Variable Mu and beginners quest!

Hi folks!
I green in tube design (but im learning slowly),
have some questions.

Have there been a variable Mu tube compressor design
where the peak limiting rectifier sensed the signal "before"
the compressor stage instead of after?

Have anyone any experience of the 6L7GT sovtek tubes
in a variable Mu application?

This might be a corny qestion but what is the crosstalk
figure in a dual triode? Between the two triodes that is?
Im curious to know!

Some tube compressors feed the peak limiting signal trough two
resistors into each grid, e.g, UA175,Altec 438A, and others feed the
signal throug the resistors and the center winding of the input tranny,
e.g, BA6A, Altec 444. What are they tryng to achive by doing either
way?

Most classics are old designs but wich one is the most modern in
respect to engineering?

And wich compressors are the most classic and regarded,
in its field of usage? Im trying to learn wich one is used in wich situation.
Reg
KKRAFS

> Have there been a variable Mu tube compressor design where the peak limiting rectifier sensed the signal "before" the compressor stage instead of after?

Yes. But it is difficult. Grid-voltage curves are non-linear, and not the right way for a limiter. Sensing the output through a high-gain side-chain is a form of feedback which linearizes the curve and corrects for tube variation. Feedforward was really not too practical before Gilbert Cells.

Though there is an Australian radio limiter that used cathode-impedance vari-gain with feedforward. Problem is not a lot of limiting range.

> what is the crosstalk figure in a dual triode?

A problem at 50MHz. At audio, crosstalk among the components hanging off the socket is more than crosstalk inside the tube. And of course, in a balanced push-pull pair the only effect of crosstalk is a slight reduction in gain.

> feed the peak limiting signal trough two resistors into each grid,.., and others feed the signal throug the resistors and the center winding of the input tranny,

The transformer usually needs a loading resistor(s) for flattest response.

If you don't have a center-tapped transformer, center-tapped loading resistors do nearly the same thing.

> wich one is the most modern in respect to engineering?

Who cares? Modern is not better. (I designed one of the more "modern" limiters around, and I don't claim any special virtue except very low cost using $1 parts to replace $100 of hard-way stages.)

>Though there is an Australian radio limiter that used cathode-impedance >vari-gain with feedforward. Problem is not a lot of limiting range.

Say, less then 15db?

> what is the crosstalk figure in a dual triode?

>A problem at 50MHz. At audio, crosstalk among the components hanging >off the socket is more than crosstalk inside the tube. And of course, in a >balanced push-pull pair the only effect of crosstalk is a slight reduction >in gain.

So, then one could infact use dual triode as a single stage input amp
in a sterro application!? or was that a dumb thing to do because of "other"issues?

> feed the peak limiting signal trough two resistors into each grid,.., and others feed the signal throug the resistors and the center winding of the input tranny,

The transformer usually needs a loading resistor(s) for flattest response.

If you don't have a center-tapped transformer, center-tapped loading resistors do nearly the same thing.

Hmm!

> wich one is the most modern in respect to engineering?

>Who cares? Modern is not better. (I designed one of the more "modern" limiters around, and I don't claim any special virtue except very low cost using $1 parts to replace $100 of hard-way stages.)[/quote]

I didnt say that modern is better, i was more of curious to
different design aproches that have been used over the years.

Otherwhise, wich one was that?

> So, then one could infact use dual triode as a single stage input amp in a sterro application!?

Why not? Nearly ALL hi-fi in the golden years did.

The problem isn't the few-mm wide plates, or the smaller grids inside the plates. It is the pair of many-mm-long 0.047uFd coupling capacitors, often mounted side by side. Say 10pFd leakage capacitance between them, against the ~50K plate circuit resistance. That gives maybe -26dB crosstalk at the top of the audio band, -40dB midband. Common B+ capacitors might leak -26dB at the bottom of the audio band. This is quite acceptable for most FM and phono recordings. Better designs spread channel-parts further away.

[quote author="PRR"]> So, then one could infact use dual triode as a single stage input amp in a sterro application!?

>Why not? Nearly ALL hi-fi in the golden years did.

>The problem isn't the few-mm wide plates, or the smaller grids inside >the plates. It is the pair of many-mm-long 0.047uFd coupling capacitors, >often mounted side by side. Say 10pFd leakage capacitance between >them, against the ~50K plate circuit resistance.

Are you talking about the feedback paths or the inputs to the grid?
I was to use SMD close to grid pin, large PCB track etc, to minimise
effect of crosstalk. but then we have the pin to pin distance thats
always the same depending of tube used ofcourse.
But 10pF isnt much even a PCB track would add more!! I would say that
the 50k is far more sever.

Tanks for the links jaakko!