6AV6 => 12AX7 question

[AmadeusMozart]

This question is one for the more knowledgable (at least more than me) contributors on this forum.

I have been looking at the RCA Handbook # 20 and it has a microphone / phonograph amplifier (8 watts with 6L6-GC as output tube).

The microphone and tonecontrol are using a 12AX7 and the driver stage has a 6AV6 which according to the tube manuals is the same as 1/2 12AX7.

I never like the idea of using half of a double triode for one channel and the other half for the other channel so will resort to using a full 12AX7 as driver.

Knowing that the 6L6-GC can demand quite a bit of drive when driven in positive grid territory I am thinking if I should use both halves in parallel and adjust the resistor values accordingly. Am a bit reluctant to do this (although it is done in my Ming Da MC34-B) since I have read some observations that when using parallel output tubes it (supposedly?) affects the sound quality.

Gents, (and gals if they are around) please let me know of your opinions,

Thanks, AM.

[aardvarkash10]

schematic?

[AmadeusMozart]

As requested the schematic is attached. One thing that is attracting me to this design is the feedback from the plate rather than the secondary of the transformer. It will need a good transformer but in the Radiotron Handbook this form of feedback is regarded as "very statisfactory" and also Alex Kritic's RH807 has feedback from the plate. This avoids a lot of the phase shift that happens in the transformer and this has a profound effect on the performance at both top and bottom end.

What I am proposing is to leave away the 12AX7 in total but am not too sure if I'll have enough gain in that setup. Am not too keen on going with a long tail design since then we start to run into trouble with heater-cathode voltage levels exceeding max values.

Thanks in advance for your input.

[Triode Kingdom]

The tube driving the 6L6 isn't overly critical. C13 will prevent you from ever driving the 6L6 into "positive grid territory." As soon as the signal becomes that large, the cap will charge up and the outpout stage will distort. Driving the grid positive without distortion requires either an interstage transformer or direct coupling from a low Z source such as a cathode follower. Also, there is absolutely no reason not to share twin tubes between channels. For 9-pins, I would use a 12AU7 as driver, only because it's more robust than the 'AX7 and can drive a smaller grid resistor.

[aardvarkash10]

don't know what your intended end use is, but worth noting that the set-up is for a crystal input ie higher voltage than a modern mm cartridge or mike will provide.

In fact, for use with a modern line level source it appears to have PLENTY of gain. Proviso here - I haven't done the math, I'm just looking at the magnitude of values used.

Lots of discussion on this site around the partial feedback / Schade feedback approach with the conclusion that it operates best against a high impedance driver, so a pentode is indicated rather than the triode like the 6AV6. So, a 6AU6 is a more likely candidate.

In short, unless you have a specific requirement to use the voltage amp and driver topology that is shown, there are better alternatives around that would perform better...

[SY]

First, understand the context- this circuit was meant for cheap and cheesy low end phonographs. Goals are low cost and mass production, not high quality. As Aard points out, the gain structure is set up for the cheapest and poorest quality cartridges, not for modern MMs (or even the MMs of its day).

The feedback is arranged as it is because of the use of the cheapest possible transformer. Driven by a pentode, the distortion would be too high because of the source impedance. The feedback's main purpose here is to reduce the effective plate resistance of the output tube.

The coupling between sections of a 12AX7 is several orders of magnitude lower than the crosstalk from any phono cartridge, i.e., don't worry about it. The thing about paralleled output tubes is nonsense.

[AmadeusMozart]

Thanks for the replies and pointing out that it was meant for the cheapest possible iron. We can turn that equation around and with good iron the results (omitting the tone control and microphone stage) should be quite good. ( After all the OPT is normally the weakest point and if we can get decent result using this topology with the lowest quality iron then what is waiting for us when using the good stuff?) It is interesting to read in the Radiotron HB (both 3 & 4) that this form of feedback from the plate gives very satisfactory results. And so it should be since we have less issues with phase shift. And the standard second harmonic produced by an OPT is not objectionable. (if I am wrong please correct me, after all I am a newby to hi end tube audio amps)

AM

[kevinkr]

Originally Posted by AmadeusMozart

Thanks for the replies and pointing out that it was meant for the cheapest possible iron. We can turn that equation around and with good iron the results (omitting the tone control and microphone stage) should be quite good. ( After all the OPT is normally the weakest point and if we can get decent result using this topology with the lowest quality iron then what is waiting for us when using the good stuff?) It is interesting to read in the Radiotron HB (both 3 & 4) that this form of feedback from the plate gives very satisfactory results. And so it should be since we have less issues with phase shift. And the standard second harmonic produced by an OPT is not objectionable. (if I am wrong please correct me, after all I am a newby to hi end tube audio amps)

AM

I'm not convinced, it seems to me there are much better more modern designs floating around right here. I grew up in the 1960s and well remember the sound of some of these very inexpensive phonograph amplifiers. I wouldn't bother to replicate those designs even with decent transformers, the designs were just not that good. Instead and for roughly the same money I would recommend Tubelab's simple single ended amplifier. If you want to build it point to point you can do so, this I think would give you a much more predictable end result.

[Miles Prower]

Originally Posted by AmadeusMozart

This question is one for the more knowledgable (at least more than me) contributors on this forum.

I have been looking at the RCA Handbook # 20 and it has a microphone / phonograph amplifier (8 watts with 6L6-GC as output tube).

That design is absolutely hideous. I wouldn't bother.

The microphone and tonecontrol are using a 12AX7 and the driver stage has a 6AV6 which according to the tube manuals is the same as 1/2 12AX7.

You see that all the time, and it just ain't so. Take a look at the specs: 12AX7: Pd= 1200mW. 6AV6: Pd= 500mW. That, right there, ougth to tell you that one 6AV6 != 1/2 12AX7. A 6AV6 isn't up to driving a load like a 6L6. You'll probably have slew rate troubles at the high end, and the 6AV6 will roll over and die when that high level transient comes along and forces the 6L6 grid positive. The clipping behaviour on overdirve is going to be simply gawdawful.

I never like the idea of using half of a double triode for one channel and the other half for the other channel so will resort to using a full 12AX7 as driver.

Why? Take a look at the specs again. The capacitances between sections are a fraction of a pF, not big enough to make any meaningful difference until you're approaching VHF.

Am a bit reluctant to do this (although it is done in my Ming Da MC34-B) since I have read some observations that when using parallel output tubes it (supposedly?) affects the sound quality..

There is some controversy over this. Some say it makes a difference; some say it doesn't. Paralleling VTs can help with noise reduction. It's one of those areas where you have to try and see for yourself.

[AmadeusMozart]

Originally Posted by Miles Prower

That design is absolutely hideous. I wouldn't bother.

.. some say it doesn't. Paralleling VTs can help with noise reduction.....

Miles

Thanks for chiming in. I've read anumber of your posts and I greatly appreciate your opinions.

On this note - I see you have had a close look int he past at the RH84 / RH807 and also chimed in on the "New KT66 - 6AU6 SE' thread ( NEW KT66 - 6AU6 single ended ) where you did some simulation / calculations. Those two designs (as well as the Tubelab Simple SE and J C Morrison's design with the EF86 as driver) are another couple of designs that I am closely looking at.

I like in particular the local feedback but I am not convinced that I want to use a 12AT7 (ECC81). In addition I would prefer to use a tube which is in current production (e.g. the ECC83 / 12AX7) which is why I did not want to use the 6AV6 in the first place.

Could you be so kind and vent your view on Jorgetronic's KT66 - 6AU6 SE design?

(PS I have also looked at a kit which is for sale on eBait from a Hong kong seller which is using an 12AX7 as driver where the first part is the voltage amplifier and the second part of the 12AX7 is a cathode follower. Got my reservations of having the cathode sitting at the anode potential of the first section since I am concerned about heater - cathode breakdown and longevity of the tube when used in that manner).

Regards, AM