• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Why no simple 2-stage EL34 PP amps?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I was thinking (never a good idea, but it happens)...

I've seen lots of PP 6L6 amps made with the "Mullard 5-20" and Williamson topologies. Basically, three stages of gain. They always have very high sensitivity so a lot of global NFB can be applied.

But I like my 6L6's in triode, [but they do need a bit of gNFB to get away with it]. Even better, EL34's in triode.

I have built a Mullard-style PP EL34-triode amp, with a 6DJ8 as first stage and 6SN7 as the LTP.

I have built a 'dual-diff' PP 6L6-triode amp, with 12AT7 LTP to 6GU7 PP-driver.

Both turned out too sensitive open loop. As little as 300mV will drive them to clipping.

So, the questions...

1) If I can get about 40V peak out of a 12AT7, that should drive an EL34-triode with no problem, right? Why wouldn't a 2-stage amp like Eli Duttman's El Cheapo work with EL34 triodes? Why would the source followers be necessary? Is it that the 12AT7 "loads down" in some way trying to drive the EL34-triode grids? It's not simply a matter of Miller capacitance, is it?

2) Or, why not build something like Morgan Jones' Bevois Valley, but with EL34-triode finals? Why can't a 6DJ8 cathodyne drive a pair of EL34-triode in class A (only, no AB)?


--
 
Last edited:
I built my own version of "El Cheapo" and there is not enough drive for an EL34. I have been trying to come up with a 2 stage amp to drive a pair of 2A3s. Now that's a challenge. The attached schematic is a design exercise. It will require about 2VRMS to drive to full output (according to rough calculations) which is really not quite enough drive without a preamp. Further, since I have conflicting data sheets WRT operating points for a triode strapped 6688 this will have to be built in order to get the proper operating points. One point: a 6L6 or EL34 is much easier to drive than a 2A3, so once the operating points for the 6688 are worked out it might be possible to use this transformer combination for a 6L6 or EL34 since the feedback is acquired from the cathode winding on the Op transformer.
 

Attachments

  • PP 2A3 6688 NC14.jpg
    PP 2A3 6688 NC14.jpg
    272.1 KB · Views: 1,142
Thanks for the replies.

I built my own version of "El Cheapo" and there is not enough drive for an EL34.

Triode outputs have more input capacitance than other types, so need a lower impedance driver/PS.

I can see how an El Cheapo 12AT7 LTP wouldn't have the low impedance and current sinking ability to drive EL34 triodes. I guess the 12AT7 rp is simply too high.

However, there is stil the controversial subject of whether or not a cathodyne (concertina, split-load) PS has output impedance akin to a cathode follower.

Taking MJ's Bevois Valley, I see that a 6DJ8 (ECC88) common cathode stage with 2.5V bias should accept 2V peak (1.4V rms) at its input. I figure 2V pk should be amplified to about 50V pk to the next stage. Theoretically, that should be enough voltage swing to drive an EL34 with grid bias of -32V, even after the loss of gain from the cathodyne stage.

The question remains then... Does the 6DJ8 cathodyne PS have the low impedance to properly drive a pair of EL34-triodes? Taking into account the slight loss of gain, of course.

There's another thread here where the consensus was that a 6SL7 can drive a 2A3 in an SE amp. If that's the case, why can't a 6DJ8 cathodyne drive PP 2A3's? The 6SJ8 common cathode stage should give you at least 50V pk volt swing, and the 6DJ8 cathodyne with 10mA through it should give you a low impedance drive. No? What am I missing here?
 
"Looks like the connection from Common to Gnd on the secondary of the OT will need to be removed."

Good call! Thanks

I checked the specs that came with the transformer and there is no electrical connection between the cathode feedback and the secondary. So it's actually sloppy drawing on my part.
 
Last edited:
rongon said:
The question remains then... Does the 6DJ8 cathodyne PS have the low impedance to properly drive a pair of EL34-triodes? Taking into account the slight loss of gain, of course.
(Without straying too far into choppy waters) the cathodyne requires identical impedances at anode and cathode in order to guarantee balance. If this is true for the circuit, then you can regard it as a low impedance source. However, if the load contains a significant Miller capacitance then this will vary somewhat with output valve voltage gain which will be different for the two halves on larger signals so balance will degrade. On the other hand, a good OPT will enforce balance of anode voltages and hence Miller effect. So my answer is 'it all depends' - a good OPT which retains balance at the HF end and well-balanced output valves should work OK.
 
Ah, thank you, that does make things clearer.

It all depends on the quality of the OPT and the matching of the tubes to help maintain balance at higher frequencies, and even then nothing's really guaranteed.

I can see why you would not want to design an amp like this and try to sell it. Its performance would likely fall apart with aging tubes or poorly balanced output tubes.
 
Rogon'

A few considerations regarding an input stage feeding a cathodyne circuit driving a p.p. power stage.

The great attraction is the simplicity, particularly with direct coupling from input stage to the cathodyne. But there is yet again no free lunch here!

Firstly one considers the grid signal voltages required here (as said). EL34/6L6 g1 signal for full output may be of the order of 25 - 40Vp depending on class of operation. To stay with class-A, the peak-peak output signal to be supplied by the cathodyne is then say 110 - 130Vpp. Examining the distortion for ECC81 triodes, they will need to work at about >180Va, giving with an available h.t. of 400V, 100V across each load resistor. (Oops! I have tested this just a few weeks ago against distortion, but am now quoting from memory!)

Perhaps fine, because one has the NFB of half the stage gain to reduce distortion.

BUT (secondly)!

The input stage must also be able to provide this amplitude or then at least half of it, that is; say 60Vp-p. That means an anode potential of some 130+V if part of the same tube, for the distortion to stay low. No go then for direct coupling, except when one uses a triode-pentode like an ECF80 or such, and even then one will be pushing it (pentodes can reach lower anode voltages than triodes).

Thus the simple direct-coupled cathodyne kind of circuit is really idealy suited only for use with the very sensitive EL84 output tubes.

In closure, one notices the use of 4 x EL84 rather than 2 x larger output tubes in the Manley Stingray (though they are fed by an ltp driver, not cathodyne.

And, the output impedance of a cathodyne is low; twice that of its use as a full cathode follower. The output voltages will remain the same as long as both outputs see the same following impedances. The current through both loads are the same and thus react the same way to output loads. The anode is not a high impedance output; it is also governed by the cathode feedback.

Finally, I use an ECC88 in the above application, more linear than an ECC81 (or whatever USA equivalents).
 
Johan,

Yes, for cathodyne (and for common cathode stage as well) I would use ECC88 (6DJ8).

For LTP a lot of people like ECC81 (12AT7).

For the cathodyne, what if one were to AC-couple the voltage amp to the cathodyne? Also, looking at SY's Impasse Preamplifier, I see it is also possible to use a voltage divider to drop the plate voltage of the voltage amp down to what is needed at the grid of the cathodyne.

So if the voltages can be adjusted so that both the voltage amplifier and the cathodyne work close to optimally, wouldn't that make it possible to drive a pair of power triodes like EL34-triode or 2A3 from a cathodyne? (Assuming a top-shelf OPT and well-matched output tubes, of course.)
 

Attachments

  • SY_impasse_preamp.jpg
    SY_impasse_preamp.jpg
    112.3 KB · Views: 1,034
Blinking eye

I have been trying to come up with a 2 stage amp to drive a pair of 2A3s. Now that's a challenge. The attached schematic is a design exercise. It will require about 2VRMS to drive to full output (according to rough calculations) which is really not quite enough drive without a preamp.

Reminds me a little bit of the Feral Eye. What drives your use of output transformer feedback?

I presume you've read Lynn Olson's Amity pages? The challenge he ran into was the huge millar capacitance of the driver.
 
Hey nigelwright7557,

I've been eyeing those OddWatt designs, but the thing that makes me wonder is why they make so little power. I mean, why only 15 watts from a push-pull pair of EL34's in ultralinear, at 2% THD? I'm getting 11 watts from a push-pull pair of triode-wired 6L6GC's. It seems there's something about those OddWatt designs that doesn't get all the power output possible from the output tubes. It might have something to do with making the output stage the phase splitter, and with an unbypassed ccs in the output tubes' cathodes.

Maybe, in these OddWatt amps, the driver stage is the limiting factor. The driver stage clips before the output stage?

I don't know the answers to those questions, which is why I'm asking...

[Added] -- Also, the 5751, 12AX7 or 6SL7 SRPP is not likely to be able to cope with the Miller capacitance of power triodes, or at least I wouldn't think so. I know there are any number of phase splitters that can drive power pentodes or beam tetrodes well, but triodes are a different matter.
 
Last edited:
You can do a 2 stage, no NFB, EL34 amp. Poindexter has been working on his "Musical Machine", for many years. In the not too distant past, Poinz switched from 5965 LTPs to 6GK5 LTPs. That change yields sufficient drive for triode wired EL34s.

If you want to do an "El Cheapo", with its modest amount of NFB, and triode wired EL34 "finals", buffer the 'T7 anodes with DC coupled ZVN0545A source followers. Between the larger load resistors the "taller" B+ rail used with EL34s permits and buffering, open loop gain should be sufficient.

The ZVN0545A source follower "trick" will also work with the high μ/high RP twin triodes. However, I would not use NFB with that sort of set up, due to low gm. High gm in circuits with loop NFB provides resistance against HF error correction signal induced slew limiting.

"Fixed" bias is definitely an option whenever source followers are employed, as even a 10 Kohm net grid to ground resistance is acceptable.
 
A circuit came out in AudioXpress using a Ecc99 cathodyne to drive triode El34's. triode 6l6's require a lot more drive.
I have a circuit developed by AWA Australia, Using a 6j7 and 6V6 to drive 2a3's for 13 watts. I would think an EL84 for work even better.
I am on Holidays so can not provide the circuits.
Phil
 
Hey nigelwright7557,

I've been eyeing those OddWatt designs, but the thing that makes me wonder is why they make so little power. I mean, why only 15 watts from a push-pull pair of EL34's in ultralinear, at 2% THD? .

They use El84's which put out about half of an EL34.

I was looking for something without nfb to give my guitar some harmonics.
Its loud enough for home use.
 
I built this circuit based on parts I had available. Might be better to use a SS CCS on top of the driver instead of using an extra tube. Also a couple of other changes, the cathode bypass for the driver at 100uF instead of 22uF, and the coupling caps to the power section at 0.1uF instead of 0.047uF. Other than that, the 6N23P cathodyne drives the 6CA7/EL34s just fine.

Scott
 

Attachments

  • 6CA7ppschematicR5.jpg
    6CA7ppschematicR5.jpg
    203.2 KB · Views: 709
There are basicly two problems.Since the cathodyne gives no voltage amplification,everything has to come from the first stage.With a direct coupling there is little voltage left for that first stage.
By using a penthode the anodevoltage can get fairly low and you get more amplification than a triode.So I suggest you use a triode/penthode + 2x EL34(as triode).A ECF82 / 6U8 can do the job.
Mona
 

Attachments

  • FC-34TPP.JPG
    FC-34TPP.JPG
    126.4 KB · Views: 656
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.