Lets Improve John Byrns 25L6 Amplifier

[John L Stewart]

John’s short note accompanying his amplifier invites us to comment & offer suggestions. I did 7 or 8 years ago & found that John was already happy with his amplifier just the way it was then. And I guess he still is. And John comments freely on the work of others.

Anyway, here is my take on some minimal changes that would result in a substantial performance improvement.

Cosmetically the photos show a clean but crowded chassis top. Not sure of the underside, the photo does not show enough detail.

John’s hand drawn schematic would not pass on a high school boy’s test. I mentioned this to him. John said he would redraw the schema but again that was many years ago. Still the same as it was in 1998.

The power output, distortion & bandwidth specs are given nowhere. We are not told what the OPT’s are but they do look like substantial Hammond iron.

John’s amp uses broadcast quality 600R to PP grid input transformers. Very expensive indeed, but thereby gets past the need for a PT.

First mod I would recommend is to add another 1N4007 & an electrolytic to essentially double the plate supply volts with a half wave voltage doubler. That would result in doubling the audio watts available. The OPT’s look to be well up to that.

John has posted the schema of a WE 140A amp along with his amp. This appears to be where he got the idea for an amplifier without a PT. The WE140A can run on an AC or DC supply. The possibility of finding a 110 vdc supply to run John's amp on at this point in time seems rather unlikely. The WE140A makes use of a pair of 25Z6 rectifiers as a full wave doubler while running on AC.

The output stage is a pair of beam tetrodes. Damping factor for the speakers is essentially non existent, so they are sure to flop around like a fish out of water. This amp does not have enough gain available to allow an effective application NFB.

Unplugging the 12SN7 drivers I recommend the 6SL7 with 100K plate resistors.
The gain improves by about 10 db. The 25L6 does not need the drive performance available from a 12SN7.

Two resistors of 270K connected from the 25L6 plates to the corresponding 6SL7 plates gets you approx 5 to 7 db NFB, depending on supply voltages. This results in a DF of about one, what used to be called ‘Unity Damping’.

Are these simple mods worthwhile? I would say so. What do others think? And what about John B who never seems to complete the myriad projects he talks about.

Would I build this kind of amp? Not likely. A PT is far less money than a pair of expensive audio line transformers.

Cheers, John S, who is still able to ID the hot end of a soldering iron!

[Patrick Turner]

On Jun 26, 8:05*am, John Byrns > wrote:
> In article >,
> *John L Stewart > wrote:
>
> > John s short note accompanying his amplifier invites us to comment &
> > offer suggestions. I did 7 or 8 years ago & found that John was already
> > happy with his amplifier just the way it was then. And I guess he still
> > is. And John comments freely on the work of others.
>
> I don't think I ever liked it "just the way it was", there have always been a
> few details that need improving. *
>
> First, I would like to change the input tubes to the equivalent of 12AU7s,
> unfortunately I don't think there is an octal equivalent for the 12AU7, so I
> would have to somehow adapt 9 pin miniature sockets to the existing chassis
> holes. *I remember that 9 pin miniature sockets used to be available that would
> fit the cutout for an octal socket, does anyone know where I might find a couple
> of those?
>
> Second, I would like to try connecting the output tubes as triodes. *This would
> require adding an LC section to the power supply, and would unfortunately cut
> the power output to something less than the half the power output available in
> pentode mode.
>
> > Anyway, here is my take on some minimal changes that would result in a
> > substantial performance improvement.

> snip interesting read for brevity.

Hmm, with 240V mains here I won't be rushing into building any amp
without a PT.

When wanting to fit a ninepin socket into a larger 28mm octal socket
hole I just bore a 19mm hole in a bit of scrap metal and bolt it in
under the existing 28mm hole. I often have to replace ALL the 1935
tubes in some horrible old radio chassis made in 1935 and fitted with
4V heater european tubes with queer sockets. I usually strip the whole
chassis, sand it dowm and paint it with a pale greeny beige color,
Circuit is totally revised, 1950s tubes are used, and I will cut out
plates of color-bond sheet metal to overlay large areas of the chassis
with the 50mm holes for euro-tube sockets. When finished, nobody cares
about the "**** that was there" and they marvel at the sound because
I've incorperated all the best ways to build an AM radio.

Perhaps there's a parable there...

But anyway, there's no reason why the SNR for such amps can't be good
if all circuitry is referred to a 0V rail. The use of an input tranny
with a floating input coil would also help against hum. Just what
other gear was used may affect noise though, TT, preamp perhhaps.

I better get on my bike and ride out of range of the old rubbish,
boots, old bottles being hurled my way because I humbly suggested the
merits of major parts upgrading to ancient old junk.

Patrick Turner

[Engineer[_2_]]

On Jun 29, 7:34*pm, Patrick Turner > wrote:
> On Jun 26, 8:05*am, John Byrns > wrote:
>
>
>
>
>
>
>
>
>
> > In article >,
> > *John L Stewart > wrote:
>
> > > John s short note accompanying his amplifier invites us to comment &
> > > offer suggestions. I did 7 or 8 years ago & found that John was already
> > > happy with his amplifier just the way it was then. And I guess he still
> > > is. And John comments freely on the work of others.
>
> > I don't think I ever liked it "just the way it was", there have always been a
> > few details that need improving. *
>
> > First, I would like to change the input tubes to the equivalent of 12AU7s,
> > unfortunately I don't think there is an octal equivalent for the 12AU7, so I
> > would have to somehow adapt 9 pin miniature sockets to the existing chassis
> > holes. *I remember that 9 pin miniature sockets used to be available that would
> > fit the cutout for an octal socket, does anyone know where I might find a couple
> > of those?
>
> > Second, I would like to try connecting the output tubes as triodes. *This would
> > require adding an LC section to the power supply, and would unfortunately cut
> > the power output to something less than the half the power output available in
> > pentode mode.
>
> > > Anyway, here is my take on some minimal changes that would result in a
> > > substantial performance improvement.
> > snip interesting read for brevity.
>
> Hmm, with 240V mains here I won't be rushing into building any amp
> without a PT.
>
> When wanting to fit a ninepin socket into a larger 28mm octal socket
> hole I just bore a 19mm hole in a bit of scrap metal *and bolt it in
> under the existing 28mm hole. I often have to replace ALL the 1935
> tubes in some horrible old radio chassis made in 1935 and fitted with
> 4V heater european tubes with queer sockets. I usually strip the whole
> chassis, sand it dowm and paint it with a pale greeny beige color,
> Circuit is totally revised, 1950s tubes are used, and I will cut out
> plates of color-bond sheet metal to overlay large areas of the chassis
> with the 50mm holes for euro-tube sockets. When finished, nobody cares
> about the "**** that was there" and they marvel at the sound because
> I've incorperated all the best ways to build an AM radio.
>
> Perhaps there's a parable there...
>
> But anyway, there's no reason why the SNR for such amps can't be good
> if all circuitry is referred to a 0V rail. The use of an input tranny
> with a floating input coil would also help against hum. Just what
> other gear was used may affect noise though, TT, preamp perhhaps.
>
> I better get on my bike and ride out of range of the old rubbish,
> boots, old bottles being hurled my way because I humbly suggested the
> merits of major parts upgrading to ancient old junk.
>
> Patrick Turner

No boots or old bottles from me... As we all know, the range of radio
"restoration" goes from "everything authentic to period" (re-stuff old
caps, keep dog-bone resistors, use DCC wire, etc.) to "rebuild the
whole thing" with modern components. I'm somewhere in the middle with
a tendency to keep as much as possible if it works, but I always
replace all electrolytics and wax caps (time bombs!), but I don't re-
stuff any boxes and I don't leave the old dead ones littering the
chassis. I replace tube detector diodes with 1N34A point contact
diodes and I add audio NFB to nearly all the radios I restore (taking
out the useless tone controls - why top cut already crappy AM?) If
the original wiring is stable, i.e. insulation cracked but not
"leaking" and I don't have to move it, then I leave it... else I use
salvaged plastic covered wire in bright colours to replace. So far,
I've found no dead large IFT's but if I did I'd replace with a NOS
small, late 1950's types (but I'd likely put these inside the old
can!) Mains wire is always replaced with new plastic stuff, even 3
pin plug and wire, if to hand.
On the contrary (perhaps the main topic of this post), I invariably
rebuild tired old P-P "Hi-Fi" audio amplifiers completely because 1)
it's fun, and 2) I can do it better than the original in most
cases.
Cheers,
Roger

[Patrick Turner]

On Jul 2, 3:08*am, Engineer > wrote:
> On Jun 29, 7:34*pm, Patrick Turner > wrote:
>
>
>
>
>
> > On Jun 26, 8:05*am, John Byrns > wrote:
>
> > > In article >,
> > > *John L Stewart > wrote:
>
> > > > John s short note accompanying his amplifier invites us to comment &
> > > > offer suggestions. I did 7 or 8 years ago & found that John was already
> > > > happy with his amplifier just the way it was then. And I guess he still
> > > > is. And John comments freely on the work of others.
>
> > > I don't think I ever liked it "just the way it was", there have always been a
> > > few details that need improving. *
>
> > > First, I would like to change the input tubes to the equivalent of 12AU7s,
> > > unfortunately I don't think there is an octal equivalent for the 12AU7, so I
> > > would have to somehow adapt 9 pin miniature sockets to the existing chassis
> > > holes. *I remember that 9 pin miniature sockets used to be available that would
> > > fit the cutout for an octal socket, does anyone know where I might find a couple
> > > of those?
>
> > > Second, I would like to try connecting the output tubes as triodes. *This would
> > > require adding an LC section to the power supply, and would unfortunately cut
> > > the power output to something less than the half the power output available in
> > > pentode mode.
>
> > > > Anyway, here is my take on some minimal changes that would result in a
> > > > substantial performance improvement.
> > > snip interesting read for brevity.
>
> > Hmm, with 240V mains here I won't be rushing into building any amp
> > without a PT.
>
> > When wanting to fit a ninepin socket into a larger 28mm octal socket
> > hole I just bore a 19mm hole in a bit of scrap metal *and bolt it in
> > under the existing 28mm hole. I often have to replace ALL the 1935
> > tubes in some horrible old radio chassis made in 1935 and fitted with
> > 4V heater european tubes with queer sockets. I usually strip the whole
> > chassis, sand it dowm and paint it with a pale greeny beige color,
> > Circuit is totally revised, 1950s tubes are used, and I will cut out
> > plates of color-bond sheet metal to overlay large areas of the chassis
> > with the 50mm holes for euro-tube sockets. When finished, nobody cares
> > about the "**** that was there" and they marvel at the sound because
> > I've incorperated all the best ways to build an AM radio.
>
> > Perhaps there's a parable there...
>
> > But anyway, there's no reason why the SNR for such amps can't be good
> > if all circuitry is referred to a 0V rail. The use of an input tranny
> > with a floating input coil would also help against hum. Just what
> > other gear was used may affect noise though, TT, preamp perhhaps.
>
> > I better get on my bike and ride out of range of the old rubbish,
> > boots, old bottles being hurled my way because I humbly suggested the
> > merits of major parts upgrading to ancient old junk.
>
> > Patrick Turner
>
> No boots or old bottles from me... *As we all know, the range of radio
> "restoration" goes from "everything authentic to period" (re-stuff old
> caps, keep dog-bone resistors, use DCC wire, etc.) to "rebuild the
> whole thing" with modern components. *I'm somewhere in the middle with
> a tendency to keep as much as possible if it works, but I always
> replace all electrolytics and wax caps (time bombs!), but I don't re-
> stuff any boxes and I don't leave the old dead ones littering the
> chassis.

When guys try to restore a tubed radio and try to use NOS resistors
and put smaller elcaps into the old cans, methinks they are slaves to
mediocrity. I'd quite like to have a Model T Ford, but before I'd
drive it I'd want a few things done to make it safe, like better
brakes and that still leaves 100 other things out of 101.
But I would never try to put a big Chevvy V8 in a model T, but an
electric motor would be real fun!


*I replace tube detector diodes with 1N34A point contact
> diodes and I add audio NFB to nearly all the radios I restore (taking
> out the useless tone controls - why top cut already crappy AM?)

I put in tone controls that actually boost the treble. Usually it
takes a 12AU7 and passive tone control but without bass cut/boost,
very easy, and the 500k log pot usually there for treble cut can be
used.

> the original wiring is stable, i.e. insulation cracked but not
> "leaking" and I don't have to move it, then I leave it... else I use
> salvaged plastic covered wire in bright colours to replace. *

Old cloth covered rubber insulated multi-strand hook up wire is often
in very brittle rotten condition and the best place is the bin.
Usually I find so many mods are needed the whole darn thing needs
rewiring, so PVC insulated wire is used.

> So far,
> I've found no dead large IFT's but if I did I'd replace with a NOS
> small, late 1950's types (but I'd likely put these inside the old
> can!) *Mains wire is always replaced with new plastic stuff, even 3
> pin plug and wire, if to hand.

I have often found 235kHz IFTs which have hi Q and thus give AF BW of
oly 2kHz. Many old sets were cherished because ppl could pick up
unusual quirky stations outside they area, but now ther'es networking
and nearly all the AM radio stations have talk-back or they run pop
music from a selection of 300 MP3 files of old favourites that don't
attract royalties. There are huge amounts of adverts ( airborne spam )
so for most ppl AM radio is ****ed. Here in Canberra we have about 7
local AM stations with 4 running total crap while 2 are Govt owned
with no adds for ppl with IQ above 110. 1 is for print handicapped, so
they read the paper to you while you have a cup of tea. Or they have a
very long drawn out story about the Japanese bombing of Darwin during
WW2 when a large number of Oz and US service ppl illegally declared
martial law while they looted homes and govt stores while the air
raids were going on. **** happened in Darwin, and the tales delight
the real old timers of over 85 who remember all that - they have bad
eyesight, hence the print handicapped AM steam powered radio station.
But the signal IS GOOD. It needs to be because the old farts have lost
most of their hearing as well, like many of us will, when we graduate
to being a TOF, ( truly old fart ).

Anyway, the IFTs sometimes have green spot corrosion and broken
strands of litz wire. I sometime remove turns to each of 4 coils and
add more fixed C and raise the IF to 455kHz.

Some say IF = 2MHz would be MUCH better, but probably 3 IFTs would be
needed. Might be able to convert old TV IFTs for the job - I ain't
tried that yet.

> On the contrary (perhaps the main topic of this post), I invariably
> rebuild tired old P-P "Hi-Fi" audio amplifiers completely because 1)
> it's fun, and 2) I can do it better than the original in most
> cases.
> Cheers,
> Roger-

Well, like me you don't have to worry about an accountant clubbing you
with a baseball bat everytime you do something that would have sent
some unremarkable tin-pot wannabe old timey amp maker broke because
you wished to use 3 more resistors than the competition.

I always try to simplify, but avoid being overly simplistic.

Patrick Turner.