Philips EL86 SEPP OTL Double Chokes Information

[Lemoncha]

Hi group,

I bought some EL86 recently, and want to build an OTL amplifier for my
800 ohm speakers. Does anyone have some cheap, quality 60Hy double
chokes to spare for this application, or can anyone tell me where to
have them made? I discussed this matter with Sowter Transformers but
they don't know what a double choke is. The Philips book says to use
60H double chokes. There is also a EL86 SPP article in the May 2003
issue of Audioxpress which tells me to use 30H (I think there will be
less bass if a 30H DC is used). Both authors never specify the
details about this double choke. I need more details about them, ie
winding details.

Thanks a lot!

[Ross Matheson]

(Lemoncha) wrote:

: Hi group,
:
: I bought some EL86 recently, and want to build an OTL amplifier for my
: 800 ohm speakers. Does anyone have some cheap, quality 60Hy double
: chokes to spare for this application, or can anyone tell me where to
: have them made? I discussed this matter with Sowter Transformers but
: they don't know what a double choke is. The Philips book says to use
: 60H double chokes. There is also a EL86 SPP article in the May 2003
: issue of Audioxpress which tells me to use 30H (I think there will be
: less bass if a 30H DC is used). Both authors never specify the
: details about this double choke. I need more details about them, ie
: winding details.
:
: Thanks a lot!

These are the double chokes to feed the SEPP screen supplies?
I only photocopied the inroduction and a few of the PP amps in the Phillips
(or Mullard - by Rodenhuis?) book when I had it on loan, and didn't get the
SEPP circuit values - although there is a negative image of the bare circuit
on the cover of the one I am thinking of. I didn't know they were specified at
60H, for example. If your book is the one that has an analysis of the circuit
and the problems feeding the screen (for which the double choke is the best
solution) why not show Sowter that page, or pages?

My *guess* is that a bifilar choke might do the trick - they could do that.
If they understand the reasoning given in those pages they would get it, IMO.
Essentially the two windings feeding each screen have DC current in opposing
directions, so the magnetic field cancels, and no gap is required, IIRC.

If your book is different and doesn't have that analysis I can scan and post.
RdM

[Ross Matheson]

"Ronald" > wrote:

: Curiosity killed the cat , so kill me with that file ;-)

On it's way as I write ...

[Ronald]

Thanks !!

I'll take a look ....

Quote from your mail :
> Wish I had copied the actual circuit section - but I doubted that I'd ever
find
> 800ohm speakers, and it still seems unlikely;-)

It's not that hard to put a tranny in . It can be a normal PSU tranny since
there's
no DC current . A 220V pri. - 22V sec. will work for an 8 ohms speaker .
Might do nicely for multy-media .......
800 ohms speakers are very common in old Philips tube-radio's over here in
NL .

Ronald .

"Ross Matheson" > schreef in bericht
...
> "Ronald" > wrote:
>
> : Curiosity killed the cat , so kill me with that file ;-)
>
> On it's way as I write ...

[Ronald]

"Lemoncha" > schreef in bericht
om...
> Hi group,
>
> I bought some EL86 recently, and want to build an OTL amplifier for my
> 800 ohm speakers. Does anyone have some cheap, quality 60Hy double
> chokes to spare for this application, or can anyone tell me where to
> have them made? I discussed this matter with Sowter Transformers but
> they don't know what a double choke is. The Philips book says to use
> 60H double chokes. There is also a EL86 SPP article in the May 2003
> issue of Audioxpress which tells me to use 30H (I think there will be
> less bass if a 30H DC is used). Both authors never specify the
> details about this double choke. I need more details about them, ie
> winding details.
>
> Thanks a lot!

I had a look at the file Ross was talking about and there it says 2 x 5H is
enough
for 30Hz .
So the first thing that came to mind was using 2 power trannies and try to
get both
primaries on one bobine/core .
That wouldn't be to hard if the trannies have 2 chambers and a core that can
be taken
appart (not welded or something) .

Ronald .

[Ross Matheson]

"Ronald" > wrote:

: Thanks !!
:
: I'll take a look ....

You're welcome:)

: Quote from your mail :
: > Wish I had copied the actual circuit section - but I doubted that I'd ever
: find
: > 800ohm speakers, and it still seems unlikely;-)
:
: It's not that hard to put a tranny in . It can be a normal PSU tranny since
: there's
: no DC current . A 220V pri. - 22V sec. will work for an 8 ohms speaker .

I suppose that's a point, although I doubt much useful HF from such.
I've saved a 26V (and 28V, too) xfmr, but for the purpose of a filament
transformer for a quad of 6159B, (26.5Vfil 6146) when I get around to it!

: Might do nicely for multy-media .......

Like freq-resp limited rear channels, as in Dolby Pro-Logic or simpler:-?)

: 800 ohms speakers are very common in old Philips tube-radio's over here in
: NL .

I've never seen any here, although not to say none came out ...
I do have a pair of "full range" 12" AD1256-M8 drive units though. In other
discussions I've been told their bass is not so great compared to 1256W.
My Philips speaker book doesn't cover them, although it has the 1256W.
I guess the "M" essentially means "midrange", but they might be fun ...

But back to the subject - how do you reckon these chokes are constructed?
The diagram shows them looking like a 1:1 transformer - in a sense it is.

"As the direct currents flow through the windings of the choke in opposite
directions, the magnetic fields neutralise each other and the core is thus not
magnetised; there is thus no need for an air-gap, resulting in a component
which is small but which nevertheless has a high inductance."

This shouldn't be too hard to make - and the screen currents are not high.

[Ian Iveson]

Although I don't have the circuit, I assume bandwidth is important?
If so, then the capacitance and leakage of the tranny/choke must be
considered. Maybe leakage isn't crucial...depends.

Simulation would be dead easy for anyone with a circuit and the
right valve models. Unfortunately no EL86 in my library.

As Ross says, Sowter could undoubtedly help if they knew what it is
for. They may well have a stock interstage transformer that would
do the job, or they can wind an OPT without the secondaries,
presumably. They don't charge extra for one-offs, or for advice.
They can take a long time to deliver though...

cheers, Ian


"Ronald" > wrote in message
...
>
> "Lemoncha" > schreef in bericht
> om...
> > Hi group,
> >
> > I bought some EL86 recently, and want to build an OTL amplifier
for my
> > 800 ohm speakers. Does anyone have some cheap, quality 60Hy
double
> > chokes to spare for this application, or can anyone tell me
where to
> > have them made? I discussed this matter with Sowter
Transformers but
> > they don't know what a double choke is. The Philips book says
to use
> > 60H double chokes. There is also a EL86 SPP article in the May
2003
> > issue of Audioxpress which tells me to use 30H (I think there
will be
> > less bass if a 30H DC is used). Both authors never specify the
> > details about this double choke. I need more details about
them, ie
> > winding details.
> >
> > Thanks a lot!
>
> I had a look at the file Ross was talking about and there it says
2 x 5H is
> enough
> for 30Hz .
> So the first thing that came to mind was using 2 power trannies
and try to
> get both
> primaries on one bobine/core .
> That wouldn't be to hard if the trannies have 2 chambers and a
core that can
> be taken
> appart (not welded or something) .
>
> Ronald .
>
>

[Ian Iveson]

"Ronald" > wrote

> If you test with cheap trannies first , you better know what to
order I
> think .....
>
Ah but Ronald then it will either work OK or not. If not, you will
have no extra clues about why not. Hence no further forward. Cheap
trannies would be useful to establish that the basic idea is
workable, but we know that already.

With simulation you can alter parameters to see what improvements
result, jiggle until you get it right, and then order with
confidence from a winder who knows how to translate from performance
requirements to manufacturing parameters.

With correct analysis you can predict requirements and should not
need to experiment.

Brian Sowter knows his stuff, BTW, and if he is unsure he will say
so.

Just my view.

cheers, Ian


> Hi Ian ,
>
> > Although I don't have the circuit, I assume bandwidth is
important?
> > If so, then the capacitance and leakage of the tranny/choke must
be
> > considered. Maybe leakage isn't crucial...depends.
>
> Might be , but using normal trannies is cheap so testing with it
won't hurt
> much ....
>
> > Simulation would be dead easy for anyone with a circuit and the
> > right valve models. Unfortunately no EL86 in my library.
>
> I don't simulate , but maybe PL84 is an option . Only the heaters
are
> different .
>
> > As Ross says, Sowter could undoubtedly help if they knew what it
is
> > for. They may well have a stock interstage transformer that
would
> > do the job, or they can wind an OPT without the secondaries,
> > presumably. They don't charge extra for one-offs, or for
advice.
> > They can take a long time to deliver though...
>
> > cheers, Ian
>
> Best regards ,
>
> Ronald .
>
>
>

[Ronald]

And if you can read German , this might be interesting :
http://www.jogis-roehrenbude.de/Leserbriefe/Hans-Heyligers-Amp.htm

[Ian Iveson]

Thanks, Ronald. No I don't read German (wish I did...only foreign
language I know even a bit of is Latin, unfortunately. I blame the
Jesuits.) but I can read the pictures and the flow of equations :-)
Now I remember what a SEPP is and hey, I have the parts to build one
right now, with ECC81 and a spare pair of 6CH6. PS would be awkward
though. No chance of driving these speakers either, says the humble
Choky.

But where does the "double choke come in"? Still don't quite know
what I'm on about. Looking through the thread I can't see a current
link to that circuit. Coupling the screens? At a guess only
problem might be capacitance.

I found with my simulated parafeed circlotron, and my various
simulated totem-poles including SEPP and mu-follower, that they
seemed very wonderful in almost every respect, but there is always a
quirk. Some are too easily provoked into current limiting, and some
have output impedance that behaves strangely. And some seem like
they can source current but can't sink it so good. Perhaps I just
need more volts. Transformers that may seem adequate with casual
investigation can turn out to be not so.

One mechanism is where low-frequency response is extended by
feedback...particularly cathode feedback which reduces Vak...so
early clipping results at low frequencies if you aren't careful, and
also possibly slew-rate limiting for roughly the same reason. A
related problem with cheap trannies is saturation. Another is
where leakage doesn't seem important until you consider
high-frequency output impedance.

None of which may apply to the mysterious double chokes.

I don't mean to detract from the fun of experimentally discovering
such things, and trying to work out why. I just don't like the
smell of solder ;-) Simulation is fun too. And easy and cheap and
safe.

In both cases, however, if you don't grasp the maths, you're
stumbling. Transformers have too many equations bundled together to
be encompassed by a single thought, and sometimes I get lost.

best wishes

Ian

[Ross Matheson]

"Ian Iveson" > wrote:

: Thanks, Ronald. No I don't read German (wish I did...only foreign
: language I know even a bit of is Latin, unfortunately. I blame the
: Jesuits.) but I can read the pictures and the flow of equations :-)

Here's a link to the first part of the page translated by Google
(unfortunately Google translations have a word/character limit)
http://tinyurl.com/i3fi

But he doesn't use a double screen choke, and appears in fact to be
running them as triodes with 1k (G2) grid stoppers.

: Now I remember what a SEPP is and hey, I have the parts to build one
: right now, with ECC81 and a spare pair of 6CH6. PS would be awkward
: though. No chance of driving these speakers either, says the humble
: Choky.
:
: But where does the "double choke come in"? Still don't quite know
: what I'm on about. Looking through the thread I can't see a current
: link to that circuit. Coupling the screens? At a guess only
: problem might be capacitance.

I've just emailed you the pdf discussion pages and the basic circuit gif.
I've also just posted the basic circuit (no values) gif on a.b.s.e.

I'll see if I can figure out a way to reduce the size of the discussion
pages for an easier download post, as I suspect a ~390kb pdf might be too
large for some servers (although I did post it once late last year, and
the binaries archive did pick it up and hold it for some time at the time)
<http://www.usenet-replayer.com/groups/alt%2Ebinaries%2Eschematics%2Eelectronic.en.html.g z>
OTOH, just doing a search there (choosing "More" and selecting pdf only,
long time-base, 300 results) shows at least a 449kb pdf archived, so what
the hell, I may just post it again as-is if I can't reduce the filesize.
Look out for it soon ... (not quite yet) [ie earlier the one is long gone]

[Not suggesting anybody bother with that search, but perhaps a reminder of
this useful archive and its search functions is timely for future ref.:-)]

Regards, Ross

[Ross Matheson]

: what the hell, I may just post it again as-is if I can't reduce the filesize.
: Look out for it soon ... (not quite yet) [ie earlier the one is long gone]

Done. A bit tough at 7044 lines for anyone who has auto-download set
(but I reckon you'd be a bit mad to on a binaries group anyway:-)

Once again, the archive at
<http://www.usenet-replayer.com/groups/alt%2Ebinaries%2Eschematics%2Eelectronic.en.html.g z>
should have it shortly ...

[Henry 007]

"Ronald" > wrote in message
...
>
> > : It's not that hard to put a tranny in . It can be a normal PSU tranny
> since
> > : there's
> > : no DC current . A 220V pri. - 22V sec. will work for an 8 ohms speaker
..
> >
> > I suppose that's a point, although I doubt much useful HF from such.
> > I've saved a 26V (and 28V, too) xfmr, but for the purpose of a filament
> > transformer for a quad of 6159B, (26.5Vfil 6146) when I get around to
it!
> >
> > : Might do nicely for multy-media .......
> >
> > Like freq-resp limited rear channels, as in Dolby Pro-Logic or
simpler:-?)
>
> Something like that ......
>
> > I guess the "M" essentially means "midrange", but they might be fun ...
>
> "M" was also used for full-range . "W" was a "real" woofer .
> You might like the AD1256-M8 in a classic transmision line . You just need
> the effective cone area and F-res. to design one ......
>
> > But back to the subject - how do you reckon these chokes are
constructed?
> > The diagram shows them looking like a 1:1 transformer - in a sense it
is.
>
> That's how I looked at it ....
> It might just be the cheap way to test things first ...
>
> > "As the direct currents flow through the windings of the choke in
opposite
> > directions, the magnetic fields neutralise each other and the core is
thus
> not
> > magnetised; there is thus no need for an air-gap, resulting in a
component
> > which is small but which nevertheless has a high inductance."
> >
> > This shouldn't be too hard to make - and the screen currents are not
high.
>
> I think I got some ideas too by reading the article ...... I have some NOS
> PL504's .....
>
> Best regards ,
>
> Ronald .
>
>

If its any help at all, I have an old philips stereo thing with 2 800 ohm
full-rangers in it, and all the appropriate circuitry... I have scavenged
the power tranny and the EL84s that originally came in it, but the rest is
due to be thrown out...

[Ross Matheson]

"Henry 007" > wrote:

: If its any help at all, I have an old philips stereo thing with 2 800 ohm
: full-rangers in it, and all the appropriate circuitry... I have scavenged
: the power tranny and the EL84s that originally came in it, but the rest is
: due to be thrown out...

Sheesh, I'm keen, Henry!!!
You're in Wellington, aren't you?
Maybe we could work out a postage/packaging deal??
I've got EL86's - I'm into mad retro-tech experiments.
Just the speakers would do, most probably - and easy to send.
Please, send me an email and let's see what we can work out!
Regards, Ross

[Ronald]

Hi Ian ,

The guy is talkink High-End with normal PSU iron .
So it should be at least Hi-Fi I think .
I also have good experience using C-core PSU trannies in PP .
So at least this will be good enough for a low budget experiment I think
.....

I also acted a bit (to) quick sending you the same file Ross did .
I saw that to late . Hope you don't mind getting it twice .....

Ronald .

"Ian Iveson" > schreef in bericht
...
> Thanks, Ronald. No I don't read German (wish I did...only foreign
> language I know even a bit of is Latin, unfortunately. I blame the
> Jesuits.) but I can read the pictures and the flow of equations :-)
> Now I remember what a SEPP is and hey, I have the parts to build one
> right now, with ECC81 and a spare pair of 6CH6. PS would be awkward
> though. No chance of driving these speakers either, says the humble
> Choky.
>
> But where does the "double choke come in"? Still don't quite know
> what I'm on about. Looking through the thread I can't see a current
> link to that circuit. Coupling the screens? At a guess only
> problem might be capacitance.
>
> I found with my simulated parafeed circlotron, and my various
> simulated totem-poles including SEPP and mu-follower, that they
> seemed very wonderful in almost every respect, but there is always a
> quirk. Some are too easily provoked into current limiting, and some
> have output impedance that behaves strangely. And some seem like
> they can source current but can't sink it so good. Perhaps I just
> need more volts. Transformers that may seem adequate with casual
> investigation can turn out to be not so.
>
> One mechanism is where low-frequency response is extended by
> feedback...particularly cathode feedback which reduces Vak...so
> early clipping results at low frequencies if you aren't careful, and
> also possibly slew-rate limiting for roughly the same reason. A
> related problem with cheap trannies is saturation. Another is
> where leakage doesn't seem important until you consider
> high-frequency output impedance.
>
> None of which may apply to the mysterious double chokes.
>
> I don't mean to detract from the fun of experimentally discovering
> such things, and trying to work out why. I just don't like the
> smell of solder ;-) Simulation is fun too. And easy and cheap and
> safe.
>
> In both cases, however, if you don't grasp the maths, you're
> stumbling. Transformers have too many equations bundled together to
> be encompassed by a single thought, and sometimes I get lost.
>
> best wishes
>
> Ian
>
>
>

[John Stewart]

I just took a look at that article in aX. No need for a hard to get
double choke. A pair of single 20H PS chokes will work OK.
One end of each choke is at AC ground while the other
is at the output AC signal potential. The two chokes are already
tightly coupled through those connexions. The original inventor
of the subject circuits objective was to avoid iron cored
devices completely.

Stray winding C is a consideration, but for an experimental
try that should not stop you. I built something similar to this
around 1960 using 20 of 6AQ5. It made an easy 100 Watts
into a 100 ohm load. I used ordinary PS chokes. Author of
this article even suggests the use of line matching transformers
as an output impedance match to regular 8 ohm speakers.
In general they are no better than a PS choke for audio
applications.

I've posted yet another version of this circuit at ABSE
for your interest.

Good Luck, John Stewart

Lemoncha wrote:

> Hi group,
>
> I bought some EL86 recently, and want to build an OTL amplifier for my
> 800 ohm speakers. Does anyone have some cheap, quality 60Hy double
> chokes to spare for this application, or can anyone tell me where to
> have them made? I discussed this matter with Sowter Transformers but
> they don't know what a double choke is. The Philips book says to use
> 60H double chokes. There is also a EL86 SPP article in the May 2003
> issue of Audioxpress which tells me to use 30H (I think there will be
> less bass if a 30H DC is used). Both authors never specify the
> details about this double choke. I need more details about them, ie
> winding details.
>
> Thanks a lot!

[Ross Matheson]

John Stewart > wrote:
in >,
on Sat, 26 Jul 2003 07:26:24 -0400,
: I just took a look at that article in aX. No need for a hard to get
: double choke. A pair of single 20H PS chokes will work OK.
: One end of each choke is at AC ground while the other
: is at the output AC signal potential. The two chokes are already
: tightly coupled through those connexions. The original inventor
: of the subject circuits objective was to avoid iron cored
: devices completely.

Are you sure? Completely? I obtained the impression it was ungapped and of
small size, but I assumed it still had a core! (Smaller than OPT's, natch:)
It'd be good to get an impression of what this dual choke looked like ...

: Stray winding C is a consideration, but for an experimental
: try that should not stop you. I built something similar to this
: around 1960 using 20 of 6AQ5. It made an easy 100 Watts
: into a 100 ohm load. I used ordinary PS chokes. Author of
: this article even suggests the use of line matching transformers
: as an output impedance match to regular 8 ohm speakers.
: In general they are no better than a PS choke for audio
: applications.
:
: I've posted yet another version of this circuit at ABSE
: for your interest.

Coincidentally, I had just gone through the usenet-replayer archive with
the longest timebase option and found that (previously posted) jpeg and
also the WW May 1952 article (a few months younger than I am:-) you also
posted earlier, *still there*! (just), which might give some impression of
it's archive life. (it doesn't appear to date them, as far as I can see...)

Also, there's a 5-part dualSE Glass Audio extract that I'd missed earlier.
A bit odd, that one. I have yet to assemble the collected Matrix Amp parts!
Regards and Cheers! - RdM.

: Good Luck, John Stewart
:
: Lemoncha wrote:
:
: > Hi group,
: >
: > I bought some EL86 recently, and want to build an OTL amplifier for my
: > 800 ohm speakers. Does anyone have some cheap, quality 60Hy double
: > chokes to spare for this application, or can anyone tell me where to
: > have them made? I discussed this matter with Sowter Transformers but
: > they don't know what a double choke is. The Philips book says to use
: > 60H double chokes. There is also a EL86 SPP article in the May 2003
: > issue of Audioxpress which tells me to use 30H (I think there will be
: > less bass if a 30H DC is used). Both authors never specify the
: > details about this double choke. I need more details about them, ie
: > winding details.
: >
: > Thanks a lot!

[John Byrns]

In article >, Ross Matheson
> wrote:

> John Stewart > wrote:
> in >,
> on Sat, 26 Jul 2003 07:26:24 -0400,
> : I just took a look at that article in aX. No need for a hard to get
> : double choke. A pair of single 20H PS chokes will work OK.
> : One end of each choke is at AC ground while the other
> : is at the output AC signal potential. The two chokes are already
> : tightly coupled through those connexions. The original inventor
> : of the subject circuits objective was to avoid iron cored
> : devices completely.
>
> Are you sure? Completely? I obtained the impression it was ungapped and of
> small size, but I assumed it still had a core! (Smaller than OPT's, natch:)
> It'd be good to get an impression of what this dual choke looked like ...

Of course he is sure, completely! The two chokes should work just fine,
you might even get away with just one if you have a convenient source at
the correct voltage for feeding the screen of the lower tube. The use of
the double choke cuts down on the amount of iron necessary, since the DC
current for the two screens flows in opposite directions, the effect of
the DC tends to cancel as in a push pull output transformer, eliminating
the need for an air gap, or at least a large air gap. Eliminating the air
gap means less iron is needed to maintain the required inductance.

I suspect that the "dual choke" looked pretty much like a very small push
pull output transformer, with four lead wires exiting the windings.

Someone posted a link to a neat design for an amplifier like the Philips
about half a year ago. Unfortunately I don't remember who the designer
was, or who originally posted the circuit. You can find a copy on my web
pages at this URL:

http://users.rcn.com/jbyrns/stuff/scan0022.png

Notice that the plate current for the tubes doesn't pass through the "dual
choke", and a secondary winding has been added to the "dual choke" to
allow the connection of an ordinary loudspeaker.


Regards,

John Byrns


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