Explain me vactrol specs please

[Max9999]

I am trying to do some signal switching with vactrols. I found a nice datasheet of Perkin Elmer that details the differences between the models in the VTL5C series.

Below is a screenshot of a comparison from the datasheet. What exactly are the dynamic range and slope specs?
Temperature coëfficient is how stable the resistance is regarding to temperature right? Light history effect is how much the values change according to how often you turn on the light on the cel?

[Rob Strand]

What exactly are the dynamic range and slope specs?

Dynamic range is the range of resistance expressed in dB.  Essentially 20*log10(Rdark/Rlight).
Rlight will be at some high LED current, say 20mA.

Slope relates to how the resistance changes with light level between defined two points.
slope = R @ light level 2 / R @ light level 1 = R @ LED current 2 / R @ LED current 1

Typically LED currents are spread one decade apart like 0.5mA and 5mA.

Plain LDRs which aren't in an enclosed LDR+LED assembly give a spec called gamma
which represents how much the resistance changes with light level.  It contains
the same information as slope but it's not the same thing.   A gamma of 1.0 would
be the same as a slope of 10.   (The advantage of using gamma is (in theory) it doesn't
care how far the light levels are spread apart.)

Light history effect is how much the values change according to how often you turn on the light on the cel?

If an opto is left dark for a long time then current is applied the resulting resistance is different to when the
the opto is left on for a long time.   The LDRs adapt to the light level.   That's just something they do.
(Actually I think it does a similar thing for dark.)

Temperature coëfficient is how stable the resistance is regarding to temperature right?

Yes, that's a common parameter for resistors and capacitors as well.  With LDRs the tempco depends on the light level.

[PRR]

(Actually I think it does a similar thing for dark.)

Yes, flip-side of light resistance and time.

Put a clean LDR in dead-dark. Resistance keeps rising, at least for weeks, if you have a sensitive high-R ohm meter.

So the "Dark" resistance always has to be quantified "after 30 seconds" or whatever.

Ah, this chart just has small or large. Some recipes recover quick, some linger a long time.

[Max9999]

Rob Strand, about the dynamic range spec. How can VTL5C4 have 72dB dynamic range with a dark resistance of 400M, and VTLC1 100dB with 50M dark resistance? On resistance is usualle very low in all cases right?

So I would expect a larger dynamic range then with the VTLC4 .. could you explain it in another way so I can understand it this time ( maybe) Sorry.

[Rob Strand]

Rob Strand, about the dynamic range spec. How can VTL5C4 have 72dB dynamic range with a dark resistance of 400M, and VTLC1 100dB with 50M dark resistance? On resistance is usualle very low in all cases right?

So I would expect a larger dynamic range then with the VTLC4 .. could you explain it in another way so I can understand it this time ( maybe) Sorry.

As far as the general idea goes, I don't think there's a problem with what I've said.   It's only definitions.

Unfortunately there's a lot of inconsistencies with the data on these things.    About 20 years ago I entered the tabulated data and graphs and tried to turn it into something consistent.   I also dug up old and new datasheets.  Over the years there have been a few different sets of data issued.    It was a big mess, and because of that I can't remember what ended-up making sense and what was left with contradictory info.

I have a feeling there was another opto "info" document on the vactrol's web site.  That disappeared at some point but I remember a few things in that didn't match the catalog.   It might have had a few extra nuggets of info.

In some cases the dynamic range doesn't match-up with calculations based on the tabulated data.     So one source of error is the dark resistance is specified as a minimum value not a typical.  However, I don't think that's the cause of the discrepancies.   Another source of error is what time scale the dark resistance is measured at when computing dynamic range.   For the fast optos the dynamic range calculation seems close to the tabulated values (when using the Min(!) dark resistance).  For the slower optos the dynamic range is way off, almost like it's calculated on fast time scale and doesn't let the dark resistance reach it's high value.   However the fast time scale theory means the dynamic range of the fast opto should also be a little lower than the table.

Another example of the discrepancies is if you compute the slope from the tabulated ON resistances (allowing for the different current) you don't always get the slope in the tables.

[Rob Strand]

Here's another issue with the datasheets, possible typos.   Sometimes when you fit curves to manufacturers data the match is very poor and it makes you think the numbers could be missing a k  or an M should be a k.   It doesn't help to add any confidence to the results!

So look at these measurements,
https://modularsynthesis.com/vactrols/vactrols.htm

Check out the specs for the VTL5C3/2
Tabulated spec is 2 ohm at 40mA
The measurements looks more like it should be 2k ohm at 40mA

At 1mA the spec is 55k which looks OK.

So at 40mA what do we believe 2 ohm or 2k or other?

The VTL5C3/2 is made from material type 3.   There is a generic graph for material type 3.
The tabulated data is for 1mA and 40mA, which is about 40:1 change in illumination.
Look at the resistance for material 3 with 40:1 change in illumination,
Pic any two points, I'm seeing 20k down to 700 ohms on the lower curve and 200k to 7k on the upper curve,
in both cases the resistance changes by a factor of 20/0.7 = 29.

So if we take the tabulated data of 55k at 1mA as correct and estimated what it might
be as 40mA we get 55k/29 = 1.9k.
Which looks a lot like 2k, not the 2 ohms in the tabulated data for 40mA.

Smells like a typo.

With that correction, we can then estimate the On resistance at 20mA for the VTL5C3/2 to be about 3.7k.
We need the resistance at *20mA* to compute the dynamic range.
Now take the dark resistance as 10 M ohm.  Typical will be a bit higher.
Then calculate the dynamic range as 20*log10(10MEG / 3.7k) = 69dB

That's a hell of a lot closer to the 71dB in the tabulated data.
(If the *typical* dark resistance was just 10% higher than the minumum that would bump the dark resistance
to 11M and bump the dynamic range to 70dB.)

So *in this case* the dynamic range calculates wrongly because the tabulated On resistance is out by a factor of 1000!
The material curves are correct.

This is the type of junk I tried to back engineer and fix 20 years ago!!!  You have to cross check all the tabulated values for each opto and unravel the mystery/errors.

[Max9999]

That's some crazy stuff Rob Strand. I am having the same posture as your avatar right now! Is that not the neighbour kid of Beavis and Butthead btw?

What I want to do with the ldr's is something like this potentiometer switching scheme of the Soldano Slo ( see below).
He selected the VTL5C1.

As I see it how it works is that when the ldr resistance is high it forms a voltag divider with the (+/-) 3M input impedance of the phase inverter and drops the signal. Do I see that correct?

Wouldn't VTL5C4 or VTL5C10 then be  better choice with their 400M dark resistance?
What could have been a reason for Soldano to select the VTLC1?

[Eddododo]

It may have simply been a matter of that being the part they had in the room already for some other circuit.

It may also have been that the on resistance was more ideal, or that the current needed for the LED side was more optimal

Because if there's literally any reason, including that '1' comes before '4' and '10,' 50 meg resistance from the VTL5C1 is plenty for a practical NC  in most situations

[Rob Strand]

That's some crazy stuff Rob Strand. I am having the same posture as your avatar right now! Is that not the neighbour kid of Beavis and Butthead btw?

I'm the same.  Your post made me think back to all the stuffing about I did with these things.  If you look at the material graphs you would expect similar slope and dynamic range from LDRs made from the same material *but* the table shows something different.   We can try to fob it off by saying different parts operate on different parts of the curve but I seem to remember a deep dive doesn't come-up with that conclusion.  I can't remember all the fine details because there were so many questions! (Yes, the dude is Stuart from Beavis and Butthead, except I donated a new T-shirt.)

What I want to do with the ldr's is something like this potentiometer switching scheme of the Soldano Slo ( see below).
He selected the VTL5C1.

As I see it how it works is that when the ldr resistance is high it forms a voltag divider with the (+/-) 3M input impedance of the phase inverter and drops the signal. Do I see that correct?

Wouldn't VTL5C4 or VTL5C10 then be  better choice with their 400M dark resistance?
What could have been a reason for Soldano to select the VTLC1?

Superficially at least I'd come to the same conclusion.   It's possibly a trade-off had to be made between fast response (5C1) speed and off resistance (5C4).    I remember the Mesa Boogie amps were full of LDR's.   I'm pretty sure they used *different LDRs* in different positions in the circuit.  However, I haven't had a chance to disassemble many of those.     

I had a look over the schematics and there's no details of the LDR types.   What you might expect from Mesa equipment is if enough pro-players complained about the switching leaking or lagging they would probably have refined the design by using the most suitable opto in each position. 

Wind forward to today: I noticed spare parts places only offer a few types.  Some VTL5C1 and some VTL5C3/2 and the mesa store had VTL5C4/2.   No indications of the positions in the circuit.   Anyway it looks like people use whatever they can get and the finer engineering points have been swept away by time.

(There's a few other amps that use LDR's like Fender and Crate.   Details not always given on the schematic.)

[Rob Strand]

Back in the day I believe on company for these things was Vactec, then Perkin-Elmers bought the company and change it to Vactol - you can read google the history.

You can find a few old datasheets,
https://www.datasheetarchive.com/Vactec-datasheet.html

IIRC, Siliconix (now defunct) did equivalent optos as well.


Perhaps some of the older datasheets fill in the gaps and fix the typos.  (I'm pretty sure they helped me resolve some discrepancies in a few cases.)

[Max9999]

Hi Rob Strand, please tell me about Stewarts new shirt. I see a burning man? Alternative burning man festival logo?

I bought some VTL5C4's. Looking forward to experiment with them. If I have any questions I know who to bother .

[Rob Strand]

Hi Rob Strand, please tell me about Stewarts new shirt. I see a burning man? Alternative burning man festival logo?

I bought some VTL5C4's. Looking forward to experiment with them. If I have any questions I know who to bother .

I suspect all will work fine.
(The burning dude is a superhero, the opposite of Stewart.)

[Max9999]

Remember the first table I showed? "VTL4C4 speed of response Fast".

Well .. turn off decay till 100K is 1.5 sec in the other datasheet .. Going to order different vactrols

[Rob Strand]

Maybe you need to go over all the specs.

Don't forget the 100k point is just a number.    You need to consider what off means in your circuit.  Maybe it's 1M or 10M, which are going to be longer times that the 100k times in specs (obviously).

Tables = max
graphs = typical

VTL5C1
https://www.samodular.com/wp-content/uploads/2017/11/XVIVE-AUDIO-VTL5C1.pdf
Off 100k ohm, max time 35ms, typical 20ms

VTL5C4
https://www.qsl.net/wa1ion/vactrol/vactrol.pdf
Off 100k ohm, max time is 1.5 sec,  typical 700ms