I came across a handful of inexpensive 480/8 ohm SE output transformers. I
want to build a decent headphone amp with them.

The DC resistance of the primary winding is about 32 ohms.

What's the best choice for driving the tranny . . . Should I use a CF
configuration? The low RA tubes usually have crappy linearity.

Any good advice welcome & appreciated!

TIA,

Jon

Jon Yaeger wrote:

I came across a handful of inexpensive 480/8 ohm SE output transformers. I
want to build a decent headphone amp with them.

The DC resistance of the primary winding is about 32 ohms.

The 32 ohms DCR of the P is 6.6% of the load,
so that tranny will probably have 15% losses, since secondary resistances
are notoriously a higher % of the sec load.
But for what you want, the losses are not very important.

480 ohms is rather low for a tube such as EL84 in triode,
which would be normally very linear enough if RL = 10k.


What's the best choice for driving the tranny . . . Should I use a CF
configuration? The low RA tubes usually have crappy linearity.

The also have a very low U, 6AS7 has U = 2, so the
effect of the series voltage NFB with CF config is
that there isn't much NFB acting, and you need a high drive voltage.
However, a classA SE 6AS7 with both halves paralleled would be an excellent
current source for your trannies, but are they gapped to take about
200 mA of DC?
If not, then anode choke feed and cap coupling to the tranny will have to be
used.
Its "parafeed".

Ra of 6AS7 is about 140 ohms paralleled, and thats a 1/4 of proposed RL.
And most HPs are 32 ohms, so that makes the load even better for the output
triode.

The power could be about 8 watts max, or 60 vrms at the anode.
Even if the thd was 3% at 8 watts, 8 ohms, at 20 milliwatt headphone levels it
will be
negligible.
The 6AS7 anode gain will only be about 1.5, so you will need a driver triode
capable of about 40 vrms, and an EL84 will give you a gain of 18,
so you'd need about 2.2 vrms input for 8 watts, but only 110 mV for 20
milliwatts
or 0.4 vrms at 8 ohms ( bloody loud into headphones ).

The use of some global NFB from the OPT secs to driver cathode
may be quite feasible to control what may be too much power at low gain
settings.

But when you are not using phones, the amps would be fine for
speakers, eh.

Patrick Turner.




Any good advice welcome & appreciated!

TIA,

Jon

Jon Yaeger wrote:

I came across a handful of inexpensive 480/8 ohm SE output transformers. I
want to build a decent headphone amp with them.

The DC resistance of the primary winding is about 32 ohms.

What's the best choice for driving the tranny . . . Should I use a CF
configuration? The low RA tubes usually have crappy linearity.

Any good advice welcome & appreciated!

TIA,

Jon

Not recommended at all. Those transformers are meant for a 500 ohm sound
distribution system. As such they probably have limited frequecy response. Used
in Public Address systems. Cheers, John Stewart

I'd check out some common vertical-sweep triodes. They have plateloads in
that range.

--
Gregg t3h g33k
"Ratings are for transistors....tubes have guidelines"
http://geek.scorpiorising.ca

in article , John Stewart at
wrote on 1/19/05 12:36 PM:



Jon Yaeger wrote:

I came across a handful of inexpensive 480/8 ohm SE output transformers. I
want to build a decent headphone amp with them.

The DC resistance of the primary winding is about 32 ohms.

What's the best choice for driving the tranny . . . Should I use a CF
configuration? The low RA tubes usually have crappy linearity.

Any good advice welcome & appreciated!

TIA,

Jon

Not recommended at all. Those transformers are meant for a 500 ohm sound
distribution system. As such they probably have limited frequecy response.
Used
in Public Address systems. Cheers, John Stewart


Would the fact that they would be handling only milliwatts mitigate that
limitation?

Jon

Thanks, Patrick!



in article , Patrick Turner at
wrote on 1/19/05 11:54 AM:



Jon Yaeger wrote:

I came across a handful of inexpensive 480/8 ohm SE output transformers. I
want to build a decent headphone amp with them.

The DC resistance of the primary winding is about 32 ohms.

The 32 ohms DCR of the P is 6.6% of the load,
so that tranny will probably have 15% losses, since secondary resistances
are notoriously a higher % of the sec load.
But for what you want, the losses are not very important.

480 ohms is rather low for a tube such as EL84 in triode,
which would be normally very linear enough if RL = 10k.


What's the best choice for driving the tranny . . . Should I use a CF
configuration? The low RA tubes usually have crappy linearity.

The also have a very low U, 6AS7 has U = 2, so the
effect of the series voltage NFB with CF config is
that there isn't much NFB acting, and you need a high drive voltage.
However, a classA SE 6AS7 with both halves paralleled would be an excellent
current source for your trannies, but are they gapped to take about
200 mA of DC?
If not, then anode choke feed and cap coupling to the tranny will have to be
used.
Its "parafeed".

Ra of 6AS7 is about 140 ohms paralleled, and thats a 1/4 of proposed RL.
And most HPs are 32 ohms, so that makes the load even better for the output
triode.

The power could be about 8 watts max, or 60 vrms at the anode.
Even if the thd was 3% at 8 watts, 8 ohms, at 20 milliwatt headphone levels it
will be
negligible.
The 6AS7 anode gain will only be about 1.5, so you will need a driver triode
capable of about 40 vrms, and an EL84 will give you a gain of 18,
so you'd need about 2.2 vrms input for 8 watts, but only 110 mV for 20
milliwatts
or 0.4 vrms at 8 ohms ( bloody loud into headphones ).

The use of some global NFB from the OPT secs to driver cathode
may be quite feasible to control what may be too much power at low gain
settings.

But when you are not using phones, the amps would be fine for
speakers, eh.

Patrick Turner.




Any good advice welcome & appreciated!

TIA,

Jon

in article , John Stewart at
wrote on 1/19/05 12:36 PM:



Jon Yaeger wrote:

I came across a handful of inexpensive 480/8 ohm SE output transformers. I
want to build a decent headphone amp with them.

The DC resistance of the primary winding is about 32 ohms.

What's the best choice for driving the tranny . . . Should I use a CF
configuration? The low RA tubes usually have crappy linearity.

Any good advice welcome & appreciated!

TIA,

Jon

Not recommended at all. Those transformers are meant for a 500 ohm sound
distribution system. As such they probably have limited frequecy response.
Used
in Public Address systems. Cheers, John Stewart


I've got some PP OPTs from a Kenwood EL84 receiver. Would it be feasible to
rearrange the iron to make a SE gapped XFMR?

Jon Yaeger wrote:

in article , John Stewart at
wrote on 1/19/05 12:36 PM:



Jon Yaeger wrote:

I came across a handful of inexpensive 480/8 ohm SE output transformers. I
want to build a decent headphone amp with them.

The DC resistance of the primary winding is about 32 ohms.

What's the best choice for driving the tranny . . . Should I use a CF
configuration? The low RA tubes usually have crappy linearity.

Any good advice welcome & appreciated!

TIA,

Jon

Not recommended at all. Those transformers are meant for a 500 ohm sound
distribution system. As such they probably have limited frequecy response.
Used
in Public Address systems. Cheers, John Stewart

Would the fact that they would be handling only milliwatts mitigate that
limitation?

Jon

The 480R primary is a real problem for most tubes. There are some circuits that
can work into that but those are usually rated at several watts since it is a
power tube application. If those transformers are what I think they are their
bandpass will be something like 70-8000 Hz & that is with some luck & only when
driven by a low impedance source. And they probably have no gap since the
application is pure AC, no DC at all. What physical size are they?

If you are looking at milliwatts, better to get the transformer that fits the
application. That way you will not be disappointed with the results of your sweat.

Good Luck with your project, John Stewart

Gregg wrote:

I'd check out some common vertical-sweep triodes. They have plateloads in
that range.

For a headphone amp? Isn't that a little overkill? And what vertical sweep
tubes come to mind? JLS

Gregg t3h g33k
"Ratings are for transistors....tubes have guidelines"
http://geek.scorpiorising.ca

John Stewart wrote:


Gregg wrote:

I'd check out some common vertical-sweep triodes. They have plateloads in
that range.

For a headphone amp? Isn't that a little overkill? And what vertical sweep
tubes come to mind? JLS


The PCL805 comes to mind. I have a headphone amp with two of those
per channel as OTL amp. *BUT* it also has an op-amp p/c and lots of
NFB for DC stability. Sounds very clear and precise with Sennheiser
headphones of 600Ohms. It is a lot better than the FET-amp I had
before.

Gregg t3h g33k
"Ratings are for transistors....tubes have guidelines"
http://geek.scorpiorising.ca


Kind regards, Eike

"Eike Lantzsch, ZP6CGE" wrote:

John Stewart wrote:


Gregg wrote:

I'd check out some common vertical-sweep triodes. They have plateloads in
that range.

For a headphone amp? Isn't that a little overkill? And what vertical sweep
tubes come to mind? JLS


The PCL805 comes to mind. I have a headphone amp with two of those
per channel as OTL amp. *BUT* it also has an op-amp p/c and lots of
NFB for DC stability. Sounds very clear and precise with Sennheiser
headphones of 600Ohms. It is a lot better than the FET-amp I had
before.

Gregg t3h g33k
"Ratings are for transistors....tubes have guidelines"
http://geek.scorpiorising.ca


Kind regards, Eike

Hello Eike-

Is your headphone driver similar to that shown in the following link?

http://frank.pocnet.net/sheets/106/7/7754.pdf

Five watts at 10% distortion ain't exactly hifi.

That is one of the hookups that allow some high perveance tubes to drive loads in
the 500 ohm region. It was invented by some guys at General Radio around 1952.
There was a discussion here on RAT regarding that circuit a while ago. There were
speakers built by Philips & others for direct connexion in this circuit. No OPT.
Let me know if you would like to read a copy of the GR paper. I will post it
again at ABSE.

My objection to the use of most sweep tubes for headphone application is that in
general it is overkill. I've done quite a few tests on sweep pentodes & found
that for SE applications they have far too much distortion while hooked up as
pentodes. These tubes were developed for TV sweep applications, not audio & it
shows. And I sure don't like high voltage on top of the chassis when a plate cap
is involved. These pentodes get a lot better hooked up as triodes. But there are
some small ordinary triodes that may do just as well with the proper OPT.

How much power does one need to drive a set of headphones? And what does the
otolaryngologist (ear, nose & throat specialist) cost when the amp wrecks your
ears?

Cheers, John Stewart

On Wed, 19 Jan 2005 17:32:06 -0800, John Stewart wrote:



Gregg wrote:

I'd check out some common vertical-sweep triodes. They have plateloads
in that range.

For a headphone amp? Isn't that a little overkill? And what vertical
sweep tubes come to mind? JLS

6EM7 sound good.

--
Gregg t3h g33k
"Ratings are for transistors....tubes have guidelines"
http://geek.scorpiorising.ca

John Stewart wrote:
"Eike Lantzsch, ZP6CGE" wrote:

John Stewart wrote:


Gregg wrote:

I'd check out some common vertical-sweep triodes. They have
plateloads in
that range.

For a headphone amp? Isn't that a little overkill? And what
vertical sweep
tubes come to mind? JLS


The PCL805 comes to mind. I have a headphone amp with two of those
per channel as OTL amp. *BUT* it also has an op-amp p/c and lots of
NFB for DC stability. Sounds very clear and precise with Sennheiser
headphones of 600Ohms. It is a lot better than the FET-amp I had
before.

Gregg t3h g33k
"Ratings are for transistors....tubes have guidelines"
http://geek.scorpiorising.ca


Kind regards, Eike

Hello Eike-

Is your headphone driver similar to that shown in the following link?

http://frank.pocnet.net/sheets/106/7/7754.pdf

Five watts at 10% distortion ain't exactly hifi.

That is one of the hookups that allow some high perveance tubes to
drive loads in
the 500 ohm region. It was invented by some guys at General Radio
around 1952.
There was a discussion here on RAT regarding that circuit a while
ago. There were
speakers built by Philips & others for direct connexion in this
circuit. No OPT.
Let me know if you would like to read a copy of the GR paper. I will
post it
again at ABSE.

My objection to the use of most sweep tubes for headphone application
is that in
general it is overkill. I've done quite a few tests on sweep pentodes
& found
that for SE applications they have far too much distortion while
hooked up as
pentodes. These tubes were developed for TV sweep applications, not
audio & it
shows. And I sure don't like high voltage on top of the chassis when
a plate cap
is involved. These pentodes get a lot better hooked up as triodes.
But there are
some small ordinary triodes that may do just as well with the proper
OPT.

How much power does one need to drive a set of headphones? And what
does the
otolaryngologist (ear, nose & throat specialist) cost when the amp
wrecks your
ears?

Cheers, John Stewart


I built a stereo SE headphone amp with a couple 60FX5 tubes..Power
output is about 1 watt per channel.With my cheapie (but fairly decent
sounding) Panasonic 32ohm headphones 1 watt is WAY more than you
need.Half volume would make my ears almost literaly bleed.(It actually
was painful.)
You won't need alot of power output,a half watt is probably enough for
most anyone with most any headphones,even if they *are* half deaf.

John Stewart wrote:
"Eike Lantzsch, ZP6CGE" wrote:

John Stewart wrote:


Gregg wrote:

I'd check out some common vertical-sweep triodes. They have plateloads in
that range.

For a headphone amp? Isn't that a little overkill? And what vertical sweep
tubes come to mind? JLS


The PCL805 comes to mind. I have a headphone amp with two of those
per channel as OTL amp. *BUT* it also has an op-amp p/c and lots of
NFB for DC stability. Sounds very clear and precise with Sennheiser
headphones of 600Ohms. It is a lot better than the FET-amp I had
before.

Gregg t3h g33k

Hello Eike-

Is your headphone driver similar to that shown in the following link?
http://frank.pocnet.net/sheets/106/7/7754.pdf

No - not at all.

My amp is specified as 0.001% distortion and although I'm unable to
measure this I *can* hear over 2% distortion ;-)
Maybe I get around to put the diagram on my website for which the
the time to maintain is wanting. The circuit, however, will not be a
delight for the tube-purists because, as I mentioned already, it has
two op-amps as vital parts.

That is one of the hookups that allow some high perveance tubes to drive loads in
the 500 ohm region. It was invented by some guys at General Radio around 1952.
Ah - the notorious Philips OTL design with EL86 and 800Ohms speakers
comes into mind. NO - that just does not take the cake.

[No offense meant, but can you please limit your line lenght to 70
characters? That will avoid the "comb quotes". Maybe your news-client
has an option to wordwrap automagically using CR/LF. Thank you.]

My objection to the use of most sweep tubes for headphone application is that in
general it is overkill.

The operating word here is "voltage swing" not "power".
I was sceptical about the design of this headphone amp but I tried it
and I like it very much now - except for the cabinet, which is not one
of my masterpeaces of art :-(

I've done quite a few tests on sweep pentodes & found
that for SE applications they have far too much distortion while hooked up as
pentodes. These tubes were developed for TV sweep applications, not audio & it

That is true for many sweep tubes such as EL/PL519, EL/PL504, PL85,
PCL805. An exception is EL/PL36, which is a sweet pentode which had
been hijacked for doing sweep tube work in the first generation of
european B&W TV sets.
As I said: I was sceptical because of the use of PCL805 in the amp
but ...

shows. And I sure don't like high voltage on top of the chassis when a plate cap
is involved. These pentodes get a lot better hooked up as triodes. But there are
some small ordinary triodes that may do just as well with the proper OPT.

The operating word here being "proper" ;-)
For me it is hard to come by a good opt here in SouthAm and the only
option I have is to wind my own, which works OK for big OPTs but not
for small ones. Funny is: I stumbled over some OPTs in the past but
never managed to find two of the same kind. So the idea of an OTL
headphone amp was intriguing.

How much power does one need to drive a set of headphones? And what does the
otolaryngologist (ear, nose & throat specialist) cost when the amp wrecks your
ears?

mW - and there is a log volume pot on the amp which I use to have
half open for listening to classical music and slightly lower volume
for jazz. Having my ears tested lately the otólogo was surprised when
he discovered that I still can hear over 15kHz - I'm 52.

Cheers, John Stewart

I hope my wording was as polite as my intention is. I'm not a native
english speaker.
Kind regards, Eike

I'd try a simple symmetric SRPP (using say 6FQ7s) or a mu-follower (say
6BM8), using these transformers to give to the tube an apparent output
impedance high enough. BTW, many headphone (Beyer or Sennheiser) are
250-300 Ohms, therefore the transformer will show a MUCH higher impedance
than 480R.
Some 30-40 uF of PP capacitors, rated at 160V, should be enough to feed the
signal to the primary, stopping DC. As a result the transformer should give
a good frequency response (no DC, only few mW required, saturation=0)

Ciao

Fabio



"Jon Yaeger" ha scritto nel messaggio
...
I came across a handful of inexpensive 480/8 ohm SE output transformers. I
want to build a decent headphone amp with them.

The DC resistance of the primary winding is about 32 ohms.

What's the best choice for driving the tranny . . . Should I use a CF
configuration? The low RA tubes usually have crappy linearity.

Any good advice welcome & appreciated!

TIA,

Jon