Limiters with optocouplers

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Limiters with optocouplers

Post by mediatechnology »

For lower cost and greater availability you might look at the THAT4316 vs. Vactrols or FETS.

I know that you've discussed the THAT4305.
I've used the 4316 and the unipolar 5V supply is not a limitation since the device is current-in/current-out.
It's about half the price and uses about half the current of the 4305.

The Vactrol is going to be expensive and the LED current more than the 4316's entire Icc.
carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Limiters with optocouplers

Post by carlmart »

THAT never mentioned that part to me, maybe because at first I was using V/2 and feeding 12-14v single, and then I moved to +/-12v.

But the price being lower and current consumption less, would justify the trouble of adding a specific 5v supply for it.

What did you use it for? A limiter?

In the next few days I'm going to build the 4305 limiter at last, as the parts have arrived. We'll see how it behaves.

One thing I still don't know is how to set the threshold pot, as that is the only part I have for adjusting the limiter point.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Limiters with optocouplers

Post by mediatechnology »

I did a Guitar Compressor design for THAT using the 4316.
I've also used it in a product in development now in the beta test stage.
Both were 9V powered using an LM2950-5.0 sub-regulator.
To simplify prototyping they have DIP adapters with the 4316 mounted.

I'm assuming that you want a simple limiter to prevent overload during capture.
Based on that premise I think you would want a fixed high ratio with the threshold adjustable.
You might consider an over-easy soft knee.

The compressor is here: viewtopic.php?f=6&t=60
FWIW I would now use a TL072 for the "A" op amp in that design.
(That's because it's a preamp and the TL072 measures quieter.)
carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Limiters with optocouplers

Post by carlmart »

A limiter is a life saver when an overload noise, voice or whatever, happens when you are shooting a movie or TV.

I've heard many complaints of young TV audio "recordists' letting the audio clip all the time, and the limiter working too.

That's not what I want, I want a tool designed for an eventual overload not ruining the recording. After it happened things get to normal quickly, as if nothing happened.

Serious recordists some time even don't like to use a limiter at all, but they are very careful at setting levels and where they put the mike.

When I started in the recording business, in the early '70s (yes, I'm that old), the Nagra 4 and then 4.2 had been released, and soon took over the market.

AFAIK it was the first recorder to bring a limiter included. Even the previous Nagra III didn't have one.

I used Nagras many times, but I never owned one. My recorder for shooting movies was a Uher 4200. I recorded the camera pilot on channel 2 and audio on channel one.

Then transferred it to a Nagra, feeding the pilot track to the pilot input, and transferred to sprocketed mag tape re-synchronizing from the Nagra. It worked like a charm.

So I'm used to recording without a limiter and know how to do things, but most people don't.

I once had the Nagra limiter schematic (it's probably around yet), but it was very complicated. All discrete, of course.

The question on a FET limiter is how you feed the VCA and what VCA to use.
JohnP
Posts: 6
Joined: Sun Jun 22, 2014 12:19 pm

Re: Limiters with optocouplers

Post by JohnP »

[quote="JR."
I am partial to limiters that are out of the path unless limiting, and I even experimented with using VCA's that way to make their already good nominal performance even better (downside is subtractive VCAs are cut only, but below limiting as clean as a simple op amp.)

JR[/quote]

We implemented such a design (subtractive-VCA) at APB for bus limiters on some rack gear we made.
Actually discussed it with JR while we worked on it.
Path was opamp inverter stage with VCA fed to summing node. VCA was effectively turned OFF most of the time by controlling the resting CV.
As the signal increased above threshold, the VCA's gain was increased and its output was added to the signal path, but at opposite polarity.
So, the VCA was effectively cancelling the original signal only above threshold. When the signal was below threshold, the path was strictly opamp-only.
We used a THAT-4305; contains a VCA + RMS detector in a QSOP pkg. Worked well. THAT parts are very good anyway, but we had opamp performance when the limiter was below threshold.
A properly designed opto circuit can have the same (THD) performance when idle, but in this case, we used the precise matching of the VCA & RMS Det to implement a very well-controlled circuit.
Also, it was "easy" to cross-link the (matched) VCAs across the multiple output buses on the mixer (Left-Right-Mono-Center) so all output levels could track based on the bus with the most GR
JP
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Limiters with optocouplers

Post by mediatechnology »

At the risk of over-generalization FET, opto and other devices with highly non-linear control characteristics limit the designer to the feedback topology.
I suppose noise gates might be the exception.
IMHO the invention of the dB linear VCA opened the door to feed-forward dynamics processing.

I did some work with subtractive VCAs about a year ago and the distortion performance was very good.
In dynamics processors, where limited attenuation range is needed, subtractive is indeed a viable - if not better - approach.

Unfortunately this application required deep attenuation (>40 dB) for use as an inline VCA.
Deep attenuation requires very close gain tracking with the VCA having to run precisely at unity gain to provide full attenuation.
I found attenuation ranges of 20-30 dB to be practical.

If I recall correctly the gain response is not dB linear when subtractive.
Was the limiter design you guys did a feedback topology?
Or were you able to linearize it for feed-forward?
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Limiters with optocouplers

Post by JR. »

The subtractive VCA I did in my automatic mixer prototype a couple years ago was feed forward, as that is how the AM gain sharing algorithm works. FWIW I laid out both VCA and DPOT gain control blocks, and the DPOT section I fired up first worked so well I never bothered to finish the VCAs. I didn't write the code for controlling the VCAs. The DPOTs accept a linear gain control so I developed my internal gain crunching based on that. Controlling the VCAs that have a log response could be converted from linear, but suggest a different internal program organization. I populated one board with THAT VCAs but never debugged that section.
----------
Decades earlier I used a VCA subtractive to make a downward expander (in a Peavey AMR product). IIRC that was feed forward too. There may be a potential stability issue using subtractive VCAs in feedback mode if you drive the subtractive path beyond unity. Then the sum stops falling and rises again.

In the AMR downward expander I used another trick to improve performance.. SInce it was a given that the downward expander would only be active during low level signals, I ran the VCA path hotter and then padded it's output down before subtracting from the dry signals, so the noise from the VCA gets padded down a like amount. Of course that trick only worked for downward expansion (perhaps a noise gate too). You want the result to be quiet when attenuating. Noise floor is a downside to subtractive VCA since simple VCA paths reduce their own noise when turned down significant amounts.

The subtractive VCA really shines for clean dry path when off and for modest amounts of attenuation. Modern high performance VCAs are very good, but not as good as a single op amp.

JR

PS: Hi JP yes I remember you guys working on that, I bet it sounds good....Someday people will be ranting about how great some old APB mixers are, but they''ll actually be correct. :D :D
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Limiters with optocouplers

Post by JR. »

mediatechnology wrote:At the risk of over-generalization FET, opto and other devices with highly non-linear control characteristics limit the designer to the feedback topology.
Back around the same time I was messing with center-tapped opto's to linearize one half with the other, I recall messing with some dual JFETs. The concept being that the two halves of a dual should track each other. (IIRC I rejected the JFETs in favor of VCAs, even the lower performance VCAs we had available back then.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Limiters with optocouplers

Post by mediatechnology »

When you linearize FET and Vactrol control response you can do a lot more things with them.
I remember several Design Ideas using both FETs and linear opto-isolators.

So if I understand correctly the subtractive approach with an exponential VCA Ec the resulting response becomes logarithmic?
That does seem to jog my memory.
It's the inverse of the inverse log correct?

I also remember the deep cutoff null being almost impossible to maintain due to gain drift with temperature and the "over-the-top" control reversal.
Maybe the world needs a clone of the OmniPressor LOL.

I also recall an recent press release from an "impossible to overload" mic preamp that dynamically controlled gain using the 5171/5173 gain controller resistive U-pad.
Just like your DPOTs.
Post Reply