What is happening when TAB u73 switches from limit to compress?

Hi Guys,


  yet another vari-mu question! Here is u73 schemo etc.

http://www.irt.de/IRT/publikationen/braunbuch/U73_U83.PDF

  Can anyone please explain what is happening? I can see the gain goes up(10dB?) in the output under compress setting, but what exactly is going on win the sidechain? is that just a change in gain, and if so, how does this soften the knee so much? I guess it is more of a soft-knee limiter really. And what is happening with the bias switched off at switch 78? (My u74 and FB76 have no bias here, and they are both limiters.)

 Basically, I have u74 and FB76, and intend to fix 'em up. Neither are very usable as they are, and I am prepared to splice and dice as necessary. I had another u74 about 15 years ago, and it was great, if not very hummy and noisy. (I think I know what I was doing wrong now). I loved it on snares, bass, and sometimes on a lead vocal. it was very fast in attack, and rather savage! (Oliver Archut describes the u73 as the fastest tube limiter ever made, and there is not much difference really!) I just bypassed the mic amp. I would love to soften the knee up on this one if possible.

Here also is schematic of u74 for your interest.

   Thank you all so much. I really do appreciate your input. I have learned SO much about vari-mu, but now realise that they is a MOUNTAIN yet to climb!

     MOst appreciated!


   KIndest regards,


       ANdyP

And here is FB76 for your interest!

I haven't checked the schematic recently, but what I remember from my head is that a diode is inserted into the control voltage path - apart from the 10 dB boost. The transfer curve of the diode creates some soft knee effect, so compression is really 'soft knee limiting'. I believe those terms weren't yet invented when the umits were designed.
I won't have time to reread the Braunbuch description the next days, but maybe this points you in sosme direction.

Michael

Hi Michael,


    I can't see a diode being inserted. Could be wrong, but . . . Schematic link is on my first post btw.



    Kindest regards,


      ANdyP

The most important thing that happens in the side-chain is that a bias voltage is substracted to the side-chain rectified voltage when the unit is in limiter mode. This puts the OA200 diodes clearly out of conduction, so the rectification characteristics is very abrupt. It changes slightly the bias of the input tubes, which have probably a little more gain, but I don't think it really changes much to the transfer curve. In addition, some low cut is applied to the signal (RC at the bottom of the 2nd xformer secondary), which may be an attempt to compensate for the subjective treble loss introduced by limiting.
In compressor mode, the diodes are not biased, so the rectification obeys the exponential law, giving a progressive ratio.
Indeed it is a little surprising that the FB76 does not have this reverse bias. Maybe it is not as sharp as the U73 in limiter mode? I think if you want to make some progress on this subject, you'll have to draw the transfer curves, and compare them with the curves in the brochure.

Too lazy to look. Abbey probably nailed it:

> a bias voltage is substracted to the side-chain rectified voltage when the unit is in limiter mode.

With no diode-bias, any signal causes gain reduction. Say that 0V to 4V causes 0dB-20dB gain reduction, then even half-volt signal is reduced a couple dB.

Put 10V bias before the diode. Up to 10V out, nothing happens, no GR, signal is passed un-changed. Then from 10V-14V, GR happens, all 20dB of reduction is crunched into the 3 dB from 10V to 14V.

Since one way the action is 0V-4V, and the other way the action is near 10V-14, some gain adjustments may be needed so that the comp/lim switch does not cause wild change of level. May want some standing-bias change and other small tweaks; sometimes you never know why a designer did this or that.

Those numbers are made-up, wrong, but the idea is common to all limiters.

Abbey and PRR,


  Thank you very much for that insight. It all makes SO much sense now! Isn't hindsight a wonderful thing . . . . Guess that's good news for me, since I guess my units are both probably of the softer knee variety already! Hussah.


    Thanks again.



    Kindest regards,



      ANdyP