I am looking for a schematic for a pair of triode connected pentodes, EL84, EL34, 6L6 etc. with a DHT front end, and some sort of a phase splitter (with gain if required). I haven't found any, and was wondering if anyone has tried this. I built a single ended amp with a Type 30 driving a triode connected EL84, and really like it, and am wondering if anyone has done this sort of thing with a push pull amp. Lots of indirectly heated tubes in front of push pull DHTs, but not much of the other way around.
Is what you want a DHT input tube, to drive a push pull Triode Wired Pentode tube /Beam Power tube output stage?
Are you planning on using DC power to the DHT input tube?
Triode wired Pentode/Beam Power output tubes have fairly low gain.
The output transformer has a voltage loss, primary to secondary.
DHT tubes usually have low gain.
Given the above facts, I recommend that you use an additional stage between the DHT input tube, and the output tubes: . . .
Try using a CCS to cathode coupled phase splitter. The phase splitter gain will be medium, but it should get the total gains and losses through the amplifier to a useful level.
I would be surprised if there is a tested and working 3 stage amplifier of that description: DHT, phase splitter, push pull triode wired pentode/beam power output.
Your quest will require some real design work.
Happy designing, building, and listening!
Are you planning on using DC power to the DHT input tube?
Triode wired Pentode/Beam Power output tubes have fairly low gain.
The output transformer has a voltage loss, primary to secondary.
DHT tubes usually have low gain.
Given the above facts, I recommend that you use an additional stage between the DHT input tube, and the output tubes: . . .
Try using a CCS to cathode coupled phase splitter. The phase splitter gain will be medium, but it should get the total gains and losses through the amplifier to a useful level.
I would be surprised if there is a tested and working 3 stage amplifier of that description: DHT, phase splitter, push pull triode wired pentode/beam power output.
Your quest will require some real design work.
Happy designing, building, and listening!
Last edited:
It's not often I would disagree with 6a3!
But this is the kind of amp I also want to build. I don't see a problem with e.g. a 2P29L into triode wired EL84. Since the EL84 is high gain in triode, and 2P29L has a mu of around 9, I don't see a problem with gain. And I would definitely use 2 stages. I avoid 3 stage amps like the plague - I prefer to put a step-up transformer in front like a Hammond 1140-LN-C which is what I use myself in 1:4 with a DHT front end. It sounds a lot cleaner.
What I don't know yet is what kind of phase splitter to use - long tailed pair, concertina or interstage.
In addition to EL84 in triode, I'd be looking at EL12n in triode, mu around 15. I have a few I bought cheap in Germany, where they can be bought new. I would also consider EL38 as an output, especially the fat bottle version which sounds wonderful, much better than 6L6 or EL34 in triode and also mu around 15. Top cap, though, so needs to be enclosed in some way. I've tried all these options in SE form and they sound excellent. I have a couple of PP OPTs looking for a build.
But this is the kind of amp I also want to build. I don't see a problem with e.g. a 2P29L into triode wired EL84. Since the EL84 is high gain in triode, and 2P29L has a mu of around 9, I don't see a problem with gain. And I would definitely use 2 stages. I avoid 3 stage amps like the plague - I prefer to put a step-up transformer in front like a Hammond 1140-LN-C which is what I use myself in 1:4 with a DHT front end. It sounds a lot cleaner.
What I don't know yet is what kind of phase splitter to use - long tailed pair, concertina or interstage.
In addition to EL84 in triode, I'd be looking at EL12n in triode, mu around 15. I have a few I bought cheap in Germany, where they can be bought new. I would also consider EL38 as an output, especially the fat bottle version which sounds wonderful, much better than 6L6 or EL34 in triode and also mu around 15. Top cap, though, so needs to be enclosed in some way. I've tried all these options in SE form and they sound excellent. I have a couple of PP OPTs looking for a build.
Last edited:
A fast way out the side door is the self inverting output stage. To get good balance it needs a large impedance in the PP triode cathodes.
Most people use 3-term regulators. But an NFET works well too.
Best results I got was using a Hammond Choke as the cathode drain,
Most people use 3-term regulators. But an NFET works well too.
Best results I got was using a Hammond Choke as the cathode drain,
I’m glad you like the idea. Weird it hasn’t been explored more. My thought is that the DHT spice of the first stage should flavour the entire amp. Another couple of tubes to consider, which are internally identical, are the 7868 and 7591. They have drive requirements similar to EL84, and close to the same gain in triode.
Hi John! As always, thanks for your input. I built a self inverting output stage, the Oddwatt, a whole back. This article contains the schematic.
https://diyaudioprojects.com/Tubes/EL84-Push-Pull/
I liked it and used it for a while, but eventually sold it and moved in another direction. I think the 7868 and 7591 could serve as a higher power tube in place of the EL84s I used in the Oddwatt. Like andyjevans, I too would prefer a two stage, but it’s not a deal breaker if it’s three stages.
tizman,
It is good to have a few responders to your quest.
These amplifiers, if they use Triode Wired Pentode/Beam Power tubes, can work reasonably well even without negative feedback (no global negative feedback).
Cathodes tied together to a CCS are a form of negative feedback, they really get rid of a lot of 2nd harmonic distortion.
I agree with jhstewart9,
You could use a self inverting output stage.
The choke "CCS" works well.
And using a transformer with a high impedance primary is a good idea, for low distortion, and a reasonable damping factor.
I know, I have done that.
Usually, the gain is on the low side.
I generally also agree with andyjevans,
He does not like 3 stage amplifiers.
Even though I proposed 3 stages for you to try, I prefer 2 stage amplifiers myself;
But I just wanted to get you enough gain.
Given a u of 9 on a DHT (use a CCS plate load to get u); and a g1 to g2 u of 19 on an EL84 triode wired . . .
(gain will be less according to gain = u x (rp) / (rp + RL), the gain is up to 9 x 19 = 171. But it would take a middle stage, a concertina phase splitter, gain slightly less than 1.
And, adding the concertina makes 3 stages.
And for example, with a 6k plate to plate output transformer, in class A, the load on each plate is 3k, and EL84 triode mode rp is 1.7k.
Sotriode wired EL84 gain is 19 x (3k / (3k + 1.7k)) = 12. The gain now is 9 x 12 = 108.
And the transformer voltage transformation from 1/2 secondary is Root (3k / 8) = 19.4. The gain is 108 / 19.4 = 5.57; that is 15 dB.
Or, with a DHT pair used as a combination input gain/phase splitter/driver, you can not use a CCS on the cathodes and a CCS in each plate; even if you could, the gain of each DHT would not be 9, it would be more like 4.5. Again, 2 stage amplifier gain would be low.
A pair of DHT used in a Para-Phase splitter would give more gain. And drive the output tubes from that.
But I am not a good fan of the Para-Phase splitter.
All the above ideas from John and Andy, and me can be good for efficient loudspeakers, or near field listening, or for low level listening.
Be happy with low gain, low power, and the listening will be good.
The Odd Watt amplifiers are good, but you did not indicate if you wanted to use Global Negative Feedback or not (they often require it).
Your Odd Watt example in Post # 6 does not use global negative feedback; it only uses Ultra Linear negative feedback.
Why did you sell it?
What did you like about it?
What did you dis-like about it?
My latest amplifiers are all 2 stage push pull, of one form or another.
They are: single ended input driving self inverting output stage; CCS cathode phase splitter input/driver driving push pull output; and Balanced driver driving Balanced output.
I use 6L6GB, 5881, 7591S, and 6CK4 triode outputs.
Have fun deciding on your next amplifier topology.
It is good to have a few responders to your quest.
These amplifiers, if they use Triode Wired Pentode/Beam Power tubes, can work reasonably well even without negative feedback (no global negative feedback).
Cathodes tied together to a CCS are a form of negative feedback, they really get rid of a lot of 2nd harmonic distortion.
I agree with jhstewart9,
You could use a self inverting output stage.
The choke "CCS" works well.
And using a transformer with a high impedance primary is a good idea, for low distortion, and a reasonable damping factor.
I know, I have done that.
Usually, the gain is on the low side.
I generally also agree with andyjevans,
He does not like 3 stage amplifiers.
Even though I proposed 3 stages for you to try, I prefer 2 stage amplifiers myself;
But I just wanted to get you enough gain.
Given a u of 9 on a DHT (use a CCS plate load to get u); and a g1 to g2 u of 19 on an EL84 triode wired . . .
(gain will be less according to gain = u x (rp) / (rp + RL), the gain is up to 9 x 19 = 171. But it would take a middle stage, a concertina phase splitter, gain slightly less than 1.
And, adding the concertina makes 3 stages.
And for example, with a 6k plate to plate output transformer, in class A, the load on each plate is 3k, and EL84 triode mode rp is 1.7k.
Sotriode wired EL84 gain is 19 x (3k / (3k + 1.7k)) = 12. The gain now is 9 x 12 = 108.
And the transformer voltage transformation from 1/2 secondary is Root (3k / 8) = 19.4. The gain is 108 / 19.4 = 5.57; that is 15 dB.
Or, with a DHT pair used as a combination input gain/phase splitter/driver, you can not use a CCS on the cathodes and a CCS in each plate; even if you could, the gain of each DHT would not be 9, it would be more like 4.5. Again, 2 stage amplifier gain would be low.
A pair of DHT used in a Para-Phase splitter would give more gain. And drive the output tubes from that.
But I am not a good fan of the Para-Phase splitter.
All the above ideas from John and Andy, and me can be good for efficient loudspeakers, or near field listening, or for low level listening.
Be happy with low gain, low power, and the listening will be good.
The Odd Watt amplifiers are good, but you did not indicate if you wanted to use Global Negative Feedback or not (they often require it).
Your Odd Watt example in Post # 6 does not use global negative feedback; it only uses Ultra Linear negative feedback.
Why did you sell it?
What did you like about it?
What did you dis-like about it?
My latest amplifiers are all 2 stage push pull, of one form or another.
They are: single ended input driving self inverting output stage; CCS cathode phase splitter input/driver driving push pull output; and Balanced driver driving Balanced output.
I use 6L6GB, 5881, 7591S, and 6CK4 triode outputs.
Have fun deciding on your next amplifier topology.
Last edited:
Another alternative is the Hammond 124E IT. And there is some signal voltage gain.
The primary is a reasonable match to several of the low power
signal triodes such as 30, 1H4G, 26, 27, 37 & so on, Maybe even a 201A.
The plate families are all similar at first glance.
https://www.hammfg.com/part/124E
The primary is a reasonable match to several of the low power
signal triodes such as 30, 1H4G, 26, 27, 37 & so on, Maybe even a 201A.
The plate families are all similar at first glance.
https://www.hammfg.com/part/124E
You're exactly right - the DHT first stage flavours the whole amp. The better it is, the better the amp. That's the flaw with driving 300bs with 12A*7 tubes - they're just not good enough. It's only builders like us that know the big difference a DHT makes - you don't find them in commercial models.
Yes - if you're in the USA go for a 7591. In Europe go for a EL12n or EL38 fat bottle. Or a trioded EL84. The best DHT driver is the 10Y/801. Especially in filament bias. That's what I use.
6A3sUMMER: Thanks for your post. With my Oddwatt, I used Sansui output transformers that didn't have ultralinear taps, so the EL84s were triode connected. I sold it because I found that I preferred my other single ended triode connected amps and wasn't using it much. The Oddwatt had better bass response, but mids were better with the single ended amps I was using at the time, and the SETs just sounded more whole and real.
Hi John: I have not built an amp with an input or interstage transformer yet. The Hammond 124E that you suggest, or the Hammond 1140-LN-C that was suggested by andyjevans, are both fairly reasonable. I like the idea of using an 01A/
Last edited:
I have many 01A 26, 27, 30, and 46 tubes at hand, but no 10Y. The 01A are thoriated tungsten like the 10Y. Have you tried 01A as the front end in a 300B amp? With respect to a DHT front end for a push pull amp, How about a triode wired push pull EL34 power section with a front end based on a 01A? It could use either a step up transformer combined with a self inverting power stage, or a phase splitter with gain. The 01A should provide plenty of flavour to the power section. Why do you prefer the 10Y? I have never heard one.
Is accurate reproduction not the goal in audio reproduction equipment? Or is the goal something more along the lines of a guitar pedal? These two things cannot both be true.
Personally, I like butter pecan.
All good fortune,
Chris
Chris: The indirectly heated triode that is usually used in the first stage also flavours the whole amp. Perhaps the DHT flavour is that of improved linearity? I'm not sure why DHTs sound better to me, but they do . I wouldn't assume that less accurate reproduction is the reason though.
Maybe my reflexive response is to the term "flavour", which conventionally means some slight change that (hopefully) improves. The universal but unstated assumption, by folk who never, ever add that their opinion is in any way "blinded", meaning that they didn't listen without already knowing about the (often expensive and high status) part they'd just installed, is that their unblinded opinions apply universally. We are fallible creatures, except me of course, I'm above all that and infallible.
You wouldn't want your doctor to prescribe you medicine tested to such a poor standard of truth. Why should we make exceptions for audio, politics or religion?
All good fortune,
Chris
You wouldn't want your doctor to prescribe you medicine tested to such a poor standard of truth. Why should we make exceptions for audio, politics or religion?
All good fortune,
Chris
If you are thinking of a 2 stage amp, don't use a 01A - it runs at a current of max. 4mA. Use something with more current capability. 01A is a very good tube e.g. for a 3 stage amp or if you look at what Ale does with it in Bartola Valves. The 26 in filament bias would run at 8mA or so. The 10Y is better if you are looking for the best sound with no compromises. Again in filament bias.
For my 2 stage amps I use 1140-LN-C in 1:4 on the input, not as a phase splitter. I have 124B which is better still than 124E because part nickel, but the 1140 is clearly better - it's Hammond's Studio range.
My favourite triode strapped pentode is the EL34. The JJ EL34 is usually my choice. If properly tested and selected, they are excellent and anything more expensive is redundant.
A simple power amp could be: Sowter 1475 as 10K input phase splitter + Type 49 differential driver + EL34 running at 400-420V with 50mA and 10K plate-to-plate for 15W Class A and about 3V rms input.
The Type 49 is also a good choice for efficiency as its filament only requires 2V/120mA. This must be DC only. You can run it at 135-140V with 6-7mA and have plenty of headroom and very low distortion. With 400V supply that's very easy to get.....
Attachments
Last edited:
Chris: No doubt that the fallible nature of human perception affects our preferences when it comes to audio. Except, of course, in the case of you and I. That said, to quote Nelson Pass, “It’s entertainment, not dialysis”, so why not?
John: Yes. Stereo reproduction is already a parlour trick full of aural approximations and distortions. Maybe a bit more distortion, or distortion of a different type, is what is needed to make the parlour trick more realistic. I’m not sure how it works, but I know what works for me when I hear it.