6DJ8 as phase splitter.

[kegger]

Give some background here.
I've got a pair of mono blocks that are a 6V6 PPP setup with no phase splitter.
I'ts got 2 extra octel sockets up front and has a fairly low B+ of roughly 315volts.

I've been toying with different topologies but finding none that want to work that well
for your basic tubes used in this situation with the octel sockets and that is the 6J5
and the 6SN7 (Any others?) so I'm leaning towards replacing one of the sockets with
a 9 pin and going to the 6DJ8 type tube and a schmidt phase splitter.

I believe with the 6DJ8's MU that I should have plenty of drive plus gain for the 6V6's
and an active pre (I have 2 pre's with 24db gain each) also with the lowish B+ voltage
it's looking pretty decent for voltages/currents on the 6DJ8 I believe.

Now with the 9 pin socket also opens up a slew of dual triodes to use, if I go the 6DJ8
route a quick rewire of the heater gives quite a few other tubes to try and I have plenty
of 6DJ8's, 6922's, and the 6n23p types. The spare octel socket could use the 6SN7 if
I felt like trying it out later or add the 6J5 for a driver in front of the phase splitter, but I
doubt I'd really either need nor want the extra gain it would provide. But hey it's there.

It seems the 6DJ8 does not have the greatest following in the DIY community but I have
some gear that use them and they sound pretty good to me. Plus it would also seem a
315v supply would make them a good fit here?

With roughly 37K plate resistors, the voltages and currents don't look to bad going the
grounded cathode route even. Yes a CCS looks better but not sure I really "need" it.

Anyone have any thoughts on this or any 2C to add?
Thanks.

[ray_moth]

You didn't mention how your 6V6 OP tubes are used - triode, UL or pentode? It makes a difference to the type of driver required. PPP triodes would need the beef of a low Rp driver but pentode mode could be more easily driven.

with pentode mode, you could probably use a 6SL7 (octal, mu of 70) as a long tail pair splitter with a CCS in the tail. Elevate the voltage of the grids and cathodes (with a potential divider from B+ to ground, the tap going to the grounded grid and bypassed to ground with a suitable cap) and you could easily accommodate a CCS without needing a negative supply. 120v plate to cathode is enough for a 6SL7.

[Miles Prower]

Originally posted by ray_moth
with pentode mode, you could probably use a 6SL7 (octal, mu of 70) as a long tail pair splitter with a CCS in the tail. Elevate the voltage of the grids and cathodes (with a potential divider from B+ to ground, the tap going to the grounded grid and bypassed to ground with a suitable cap) and you could easily accommodate a CCS without needing a negative supply. 120v plate to cathode is enough for a 6SL7.

I'm not too certain of that. The 6SL7 doesn't have very much current sourcing capability, and I don't think it'll adequately drive a 6V6 (especially if paralleled) at the higher frequencies (and certainly not if the 6V6s enter overdrive). When I did a design with a 6SL7 as a differential phase splitter, this stage was connected to cathode followers to drive 807 control grids for the necessary drive current to avoid slew rate limiting at the upper end of the audio range. 6SN7 cathode followers don't have a very stiff capacitive currents at the higher frequencies to cause these problems with 6SL7s.

[Eli Duttman]

Kegger,

You need some drive, especially if the PPP 6V6s are triode wired. Both the 7 pin 6J6 and 9 pin 6n1p are "beefier" than the 6922. The mediocre linearity of the 6J6 works to your advantage into PP "finals", as the combined distortion spectrum is NICE.

Also, your high gain line stages allow you to consider the 5687 and the ECC99.

Make adapters from Octal plugs and mini sockets for trying the various types. After you decide which type will be permanent, buy the appropriate "hole shrinker'.

[amperex]

so not much to add.

I tried a lot of phase splitter tubes driving PPP 6V6 triode connected tubes. The 6922 has reasonably low plate resistance so it can drive the PPP 6V6GT triode, but does need a preamp. In other words, 1.5 volts in will not provide full output from the PPP 6V6GT triode.

Best 6922 is Siemens with A0 stamped on a tag inside the tube. Usually found in RCA label with 'Made in Germany' printed on the tube and sometimes stamped on the box in black ink.

A Telefunken 6AQ5/ECC85 has a ever so slight edge over the Siemens 6922.

The rare Raytheon 6CG7 blackplate made in USA (not Japan manu) is better than the 6922 or Tele ECC85. Other 6GC7/6FQ7s do not come close to the performance of the Raytheon. Unfortunately, stocks are about depleted everywhere. It took 2 years to find six NOS spares.

An alternate tube is simular performing 6SN7GT. The Sylvania 6SN7GT with the 'T' shape plates manufactured in the 1940s & early 1950s is a great sounding tube. Slightly smoother on the high frequencies vs the Raytheon 6CG7, it is my second choice & may be your first choice as the Raytheon 6CG7 are somewhat bright.

Another tube with an ugly pair of top connections is the single plate 2C22/7193. I have not tried this tube, but others that have simular amps as mine and tried all the tubes I auditioned state it is the best performer of all. I will try these 7193s in the future. Of course, two tubes are needed for a phase splitter.

If you go with any of the tubes, the Tung-Sol 6P5GT is a great preamp tube to drive the phase splitter. 76 are the same tube in a different socket. I tried Sylvania 6P5G (ST bottle) and the TungSol are better to the point I would not use Sylvania.

To summarize my statements, I like a neutral tube that is crystal clear & airy sounding tube without the slightest haze, edge or grain. Warm can be a bad discription here as I believe some state warm sounding as a tube that is has a hazy sound or slight smearing.

[kegger]

I'm going to start with the 6V6's in pentode mode as wired from the factory then
possably experiment later.

I all but gave up with 6SL7's with needing the higher plate resistors and not having
much voltage to feed them, HUMM...


I would like to pick up some 6n1p, 5687 and the ECC99 but I'd really like to use some
of these tubes I have laying around.

I've got plenty of 6SN7, 6SL7, 6CG7, 6DJ8/6922/6h23n/ecc88/e88cc, 12BH7, 12AT7,
12AX7, and 5965's to use though.

I just thought I could get away with the 6DJ8/6922 in this circumstance of a lower B+ and
pushing these 6V6's decently.

Am I that far off and the 6DJ8/6922 just won't work right here?

-------------------

BY the way Amperex these are the same amps as you have.