Simple EL84 amp...

[Skorpio]

Has anybody tried this circuit:

http://diyparadise.com/web/index.php?option=com_content&task=view&id=135&Itemid=5

I have not seen this on any tubeamp yet, what are the pros and cons?

Happy new year!

 

[planet10]

One of the cons is that with the output stage as phase splitter, you limit yourself to about 1/2 the power you get with a "normal" phase splitter.

There is a discussion of this circuit elsewhere on the forum, precipitating from a magazine article reproduced on someone's site (IIRC one of the US tube or transformer sippliers)

dave

 

[Skorpio]

This gyy measures 10W output with 0.6% distorsion....I don't think it is half of normal.....it is like the Baby Huey?

Can't seem to find that discussion?

 

[mjf]

tube amp

i think it is a push-pull design that runs in class a (most of the time) .10w output should be able.
greetings.........

 

[gmilitano]

This gyy measures 10W output with 0.6% distorsion....I don't think it is half of normal.....it is like the Baby Huey?

Bruce followed the idea of that amp in his EL84 Push-Pull tube amp build. He did a number of measurements and was getting less than 0.7% at 10W.

I would like to see that discussion also.
Cheers

 

[kenpeter]

I don't see it as hobbled to 1/2 whatever normal class A
power output would have been... Certainly it needs twice
the drive signal voltage swing to achieve the same output.
No big deal, that swing is still only a small fraction of B+.

These sort self splitters are limited to class A operations.
Once the driven side cuts off, it can't pass any information
to the cathode of the opposing device to fill in the other
half of the waveform. Class AB simply ain't gonna work.

A2 is also right out. Bleeds grid current pollution onto the
shared cathode node that make it impossible to split the
music signal with accuracy.

But as long as Class A is all you care about, I don't think
this circuit any less powerful than "normal". Should have
characteristic distortion half single Ended, half PushPull.
Its never going to be in perfect cancellation of the 2nd
harmonic, which is fine by me.

Antitriode is a slightly more evolved scheme that might
be worth looking into. If you like this sort of thing....

 

[planet10]

These sort self splitters are limited to class A operations.

I haven't seen it explained quite as well (at least that sunk into my head). So due to Class A about half the power possible vrs Class AB.

Here are the other threads i was mentioning earlier.

http://www.diyaudio.com/forums/showthread.php?s=&threadid=17032
http://www.diyaudio.com/forums/showthread.php?s=&threadid=9803

You might also want to search on Audio Asylum for John Swenson's variation

dave

 

[mjf]

class ab

i think heatdisspation is often the limitation to class a outputpower;
in push-pull-designs(ab and a) there is nearly no limitation of the outputcurrent(equal outputpower?),but in single ended there is it i think .................
regards............

 

[Ian444]

Skorpio, I tried Bruce's implementation of the output section on a Chinese 6L6GC push-pull amp and it immediately sounded better, wider frequency range and more clarity. Maybe the CCS helps too. Then I drove it straight from the output of the first input valve (6N2 with both halves in parallel) instead of from one half of the LTP phase splitter, this gave more detail from the music. So now I have 2 x 6N1's just sitting there, heaters glowing but no power on the plates...

So as long as you use the CCS and don't mind class A operation only, I think its a great circuit and doesn't seem to compromise sound quality at all, in my case it was a definite improvement. Less components and better sound...

Thanks for the class A explanation kenpeter, easy to understand.

 

[kenpeter]

Easy to understand??? Never been accused of that before! Must be slipping...

 

[Ian444]

Well just keep slipping mate, works well

 

[Michael Koster]

I don't see it as hobbled to 1/2 whatever normal class A
power output would have been... Certainly it needs twice
the drive signal voltage swing to achieve the same output.
No big deal, that swing is still only a small fraction of B+.

These sort self splitters are limited to class A operations.
Once the driven side cuts off, it can't pass any information
to the cathode of the opposing device to fill in the other
half of the waveform. Class AB simply ain't gonna work.

A2 is also right out. Bleeds grid current pollution onto the
shared cathode node that make it impossible to split the
music signal with accuracy.

But as long as Class A is all you care about, I don't think
this circuit any less powerful than "normal". Should have
characteristic distortion half single Ended, half PushPull.
Its never going to be in perfect cancellation of the 2nd
harmonic, which is fine by me.

Antitriode is a slightly more evolved scheme that might
be worth looking into. If you like this sort of thing....

Hi Ken,

You'll need to brush up on your obfuscation ;-) let me show you...

I refer to this as an inverting signal current mirror because the
side that gets driven from the common cathode sees a signal
current almost exactly inverse the grid-driven side. It's really
just a LTP but as a power stage it gets more interesting.

Interesting idea about distortion being between SE and PP.

It would be informative to measure the simple EL84 to see.
The builder describes a very SE-like sound.

One mechanism at work is the inversion of the signal + f2 from
the "driven" side to the "mirror" side, and re-combination in-phase
in the OPT. This results in additive f2 distortion, along with additive
signal. The result is SE distortion signature.

Another mechanism is the "mirror" side device finite gm (or gfs...)
which allows it's own non-linearity to interact with that of the
"driven" side. This could mean less f2 and more odd harmonic.

I think the higher gm of the mirror side, the more it becomes
an exact mirror of the drive side, with less and less cancellation
of distortion. My experiments with near-infinite gm (Anti-triode)
show no cancellation at all, the mirror side perfectly following the
driven side, with signal and distortion both additive (identical to SE
distortion signature).

Also, we can't ignore the behavior when the stage is overloaded.
That will happen during music playback and also becomes part of
the amp's sound signature. When a "real" PP stage is overloaded
there is more odd harmonic distortion generated as both sides limit
equally. In the mirror self-split the driven tube can reach it's
limit at one signal extreme before the other, resulting in an
increased f2 as signal increases. The ear perceives the f2 as a
loudness cue, which is presumably more pleasing than the odd
harmonic distortion of the f2 canceling pp stage.

Also interesting to note that if there is a lot of f2, the PP OPT could
see a DC component to the signal and saturate at lower signal.
I think many PP OPTs are gapped for a few mA offset so it's not an
issue in practice.

Here's another way to think about it, using dis-similar gm devices.

Michael