Lower Distortion Alternatives to 12AT7 & 12AU7 ?

[DrDeville]

Hello Fanatics,

I'm building a cathode-feedback PPP amp, and am
looking at tweaking the design of the input and
driver stages. Because of the cathode-feedback
design, the driver stages must provide hefty
voltage gain and swing.

The original design uses a 12AT7, with one
half as an input stage direct-coupled to the
other half, which is configured as a concertina
phase splitter.

The original design's driver stage uses
a 12AU7, with each half in common-cathode
configuration, driving half of the output stage.

I'd enjoy any suggestions as to an alternative
high-gain, high-voltage splitter-driver topology,
but am wondering if there are any lower-distortion
tube choices that would enhance the original
topology.

From Morgan Jone's "Valve Amplifiers", and
from the plate curves, it seems that the 6SN7
could be an improvement over the 12AU7.

Comments?

So far I haven't found a promising high-mu
(60 or more) alternative to the 12AT7 in
the input-concertina stages, but this
could be due to my inexperience. Any suggestions?

I'd be glad to go with separate-envelope
triodes in order to get lower distortion.

Thanks for your help, and happy tweaking.

Best,

George Ferguson

 

[SY]

6SN7-oids do much better than 12AU7s. Other alternatives include 6CG7/6FQ7, 7119, and 6BK7/6BX7. You can get a surprising amount of swing out of a 6DJ8, especially in a cascode.

For higher-mu tubes, the 6SL7 family is first-rate.

If you want swing, lose the concertina (split load) and look at constant-current diff amps. Or if you want to really trick things out, consider my SYclotron input stage.

 

[DrDeville]

Hey SY,

Thanks for the reply--your information
and wit are first rate, as per.

The 6SL7 looks great--its plate curves
are very evenly spaced.

Is the current production Sovtek 6SL7 any
good? I'd like to start with current production
before I start tube rolling.

Can you think of any drawbacks in my application,
compared to the 12AT7? With such nice curves,
I wonder why folks use the 12AT7 instead, unless
its lower Ri is somehow a factor

I'm still learning here...

And thanks for the alternative topology
suggestions--can you point me to info
on your SYchotic design?

Thanks again,

George Ferguson

 

[SY]

I'm not familiar with current-production 6SL7s, so someone else will have to opine on the Sovteks. My personal fave is NOS 5691s, but I was prescient enough to buy a bunch 20 years ago. The major disadvantage of the 6SL7 compared with a 12AT7 is size and (I think) higher interelectrode capacitance.

I'll do a search for the SYclotron and edit a link into this note. Basically, it's a folded cross-coupled inverter using (I think) a pretty novel DC coupling scheme inspired by David Berning and John Curl. It gets around the interelectrode capacitance issue by driving the grids from source followers.

EDIT: Here's a link. http://www.diyaudio.com/forums/showthread.php?s=&threadid=20889

 

[Eli Duttman]

Locktal equivalents of many popular Octal types are available at ATTRACTIVE prices. The Locktal equivalent of the 6SL7 is the 7F7.

BTW, interelectrode capacitances tend to be lower in Locktals than in Octals. Loctals use a glass button base instead of the stem found in Octals.

 

[EC8010]

You're talking about a Williamson driver. DC coupling to a concertina works very well, but there are better valves than the 12AT7. As has already been mentioned, 6SL7 and 7F7 are very good as gain stages. Continuing on the Loktal theme, 7N7 works well in a concertina. There is plenty of room for discussion on these choices.

The driver can definitely be improved using 7N7 or 6SN7 in place of 12AU7. No question.

Beware that these suggestions will require double the heater current and require a different setting of the HF compensation of the finished amplifier. They will be much better, though, so it's worth your while.

 

[fdegrove]

Hi,

6BK7/6BX7.

6BL7/6BX7?

Cheers,

 

[DrDeville]

Loctal, Octal, What the Foctal?

SY wrote:

The major disadvantage of the 6SL7 compared with a 12AT7 is size

What, it's smaller?

and (I think) higher interelectrode capacitance

I'll have to check that out. Thanks for the tip.

And thanks for the SYclotron (tm) link. It looks
very clever, and I may try it out once I am
finished with transistor-detox.


And Eli,

Thanks for the info--very interesting.

Can you give me any pointers (links are fine)
on the differences between these bases,
and how to figure out equivalent tubes
with different bases?

I looked on DuncanAmps.com, and as far
as I can see their "Substitute" listings
are for tubes with the same base type
as the tube being shown.

Man, the plate curves for the 6SL7 / 7F7

( http://tdsl.duncanamps.com/pdf/vm342.pdf )

look so much more evenly spaced than the 12AT7.
It looks like I am going to have to crack down and
do more studying, so I can get a handle on the
tradeoffs in my particular design (gain, rolloff, etc.)

Man, I love this stuff.

I just read a negative review of the EH KT90 s by
Jim Shane over on the Audio Asylum. I already
bought mine. I'll just have to hope
they work better in my application (his is an
upgraded Citation).

Thanks Again and Happy Tweaking,

George Ferguson

 

[DrDeville]

EC8010's dancing fingers tapped:

You're talking about a Williamson driver.

Thanks! Is that the correct term for a driver direct-coupled to a Concertina?

Beware that these suggestions will require double the heater current

No problemo! I purposely over-specified the power transformer. I have 12.8a available at 6.3v.

and require a different setting of the HF compensation of the finished amplifier.

Hmm... you may have just stepped out of my still-shallow pool of understanding, which is just fine. Are you talking about the R and C values of the interstage coupling and grid resistors?

They will be much better, though, so it's worth your while.

Excellent!

Many thanks,

George Ferguson

 

[SY]

Compensation refers to rolloffs deliberately placed in the amp circuit to limit the open-loop bandwidth and stabilize feedback. There can also be compensation components in the feedback loop itself, most commonly a capacitor acros the feedback resistor. Look up "Bode Plots" to get a better understanding.

 

[EC8010]

Is that the correct term for a driver direct-coupled to a Concertina?

Not really (I don't think it has a name), but the Williamson was the seminal circuit that incorporated a DC-coupled concertina followed by a push-pull triode pair.

I'm afraid that HF compensation is a bit more complex if you want to follow it theoretically. Fortunately, it often falls out OK, and even if you have to determine it experimentally, it's not too painful.

 

[DrDeville]

Fair enough on Compensation being beyond the scope of our present discussion.

I'll just keep reading and, like Teddy Kennedy, drive off that bridge when I come to it.

Meanwhile, it's time for fun with loadlines!

Does the fact that the design I am mangling, er improving, has no global negative feedback make things easier?

Best,

George Ferguson

 

[EC8010]

If you don't have global feedback, you don't even need to go as far as the compensation bridge, so you don't need to drown in the river clutching desperately for a useful pole and finding zero.

 

[SY]

If you don't have global feedback, you don't even need to go as far as the compensation bridge, so you don't need to drown in the river clutching desperately for a useful pole and finding zero.

That bodes ill.

 

[EC8010]

This punning is driving me round the Nyquist. Get real.

 

[SY]

To quote the late, great midget Tattoo, "Z-plane, boss, z-plane!"

 

[DrDeville]

Now that's really sad...
I don't even understand all the puns

Congrats on the new you, SY.
Even more debonair than the smoking ape!
Who is it?

Best,

George Ferguson

 

[SY]

The last one was actually (truly!) a dope-smoking monkey. The new one is my little tribute to Frank.

 

[DrDeville]

Wow, monkeys are smarter than I thought!

Hmmm... Frank looks a lot like that
"Lifestyles of the Rich and Famous" guy. ; )

Best,

George "Poor and Obscure" Ferguson

 

[fdegrove]

Hi,

Hmmm... Frank looks a lot like that

You'd even recognise me if I'd dress up a little...
SY scares the living daylight out of me sometimes.

A pic taken by a girlfriend at LGW airport on our way back from Grenada.

Cheers,