Which LDR's for DIY vactrols?

Hi all,
There are a lot of different LDR’s available, which ones are you using in your vactrols? 5516, 5537 etc?
Also,which LEDs are preferred?
Many thanks,

Most schems I’ve seen call for the VTL5CX, but I don’t see any point in that for a DIY module. Just take any bright LED and an LDR and electrical tape them together into an enclosure and it’ll work fine. Vactrols aren’t for precise work.


There’s precision and then there’s precision. Different LDRs can differ in their resistance by orders of magnitude.


Check out the gl55 series datasheet, you should be able to pick something similar to the vactrol you are trying to recreate … or just buy a mixed lot and play around to see what sounds best like I do :smile:


Thanks guys. Going to buy 5516, 5537 LDRs and VTL5CX vactrols and see what’s what

1 Like

Note that “VTL5Cx” vactrols from AliExpress or Ebay vendors generally come from no name manufacturers and from reports may have completely different characteristics.

1 Like

Thanks again.
The VTL5CX’s are pretty expensive. Going to just DIY it I think. I wanted to ask here for what other people are using.
5526 LDR’s seem like a good all-rounder.
If it’s not obvious, I am a total novice. Trying to control some different things with LFO/CVs. I will be sharing my progress so thanks.

1 Like

I think it would be wise to have a look at the circuit you want to use your DIY vactrol in before you decide to take some random LDR. If e.g. you want to control the amplification factor of an opamp by connecting a vactrol in parallel with the feedback resistor of the amplifier, then choose an LDR that is much higher than the feedback resistor when the LDR is not lit (so it won’t have an effect) and much lower when it is lit (so it will have an effect). In that way you are sure that the vactrol will have a noticeable effect and you can even calculate the gain factor in both conditions.


I had a project once and decided to make my own vactrols using LDRs and LEDs. In order to “match” the LDRs I hooked them up to a LED, adjusted the current thru the LED and measured the resistance of the LDR.
Then I simply sorted them. It had, if I give it a thought now, been better to simply make up a series of vactrols with LEDs and LDRs and measure the complete units.

I have since gotten bunches of vactrols (VTL5 and 9) from China and I guess I have to measure them one by one and see how the characteristics look like.

And yet another thing; among audiopholes, it is popular to make an opto-electrical volume control using the NSL-32 optoisolator. That one can be easier found than vintage vactrols.


So many components and types and values and serials numbers and…
I just want to make LFOs wiggle things and make bleep bloop noises. Haha

This forum is a treasure.


The 5537 is a good LDR for use with red LEDs and put into the Resonant Lopass Gate (Buchla.)


Conversely you may be able to adjust the circuit to work with a particular vactrol resistance. For example in the Music Thing Spring Reverb either a VTL5C3 or a 32SR3 is specified, going into a summing node along with another signal via a 47k resistor if a VTL5C3 is used, or a 470R if a 32SR3. Presumably you could also use a homemade vactrol if you adjust that resistor accordingly.

(I know about this for reasons.)


I’ve always just used cheap random ones from eBay or tayda or that other Thai cheap parts shop.

I have also found that sometimes I get better results putting the led sideways to the LDR, cos otherwise it’s too bright too fast and ends up having too much of an on/off scenario. But sideways you get a bit more range usability.

Either way YMMV and you should experiment!!


FWIW, I was just looking at the build guide for the NLC Segue and found this [my emphasis]:

The vactrols should have an off resistance of at least 5MΩ, tho 10MΩ or more is better (most of them are). Faster vactrols are also preferred. Silonex NSL-32 work okay and roll-your-own vactrols using GL5549 LDRs are good too.


Don’t do vactrols like this though.

  1. There’s some more discussion of DIY vactrols in the My Build Progress topic starting here:
  1. I put some information on vactrols and LDRs up on a page on my website:

Might add more later.


Very interesting information. I wonder what LDRs are inside the VTL5C2 and VTL5C4? I can’t find any LDRs with such long t(fall) times. They’d be really useful for the Worng TMLPGX but I don’t fancy paying over £50 for the required 8!

I have no experience with those vactrols and have to wonder if the datasheet numbers might be erroneous. They’re orders of magnitude larger than the t_fall for any LDR I know of.

As discussed in the linked thread, there might be some way to hack a vactrol with essentially a lag processor to create a long fall time, but it wouldn’t be a trivial modification. It’d probably involve adding an op amp to buffer it.

Added: Do you have a TMLPGX circuit diagram?

Added more: Found this discussion:

including the following:

Here is a link to Westhost Sound (Elliot Sound Products)

About halfway down the page, Rod posted an image of the aforementioned VTL5C4, with it’s long 1.5 second turn off time.
The vactrol is cut open. When powered, Rod claims to see no light and believes after measurements that the LED is an infrared.


If you look at the datasheet for the Silonex LDRs (link below), you see they offer then with a range of peak wavelengths from:
515 nm (green) - 615 nm (red) in CdS-based LDRs.
690 nm (far-red) in CdSe-based LDRs
You can also find PbS and InSb LDRs with peak response in the mid-infrared.

I hadn’t paid much attention to peak wavelength in the datasheets I went through. I’ve now added it to the web page.

But… mostly they’re 520 to 570 nm. There are two 690 nm LDRs I can find at Digi-Key (and one of those at Mouser). One is the Advanced Photonix NSL-6910. Its diameter is 14 mm so not so good for vactrols. The other is the same company’s NSL-6112, 4.19 mm diameter; it’s in my list. Neither’s datasheet says anything about rise/fall times, interestingly. I wonder if that’s the special sauce? NSL-6112 is $2.33 each or $18 for 10, a lot more than AliExpress GL5528 but a lot less than the Xvive vactrols.

Added still more: By the way, if anyone’s looking at the Xvive datasheets for plots of resistance vs. time after turn-on/turn-off… don’t bother. All their datasheets for their various vactrols show the same plot. It might be correct for one of their vactrols but certainly is erroneous for all the others. (Unless it isn’t and all their resistance and rise/fall time specs are fictitious.) There are some old PerkinElmer datasheets around that show very different R vs. t plots.

I added this little snippet to the LPG I’m making to simulate longer fall times. The fall time is R22*C11, so 100ms or 500ms with these values.


Interesting, thanks! In cases where you don’t need a summing node on the gate this can be simplified by replacing the two inverting amps with a voltage divider going into a non inverting one.