a choke input filter, we
essentially have a capacitor input filter. I don't have one of these
Lundahl chokes to try it with, but when you reverse the connection of one
winding like this the DC output voltage under load should increase
considerably, assuming tight coupling between the two windings. So now
all we have is a capacitor input filter and we might just as well throw
out the expensive choke and save some money.

But now consider what happens if we keep the choke anyway, and using our
series circuit rule slide one of the windings to the negative end of the
rectifier, keeping the series current flowing the same direction through
the translated winding as it was with the two windings directly connected
in the positive lead. While we still have a capacitor input filter, the
choke has now become a "common mode" choke. What does a "common mode"
choke do for us? It allows a "common mode" voltage to be impressed on the
transformer/rectifier block without affecting the output of the supply, in
other we can grab hold of the transformer and rectifier and drive it,
including both the positive and negative terminals together, up and down
without affecting the output of the supply.

The bottom line is that with half the winding in the positive lead from
the rectifier, and the other half in the negative lead, whether you have a
choke input filter, or a capacitor input filter with a common mode choke,
depends on the relative phasing of the two windings on the choke.

The key to all this was given in this quote you posted earlier.

I have a private letter from another qualified engineer, who has kindly
given me permission to quote it as long as his name is not dragged into
the inevitable RAT flamewar.

Then he offers an explanation: "Allow me to explain. The two chokes as
connected have their two windings effectively connected in series
aiding. In other words, the DC flux from both windings adds in the
core. Because of this connection the only common mode inductance that
the choke will provide is that due to the leakage inductance between
its two windings. I looked on Lundahl's website for this spec, and
couldn't find it, but I'll bet that the leakage inductance is pretty
low due to the C-Core construction and long low windings. Without the
specifications for both inter and intrawinding capacitance as well as
capacitance to frame for the [LL]1638s, it's impossible to analyze, but
it's certainly possible that your filter could be passing certain bands
of common-mode noise very nicely."


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

This is bril, John. It is exactly what I was waiting for. It appears to
be most of what I need to sort out the confusion of years practice,
measurements, what the late sainted Bill told me, mislabelled specs,
and insufficiently closely supervised reverse engineering. I'll get
back to you after study of many pieces of paper on a corkboard. You
will be surprised at what together we have discovered.

Andre Jute

Just to confuse things further, most PC power supplies use one to three
dual or triple winding chokes. Usually one or two in the AC line
power supply, another in the low-voltage output circuit.

The ones in the HV power part have the windings phased for common-mode
noise isolation.
The one in the output circuit has all two or three windings in the +5
and +12 volt lines. Not a clue what the advantage is here, as it seems
to force the ripple voltages to be locked to each other. Weird.