Yet another method to balance a vari-mu

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

NewYorkDave

Well-known member
Joined
Jun 4, 2004
Messages
4,378
Location
New York (Hudson Valley)
(I mean vari-mu in the generic sense, not a Manley specifically).

Feed a 20kHz sine gated on and off about once per second into the input. Dial in a decent amount of compression. Adjust balance trim for minimum thump at output. That's it!

This comes from the description of an old Westrex compressor in the Audio Cyclopedia. I tried it on my DIY tube limiter and it works great.
 
Back in the '80s I did a lot of design work on sundry dynamics related audio paths and found tone burst useful. Since the bench top instruments of the day were limited and expensive I rolled my own from a simple FET shunt and few glue parts.

I designed a glorified burst gate that accepted an audio input (sine wave or program) with several adjustments. I varied "on time" or gate open duration, amount of attenuation when gate closed, rep rate for next gate open event, and mixed in a fixed amount of ungated signal. I also incorporated zero cross detection so the gate would open and close on zero crossings. This eliminated the gate action introducing clicks at open/close transitions. I could generate anything from simple tone bursts to finite X dB level steps, which are actually more representative of how dynamics processors get exercised by music. I could also use music as a source to get a better feel for masking of artifacts.

I found from working with sundry VCAs and gain cells that there are huge variations between how they act dynamically compared to static test/adjustment. One of the things you commonly want to adjust out is "control voltage feed though". This is the thump you are minimizing in that 20 kHz burst test.

One insidious discovery when working with some gain cell approaches is that there is not always a single DC voltage or current offset that is optimal for both high level and low level signal. That really sucks. Therefore it is useful to explore all kinds of dynamic scenarios when designing dynamics processors, and the ability to make a repeatable transient is extremely useful.

JR
 
In production here, we have settled on the 20 Hz sinewave test for best sinusoidal shape coinciding with minimum thump.
I'll be curious of the setting is the same for both 20 hz and 20 khz.
 
With my compressor, and the particular tubes (random, unselected) I had in it, the 20kHz tone burst was a bit more effective in finding the sweet spot--but only slightly so! The 20Hz method gets me damned close.

This is far from conclusive as far as which might be "better." I s'pose you could say the 20kHz tone burst does have the advantage that you can do it by ear instead of looking at a meter or scope while tweaking the trimmer. (You only hear the thump unless you can actually hear 20kHz, which I can't. I had to sweep the function generator down to about 16kHz before I could definitely hear the tone. Like they say, getting old's a bitch :wink:)
 
Oddly, without any test gear besides the output meter of the limiter, if I just adjust the balance for maximum 20 hz amplitude on the VU meter, that coincides with best sine shape.

Oh, I see, the 20khz provides an ultrasonic* test signal to reveal the control voltage blow-by "residual" better. Good one!


* Having just tested this, I can assure all of you that it is not, in any way, ultrasonic to cats!
 
Exactly! You hear only the CV feedthrough... Unless you happen to be a cat, dog or infant :wink:

20kHz makes sense 'cause it's high enough to be ultrasonic to most people, yet still falls within the nominal bandwidth of the compressor signal and sidechain circuitry. Going higher than 20k could give unpredictable results.

Time to get tiny shooter's muffs for the kitties.
 
There is a similar technique that was used at Abbey Road which is discussed in the 'Recording The Beatles' book. I think it used a neon as a relaxation oscilator which fed a common mode click into the circuit. Again you adjust balance for minimum audible click.

Ian
 
Back
Top