7247 Tube Substitutes

[adam79]

Can a 7247 tube be substituted for a 12ax7?

Thanks,
-Adam

[RS]

They have similar pinouts; cathode, grid, plate are on the same pins.
But they're not electrically identical. Both sides of the 12ax7 are
high gain. The 7247 (12DW7) has one high gain section, one low.

There's no shortage of 12ax7's, so if the circuit calls for that, just
stay with a 12ax7 unless there's a compelling reason to sub another
tube.

[adam79]

On 1/26/12 3:40 AM, RS wrote:
>
> There's no shortage of 12ax7's, so if the circuit calls for that, just
> stay with a 12ax7 unless there's a compelling reason to sub another
> tube.
>

I got my hands on a quality 7247 and I'm itching to try it out. Will it
damage the amp?

Thanks for the info,
-Adam

[RS]

Probably won't sound good at all unless you find a particular circuit
where the second half of the 12ax7 is currently used as a driver--like
a Marshall-type follower circuit or a split load (cathodyne) phase
inverter. Otherwise, the bias for the 2nd 7247 stage will be way off
and gain will be a fraction of normal. But aside from that, you could
try it.

[adam79]

On 1/26/12 8:24 PM, RS wrote:
>
> Probably won't sound good at all unless you find a particular circuit
> where the second half of the 12ax7 is currently used as a driver--like
> a Marshall-type follower circuit or a split load (cathodyne) phase
> inverter. Otherwise, the bias for the 2nd 7247 stage will be way off
> and gain will be a fraction of normal. But aside from that, you could
> try it.

Right on. Thanks for the help.

[adam79]

I got one more question about the 7247/12DW7 tubes.. I noticed that the
two triodes are supposed to test a different values (one higher than the
other). This differs from the 12AX7 types I'm familiar with, where if
both triodes match, the tube is balanced. Is there a ratio that the
triodes should measure within for a 12DW7 tube to be considered balanced
(i.e. the triodes on a 12AX7 should be within 8% of each other to be
considered balanced, more or less).

Thanks,
-Adam

[RS]

Adam, you wouldn't normally care about balance betweem the two sides
of a preamp tube, even if they're specified to be the same (like a
12ax7).

In the case of the 7247/12DW7, the two sides are specifically
UNbalanced. They're geared to cover two different uses. One side is
high gain with high plate resistance, which would be a good preamp
stage. The other is low gain with low plate resistance, which would
make a better driver stage.

I mentioned a couple scenarios where it -may- be useful to you, but
it's probably not going to do you any good. If you want to send a
link to the schematic of your amp (google for it but just post the
link), I'll take a look.

[Lord Valve]

Jayzuz!

[boardjunkie1]

I'll say. Just how did you determine that the triodes were "matched"?
Just by looking at it? That does nothing...

Build a test circuit and measure amplification factor and plate
currents and get back to me on how those halves are matched....

[adam79]

You can tell the triodes match on a 12ax7 by simply using a tube tester

[adam79]

On 1/28/12 3:59 AM, RS wrote:
>
> I mentioned a couple scenarios where it -may- be useful to you, but
> it's probably not going to do you any good. If you want to send a
> link to the schematic of your amp (google for it but just post the
> link), I'll take a look.
>

I'm looking into buying an Ampeg V4, which uses 3 12DW7/7247 tubes.

[RS]

This is how Valve tries to build his rep. Pick someone who doesn't
have any experience with amps and act as pompous as condescending as
possible. See, then you look like a friggin genius.

Of course when a conversation that requires actual electronics chops
comes around, he'll run like your grandma's cat when she turns on the
vacuum cleaner. He doesn't know what it is--he just knows it's scary.

[RS]

There ya go then. Of course the V4's tubes may not need to be
replaced, but you'll have a spare.

Your original question was about subbing for a 12ax7, which is a
different thing. Ampeg would have used the 2nd half of the tube as a
driver or low gain stage, and since that was done at design time, the
different biasing (from 12ax7) would be accounted for.

[RS]

Adam, my original point about this was that there aren't many places
where you'd need to precisely match the two sides of a 12ax7. Output
tubes can benefit from matching due to their use in symmetrical
push-pull stages, but you'll rarely find 12ax7's used like that,
especially without some feedback mechanism. *

And as mentioned, the two 7247 stages are not designed to be similar
anyway.

* Note about differential stages: A long-tailed (Schmidt) phase
splitter uses a relatively large value for the cathode resistor, so
some degree of balance equalization is inherent to the circuit itself.
Plus, the way that circuit works, the output from the two tube halves
will not be quite equal anyway. Hence the different value plate
resistors (82k and 100k) that you'll see on Fender phase splitter
circuits.

[adam79]

On 1/28/12 8:27 PM, RS wrote:
>
> Adam, my original point about this was that there aren't many places
> where you'd need to precisely match the two sides of a 12ax7. Output
> tubes can benefit from matching due to their use in symmetrical
> push-pull stages, but you'll rarely find 12ax7's used like that,
> especially without some feedback mechanism. *
>

From everything I've read, tubes perform their best when the triodes
are matched, especially when in the PI position (tubes where the triodes
are the same, unlike the 12dw7s). You're the first person I've heard say
that an unbalanced preamp tube doesn't lower the sound quality.

[jh]

Am 29.01.2012 07:14, schrieb adam79:
> On 1/28/12 8:27 PM, RS wrote:
>>
>> Adam, my original point about this was that there aren't many places
>> where you'd need to precisely match the two sides of a 12ax7. Output
>> tubes can benefit from matching due to their use in symmetrical
>> push-pull stages, but you'll rarely find 12ax7's used like that,
>> especially without some feedback mechanism. *
>>
>
> From everything I've read, tubes perform their best when the triodes
> are matched, especially when in the PI position (tubes where the triodes

> are the same, unlike the 12dw7s). *You're the first person* I've heard say

> that an unbalanced preamp tube doesn't lower the sound quality.

I'm the second person.

anything about balanced is techno babble snakeoil.

I tried a 12DW7 in the PI position of a JTM45. Now this IS imbalance -
as you already learned

guess what?

Did not make a heck of difference. Sound was not deteriorated.

mucho $$$ to be collected by sellers of "matched ECC83s"

regards

Jochen

[RS]

Interesting experiment there, Jochen. Years ago, I wrote a program to
calculate effects and imbalance of split-tail/schmidt phase inverters
as component values were changed. I used idealized matched values for
the two triode sections, with presets for 12ax7, 12at7, etc. Swapping
tubes with same component values tells the tale. 12at7 can yield
higher overall gain than 12ax7 in that spot. It drops somewhat with
12au7, but not everything is directly proportional to mu. Of course
the 12au7 test was using two identical low-gain sides, so I never
computed gains with a 12DW7 equivalent.

I should dig that program up again and modify it to allow drastically
mismatched triode sections. It never occurred to me to do that, for
obvious reasons.

[RS]

Really? Maybe it's because it's an obscure goal; it isn't discussed
much. In fact, I know knowledgable techs who deliberately MISmatch
push-pull outputs. That can be done in several ways...doesn't have to
involve the phase splitter directly. But overall result is the same.

One of the reasons that -output- tubes are matched is to avoid uneven
currents through the output transformer primary. That's a valuable
thing for both guitar and hifi amps. But aside from that, much of the
'wisdom' there is based purely on specmanship from hifi'ers who are
trying to minimize distortion, which is not always the goal with
guitar amps.

I'm not saying that people should go for mismatched drive, but guitar
amps present an interesting argument for doing that.

In any case, my advice is: Don't worry about that.

[Lord Valve]

Told ya.

Lord Valve
Expert (fuck you)

[boardjunkie1]

Any garden variety tube tester connects the tube as a diode and only
tells you cathode emission. This has fuck all to do with anything
else....

[RS]

"Told ya" what? You haven't dared to stick your neck out with any
tech posts since embarrassing yourself with your Youtube clip.

But if you really want to talk about it: What bearing do you think
transconductance and plate resistance have on functionality and gain
in a long-tailed phase splitter? You do realize that it's not linearly
related to the rated 'mu' of the triode, right? Figure that out, then
maybe you'll get a better grip on the idea of 'matching.' Ex: Which
parameters would you use to match the triode sections, and how would
you do that?

Now post your weird nonsensical link and run off like a good hack.

[Lord Valve]

Put your pop quiz up your lame ass, wanker.

http://www.youtube.com/watch?v=xvDCa1NJPz0

[RS]

On Tue, 31 Jan 2012 07:40:53 -0700, Lord Valve

Thought so.

That was no pop quiz. Here's the point, since you didn't understand:
If you run the numbers or a simulation, you'd find that 12ax7's and
12at7's can have similar gain in that circuit. Given that the output
voltage can -appear- the same for two tubes with very different
transconductance/plate res/mu, what parameters would you match and how
would you do that?

No, I don't expect an answer. It's more complicated than replacing a
fuse, so you're out of your depth. Just post your usual link and run
away again.

[Lord Valve]

[RS]

On Tue, 31 Jan 2012 13:56:00 -0700, Lord Valve

<detr...@ix.netcom.com> wrote:

>RS wrote:
>
>> On Tue, 31 Jan 2012 07:40:53 -0700, Lord Valve
>> <detr...@ix.netcom.com> wrote:
>>
>> >RS wrote:

>> >> On Sun, 29 Jan 2012 07:26:49 -0700, Lord Valve
>> >> <detr...@ix.netcom.com> wrote:

>> >> >Told ya.
>> >> >
>> >> >Lord Valve
>> >> >Expert (fuck you)

>> >> "Told ya" what? You haven't dared to stick your neck out with any
>> >> tech posts since embarrassing yourself with your Youtube clip.
>> >>
>> >> But if you really want to talk about it: What bearing do you think
>> >> transconductance and plate resistance have on functionality and gain
>> >> in a long-tailed phase splitter? You do realize that it's not linearly
>> >> related to the rated 'mu' of the triode, right? Figure that out, then
>> >> maybe you'll get a better grip on the idea of 'matching.' Ex: Which
>> >> parameters would you use to match the triode sections, and how would
>> >> you do that?
>> >>
>> >> Now post your weird nonsensical link and run off like a good hack.
>> >
>> >Put your pop quiz up your lame ass, wanker.
>> >
>> >http://www.youtube.com/watch?v=xvDCa1NJPz0

>> RS wrote:
>> That was no pop quiz. Here's the point, since you didn't understand:
>> If you run the numbers or a simulation, you'd find that 12ax7's and
>> 12at7's can have similar gain in that circuit. Given that the output
>> voltage can -appear- the same for two tubes with very different
>> transconductance/plate res/mu, what parameters would you match and how
>> would you do that?
>>
>> No, I don't expect an answer. It's more complicated than replacing a
>> fuse, so you're out of your depth. Just post your usual link and run
>> away again.
>
>Put your pop quiz up your lame ass, wanker.
>
>http://www.youtube.com/watch?v=xvDCa1NJPz0

Told ya.

[John Wood]

Its like watching The Lorfd Whisperer.

You say it, he does it.

Fascinating.

:-)

John

[Phil_S]

"adam79" wrote in message
news:EaidnVBXjYG3V73S...@posted.toastnet...

Can a 7247 tube be substituted for a 12ax7?

Thanks,
-Adam

===================================
These guys are having some fun with you Adam. You must be a glutton for
punishment. Maybe there is a new rule here at AGA about answering this sort
of question? I don't think I saw anyone give you the simple answer. Half the
tube is a 12AX7 and the other half is a 12AU7. The 12AU7 probably has only
20% of the gain of the 12AX7 (actual result depends on the circuit). It just
isn't an appropriate substitute for a 12AX7, but it won't hurt your amp any
to plug it in. Chances are it will sound like poo, but there is no
accounting for taste.

[It's that Guy again...]

On Fri, 3 Feb 2012 15:23:51 -0500, "Phil_S"
<psymond...@comcast.net> wrote:

>These guys are having some fun with you Adam. You must be a glutton for
>punishment. Maybe there is a new rule here at AGA about answering this sort
>of question? I don't think I saw anyone give you the simple answer. Half the
>tube is a 12AX7 and the other half is a 12AU7. The 12AU7 probably has only
>20% of the gain of the 12AX7 (actual result depends on the circuit). It just
>isn't an appropriate substitute for a 12AX7, but it won't hurt your amp any
>to plug it in. Chances are it will sound like poo, but there is no
>accounting for taste.

My last post is not directed at you guys..

There is a few tricks ya can do, esp in the inverter
stage of a power amp with it, but it's really not worth
the effort in the long run, because you can't switch
the 'trick' on and off, and said 'trick' can be done
with a switch and/or a pot with a 12AX/T7 better.

Interesting idea though.

JJTj

[RS]

On Fri, 3 Feb 2012 15:23:51 -0500, "Phil_S"
<psymond...@comcast.net> wrote:

>"adam79" wrote in message
>news:EaidnVBXjYG3V73S...@posted.toastnet...
>
>Can a 7247 tube be substituted for a 12ax7?
>
>Thanks,
>-Adam
>
>===================================
>These guys are having some fun with you Adam. You must be a glutton for
>punishment. Maybe there is a new rule here at AGA about answering this sort
>of question? I don't think I saw anyone give you the simple answer.

?? Here's my first reply. Do you think this was complicated?:

--------

>Can a 7247 tube be substituted for a 12ax7?
>
>Thanks,
>-Adam

They have similar pinouts; cathode, grid, plate are on the same pins.
But they're not electrically identical. Both sides of the 12ax7 are
high gain. The 7247 (12DW7) has one high gain section, one low.

There's no shortage of 12ax7's, so if the circuit calls for that, just
stay with a 12ax7 unless there's a compelling reason to sub another
tube.

--------

[WB]

On 2/3/2012 2:23 PM, Phil_S wrote:
> Chances are it will sound like
> poo, but there is no accounting for taste.

poo comes in various textures and stink.

With a mu of 80, as compared to 100, I don't know
if the audible differences *could* really be heard
depending on where the au7 part ends up at.

[Phil_S]

"RS" wrote in message news:1tioi7hrnti9fr07i...@4ax.com...

===============================
Hi RS, nice to hear from you. <shrug> OK. I yield.

--------

[Phil_S]

"WB" wrote in message news:4f2c74d0$0$26970$bbae...@news.suddenlink.net...

============
Huh? Instead of two sections at 100, it is one at 100 and one at 20. Big
difference. At 80 vs 100, probably not.
IMO, these are good for a gain stage followed by a concertina. The AU makes
a great concertina.

[RS]

Same here. Hope you've been well. I was genuinely interested in
whether that sounded obscure. I try to play to the audience.

[WB]

On 2/3/2012 8:33 PM, Phil_S wrote:
>
> ============
> Huh? Instead of two sections at 100, it is one at 100 and one at 20. Big
> difference. At 80 vs 100, probably not.
>

My humble mistake. Please accept all due apologizes.

[RS]

On Fri, 3 Feb 2012 21:33:44 -0500, "Phil_S"
<psymond...@comcast.net> wrote:
>
>"WB" wrote in message news:4f2c74d0$0$26970$bbae...@news.suddenlink.net...

> With a mu of 80, as compared to 100, I don't know
> if the audible differences *could* really be heard
> depending on where the au7 part ends up at.
>
>============

>Huh? Instead of two sections at 100, it is one at 100 and one at 20. Big
>difference. At 80 vs 100, probably not.
>IMO, these are good for a gain stage followed by a concertina. The AU makes
>a great concertina.

Correct about the mu specs for 12au7. I think I had originally
mentioned that the 7247 might be useful for concertina/split load or
for Marshall- type gain-stage-into-follower (providing that pinouts
assigned the lower gain half to the follower).

Worth adding that mu is generally important, but in some cases, like a
diff stage/Schmidt/Long tail, lower plate resistance and higher
transconductance can be more important. So in many cases, a 12at7 can
yield higher actual amplification than a 12ax7. A 12au7 is an extreme
example of course, but as the long tail's plate load resistors are
lowered in value, even the 12au7 will start to close the gap, and can
probably have up to 50% of a 12ax7's practical signal amplification,
despite having only 20% of its rated mu. Note that I'm approximating,
but those figures should be close.

[adam79]

On 1/29/12 4:53 AM, jh wrote:
>>
>> From everything I've read, tubes perform their best when the triodes
>> are matched, especially when in the PI position (tubes where the triodes
>> are the same, unlike the 12dw7s). *You're the first person* I've heard
>> say
>> that an unbalanced preamp tube doesn't lower the sound quality.
>
>
> I'm the second person.
>
> anything about balanced is techno babble snakeoil.
>
>

> mucho $$$ to be collected by sellers of "matched ECC83s"
>
> regards
>
> Jochen

So on a tester where 1250ma is "good" (in the green), a 12ax7 that tests
1400/2200 is a great result, right? Normally I would pass right over a
tube like this (due to the triode imbalance), but now it's an
opportunity for me to debunk my prior thought (with my own ears) that
12ax7 type tubes don't have to be balanced.

[adam79]

Haha. I think it's impossible for someone to insult me on the internet!

[adam79]

On 1/28/12 8:20 PM, RS wrote:
> On Sat, 28 Jan 2012 15:17:26 -0500, adam79<ada...@toast.net> wrote:
>
>> On 1/28/12 3:59 AM, RS wrote:
>>>
>>> I mentioned a couple scenarios where it -may- be useful to you, but
>>> it's probably not going to do you any good. If you want to send a
>>> link to the schematic of your amp (google for it but just post the
>>> link), I'll take a look.
>>>
>>
>> I'm looking into buying an Ampeg V4, which uses 3 12DW7/7247 tubes.
>
> There ya go then. Of course the V4's tubes may not need to be
> replaced, but you'll have a spare.
>

I just bought a Mullard 10M 7247 for $20, and it tests really strong.
Hopefully the V4 will be an easy fix; I'm really anxious to try the
tube.. I've never owned a Mullard 10M "Master Series"

[jh]

...sorry adam

no way to really test a tube with a simple emission tester.

it seems your tester only "measures" quiescent current. Means only
"works" or "not works".

I'f you'd want to match them - the transconductance - slope would b what
ypu want to measure

Jochen

[adam79]

"mA" is the micro-ohm abbreviation, right? Is "Gm" transconductance?

[adam79]

On 2/7/12 2:41 AM, adam79 wrote:
>>> So on a tester where 1250ma is "good" (in the green), a 12ax7 that tests
>>> 1400/2200 is a great result, right? Normally I would pass right over a
>>> tube like this (due to the triode imbalance), but now it's an
>>> opportunity for me to debunk my prior thought (with my own ears) that
>>> 12ax7 type tubes don't have to be balanced.
>>
>>
>> ...sorry adam
>>
>> no way to really test a tube with a simple emission tester.
>>
>> it seems your tester only "measures" quiescent current. Means only
>> "works" or "not works".
>>
>> I'f you'd want to match them - the transconductance - slope would b what
>> ypu want to measure
>>
>> Jochen
>
> "mA" is the micro-ohm abbreviation, right? Is "Gm" transconductance?

Scratch that.. transconductance is measured in micro-ohms. 1400/2200 are
the micro-ohm values on each triode.

[RS]

Transconductance is actually measured in micro-Mho's. No joke--it's
Ohms spelled backward.

For the technically curious: Transconductance relates to turning
voltage at a tube's grid into current at the plate. The name derived
from permuting Ohm's law:

Ohm's law: E = IR
Solve for R: R = E/I

Since what was needed was I/E (current-per-voltage) rather than E/I,
it would effectively be the opposite of R (resistance). Hence the
units of "Mho's".

[adam79]

On 2/7/12 4:34 AM, RS wrote:
>
> Transconductance is actually measured in micro-Mho's. No joke--it's
> Ohms spelled backward.
>
> For the technically curious: Transconductance relates to turning
> voltage at a tube's grid into current at the plate. The name derived
> from permuting Ohm's law:
>
> Ohm's law: E = IR
> Solve for R: R = E/I
>
> Since what was needed was I/E (current-per-voltage) rather than E/I,
> it would effectively be the opposite of R (resistance). Hence the
> units of "Mho's".

I before E, except after C ..and when solving for R!