best balanced tube output and tube input schematics

[gyraf]

[quote author="analag"][quote author="gyraf"]..transformerless+tube does not make sense..

Jakob E.[/quote]

That's not an objective statement at all, it would have been better (...) [/quote]

What I mean is:

With tubes, if you want balanced inputs at relatively good s/n ratio and decent CMRR, it makes no real sense - other than academically - to skip the input transformer.

On outputs the same comes true; If you want balanced output, and decent-low output impedance - combined with decent CMRR, the output transformer is not the thing you'll want to skip - unless you're into the theoretical game of "what could be done" - or you're skipping on specs in order to save parts costs.

Jakob E.

 

[mikka]

Just to give another perspective......

Sometimes, the numbers might suck........ but the sound is lovely. Sometimes the circuit might seem crude or ghastly to an engineer ...... but the sound is lovely. Sometimes when you design by ear..... technical theory and numbers mean jack..... coz the sound is lovely.

I'm told that at Electrical Engineering School around here, one of the lecturers puts up a valve circuit, gets unbridled criticism from all students on its design, then explains to them that it is one of the best selling guitar amps of all time, the Fender Twin. His purpose? To show that there is an end user for the design, who doesn't care how technically astute it is, as long as it does the job....really well.

My .02c ..... :sam:

 

[analag]

[quote author="mikka"]Just to give another perspective......

Sometimes, the numbers might suck........ but the sound is lovely. Sometimes the circuit might seem crude or ghastly to an engineer ...... but the sound is lovely. Sometimes when you design by ear..... technical theory and numbers mean jack..... coz the sound is lovely.

I'm told that at Electrical Engineering School around here, one of the lecturers puts up a valve circuit, gets unbridled criticism from all students on its design, then explains to them that it is one of the best selling guitar amps of all time, the Fender Twin. His purpose? To show that there is an end user for the design, who doesn't care how technically astute it is, as long as it does the job....really well.

My .02c ..... :sam:[/quote]

The ear should be the final judge of the sound.

analag

 

[burdij]

Just to return briefly to the original topic of this thread . . . you might check out these pages. Note that in the production unit the 6SL7 was replaced with a 12AX7:

http://www.eetimes.com/anniversary/designclassics/opamp.html

http://www.philbrickarchive.org/operational_amplifier.htm

http://www.analog.com/library/analogDialogue/archives/39-05/Web_ChH_final.pdf - - starting on page 22

 

[analag]

This is a line level balanced differential with a gain of about 5. Output resistance is not as low as a transformer coupled, but then again I have no idea what the intended purpose is. So while it will work, let's use it as a discussion piece.

analag

 

[analag]

Seems the curiousity has satisfied...mission accomplished

analag

 

[rafafredd]

what´s the actual output impedance with that 6H30 cathode follower?

 

[mikka]

In light of my previous comment..... How does it sound?


But seriously ...... I've got to do some more reading before I can comment on this ..... curiousity not satisfied...... must learn more.....

 

[analag]

[quote author="rafafredd"]what´s the actual output impedance with that 6H30 cathode follower?[/quote]

No more than about 117ohms

In light of my previous comment..... How does it sound?

Worthy to be praised. Fast, punchy and not as rounded off as it normally sound with transformer strapped to it.

analag

 

[mikka]

"Better to remain silent and be thought a fool, than speak and remove all doubt......" I NEVER could live by this. So..... firstly.....

If you want attack and transients, why go to 22k on the inputs? 12AT7 has a plate resistance of about 11k at 250v.......why not go down to 1.5k or no resistor at all? Stability? Could be made relatively stable without it, surely......

Jump on me at will..... I wanna learn.

 

[Boswell]

[quote author="mikka"]why go to 22k on the inputs? 12AT7 has a plate resistance of about 11k at 250v.......why not go down to 1.5k or no resistor at all? Stability? Could be made relatively stable without it, surely......[/quote]
The 22Ks are the feed-forward resistors to the virtual earth point (the first stage grids). The feedback resistors are the 221Ks, giving a gain of something less than 10, owing to the low open-loop gain. The 1uFs in the feedback path are unnecessary, as both ends are ground-referenced.

 

[mikka]

Thanks for your patience......

Taking the feedback to the first grid seems to me (in my ignorance) to be a bit severe. Why do this instead of after the 12AT7 or the 6N1P? There is negative feedback occurring at the cathode of each valve ....... my ears like the least amount of negative feedback possible. Could someone illuminate me..........

 

[analag]

[quote author="Boswell"][quote author="mikka"]why go to 22k on the inputs? 12AT7 has a plate resistance of about 11k at 250v.......why not go down to 1.5k or no resistor at all? Stability? Could be made relatively stable without it, surely......[/quote]
The 22Ks are the feed-forward resistors to the virtual earth point (the first stage grids). The feedback resistors are the 221Ks, giving a gain of something less than 10, owing to the low open-loop gain. The 1uFs in the feedback path are unnecessary, as both ends are ground-referenced.[/quote]

Good catch....that was an error. I forgot to take them out...aarrghh.

analag

 

[analag]

[quote author="mikka"]Thanks for your patience......

Taking the feedback to the first grid seems to me (in my ignorance) to be a bit severe. Why do this instead of after the 12AT7 or the 6N1P? There is negative feedback occurring at the cathode of each valve ....... my ears like the least amount of negative feedback possible. Could someone illuminate me..........[/quote]

Can you really hear feedback? What I can hear is lack of feedback in tube circuits.
Did you know that feedback applied to the cathode sounds different from feedback applied to the grid?
Did you know that feedback applied to the cathode is not as effective as feedback applied to the grid?
Did you know that global feedback is more sonically degrading than local feedback?
When you talk about feedback...what exactly are you saying. If you take the stance against feedback you will never learn how to apply it to tube circuits.

Let the results of your experimentation guide your design decisions, you will find that there IS a lot of misinformation and misconceptions out there. In time you will get it.

analag

 

[mikka]

Thanks Boswell .... ignorance is bliss....

Let the results of your experimentation guide your design decisions, you will find that there IS a lot of misinformation and misconceptions out there. In time you will get it.

Thank you. I'm not qualified to be asking questions here..... I'd better do more reading and get the theory down first..... then I'll break the rules.

Cheers Analag ........ :grin:

 

[analag]

[quote author="mikka"]I'm not qualified to be asking questions here..... I'd better do more reading and get the theory down first..... then I'll break the rules.

Cheers Analag ........ :grin:[/quote]

No, no....bend the rules, that's the trick. My first transformerless pre was based on something like that, with more feedback to drive down the input impedance. It worked very well, but every couple years I redo my tube stuff so it eventually got redone.

analag

 

[burdij]

Feedback, used correctly, is essential to good audio. It has gotten a bad rap by some in the audio-foolery crowd because it reduces the distortion (second and even harmonics) that people perceive as "warmth".

Global high levels of feedback are definitely bad because that practice produces the evil Transient Intermodulation Distortion. There was a brilliant series of papers produced by Walt Jung and associates on SIM and TIM. They are available at http://www.waltjung.org in the Library section. I think I can deduce a kind of general rule of thumb from these and some other work by a professor at the University of Georgia, I believe, that limiting the feedback to "local", i.e. around each stage of amplification, and the level to a value that will set the stage gain at around 20db or less produces the best compromise between TIM and THD. Not all that easy to do in a mike preamp, though, because the overall chain has to produce as much as 60db of gain and so we have to bend the 20db rule a bit if we are only going to use two gain stages.

 

[mikka]

Yes Analag ... bend.. you're right. When I BREAK the rules the thing blows up! :green: Thanks for those comments and the link burdij...

I guess at the end of the day, we're trying to coax God into the room, and each application may need a different approach..... does this need max speed, warmth, natural compression...etc

Edit: Also, with valves quite tolerant of impedance mismatch.....so I've read..... we can push the boundaries to our advantage. :sam: