Do Vactrols actually work with ac LED drive?

[ruffrecords]

I have been experimenting with some Vactrols lately. In one test I drove the LED with an amplified version of an input signal. The signal also went through a resistor and the opto resistor to ground as it would in a compressor. I looked at the signal across the opto resistor. At low levels the opto resistor is nearly open circuit and the output signal is good with no distortion. At higher levels, once the LED starts to conduct, the opto resistance drops and so does the output level. However, I discovered the output signal now has about 7% distortion. Increasing the signal level increases the gain reduction and also the distortion which reached about 25% at 10dB gain reduction.

So I drive the LED with a dc source instead and arranged it to give the same 5dB or 10dB of attenuation. In both cases distortion was negligible. So I am wondering if there is something in the construction of the particular Vactrol (VTL5C3/2) that causes this. So I made my own with an opto resistor a green LED and a bit of heat shrink tubing and repeated the experiment ac and dc drive with almost exactly the same results.

So now I wonder if Vactrols will only work with dc drive unlike the LA2A which drives the electro-luminescent panel directly with audio?

(I have since changed the drive to a rectified version of the input signal and it works exactly as expected with very little distortion).

Cheers

Ian

 

[gyraf]

This all depends on the integration/release time on the LDR, there are slow ones available that works fine on AC exiting - e.g. in the LA2A/LA3A where the EL-panel is lit up directly by the AC signal voltage

In the case of the VTL5C3/2 if you have a real exelitas one, not a recent "clone", you have a medium'ish release time (T(on)=3mS, T(off)=50mS) - see attached - I really like precisely this vactrol for it's steepness, its high precision across production runs, and its low temperature dependence.. Note that I used to run these on rectified/DC when I used them in optical compressors - I would consider them too fast to be used on AC driving (damn, I miss optical compressors..)

/Jakob E.

 

[ruffrecords]

Unfortunately it is not a genuine Vactrol but the one made by Xvive. Even so, I doubt its response would be faster than the genuine article. What puzzles me is my tests were done with a 1KHz sine wave which has a period far shorter than any Vacrols response. A 50mS response corresponds to 20Hz. Maybe the response drops by 20dB per decade so at 2KHz it is 40dB down so it could still couple significantly to the opto. I don't know, just theorising. What I might try is feeding a dc current into the opto resistor and driving the LED with 1KHz and measure ant 1KHz that reaches the optoresistor.

Either way, I guess I will be going down the rectify first route.

Cheers

Ian

 

[volker]

I'm not aware of any circuit that drives the optocoupler without rectifying the signal other than some guitar pedals and analag's tube compressor. Do EL-panels have a threshold voltage similar to a diode?

Urei decided to rectify the signal as well when they switched to an optocoupler in the LA4A.

 

[Audio1Man]

The current to drive the Vactrol must not be coming from the direct audio signal output, as this causes the distortion.

#1 First add buffer to the signal driving the Vactrol.
#2 Use full wave rectification.
#3 Add other parts to control Attack & Release timing.
Duke.

 

[ruffrecords]

I'm not aware of any circuit that drives the optocoupler without rectifying the signal other than some guitar pedals and analag's tube compressor. Do EL-panels have a threshold voltage similar to a diode?

I found a surprising number of schematics that feed audio straight to the LED but the majority have the optoresistor in the feedback loop of an op amp so that will considerably reduce any distortion.

Urei decided to rectify the signal as well when they switched to an optocoupler in the LA4A.

That link will not work for me. Can you check it please?

Cheers

Ian

 

[volker]

That link will not work for me. Can you check it please?

The one I linked to had some hotlink protection it seems, I changed to a different one.

 

[NOON]

That rang a bell about one of the 'colour modules' that was doing exactly this as a form of distortion. I found the product page. Colourupter Optical Disrupter Colour

 

[L´Andratté]

That rang a bell about one of the 'colour modules' that was doing exactly this as a form of distortion. I found the product page. Colourupter Optical Disrupter Colour

xqp optical disruptor is the name of the 500 modul and there was also a guitar pedal by a sister company called Trip Hazard which seems to have closed down business.
They used to publish the schematics and I built the pedal version on veroboard.
I love the sound (using vtl5c1 for highest possible frequency of gain modulation).
I´m using it with bass guitar (we're talking about low frequency single ended distortion, the higher frequencies are too fast for the the Opto).

 

[Script]

I'm not aware of any circuit that drives the optocoupler without rectifying the signal other than some guitar pedals and analag's tube compressor.

Analag's tube gain / opto compressor, yes indeed. Did anyone notice 'increased' distortion there, probably only a problem when using just one optocells per channel ? Can't check myself right now (as I am traveling).

 

[jdiamantis]

The original Vactrol VTL5C4/2 had a response time that could be used with an ac drive signal in a bridge configuration. Manipulating the input resistor resistance in series with the photocell would vary the release response somewhat by moving it up/down from the LED current vs resistance vs response time curve. As it was, it could reasonably mimic the release response time of a Urei T4B with selected units. It was important with these cells to keep the audio input to the units no greater than 0.5 - 1.0vrms to keep the THD low. Low frequency distortion was not a factor with less than 10dB G/R.

 

[warpie]

So from what I understand, finding a commercially available vactrol, good for audio is not an option nowdays?

 

[ruffrecords]

So from what I understand, finding a commercially available vactrol, good for audio is not an option nowdays?

It seems the only current production, of what were originally called Vactrols, now takes place in China. However, there are alternatives such as the Advanced Photonix NSL-32 series.

Edit: however I have tried one of these with an ac drive and it distorts too.

Cheers

Ian

 

[abbey road d enfer]

I found a surprising number of schematics that feed audio straight to the LED but the majority have the optoresistor in the feedback loop of an op amp so that will considerably reduce any distortion.

No. Distortion is quite similar because it is due to the dynamic variation of the LDR. Whether it is in an attenuator or in the NFB loop, the effect will be the same.
NFB distortion reduction only happens when the NFB loop is linear. If it is distorted, there's no improvement in distortion.

 

[abbey road d enfer]

Urei decided to rectify the signal as well when they switched to an optocoupler in the LA4A.

Yes they do, but also they have noted memory effect, which tends to smoothe the light flow in the presence of sustained program.

 

[ccaudle]

Do EL-panels have a threshold voltage similar to a diode?

EL panels require a high voltage alternating current to illuminate. The EL material you can get from AdaFruit lists as specs:

• Operating voltage: 60-250VAC
• Operating Frequency: 50-5000HZ

You can buy little lights to plug into power outlets which as far as I can tell apply the 60Hz 120V power directly to the EL panel.
They are phosphor based, so have pretty long time constants. If you are in a dark room and pull out one of the little night lights you can see it glowing for a few seconds as it slowly fades away. Don't think it would be very useful in most opto-compressor applications.

 

[abbey road d enfer]

Don't think it would be very useful in most opto-compressor applications.

This effect, that I mentioned in an earlier post, is exactly what gives the desired character of the LA2 and LA3. Long sustained periods of program result in constant illumination and longer release time, hence less distortion.

 

[ccaudle]

is exactly what gives the desired character of the LA2 and LA3.

I didn't realize those compressors had such long release time. I found a spec brochure for LA-3A that specs release time as:
"Varies from 500 milliseconds to 5.0 seconds depending on the duration of the peak causing the onset of limiting"

 

[abbey road d enfer]

Many designers have tried to model this serendipitious property of phosphorous persistance in the side-chain timing block of compressors using faster and more linear gain cells. A classic is the dual timing circuit in teh "Auto" position of many famous compresors, e.g. Fairchild 660/670. dbx/THAT have a clever approach with their "Non Linear Capacitor" circuit.

 

[Egor Reshetnikov]

i read that ldr gives more distortion with more current through the ldr So using it for attenuating higher impedance source signal may be cleaner? in place like after stepping up transformer before 1st tube of mic preamp