The Thing (Dynamorph from DeadEndFX) LEDs are always lit (and not in a good way.)

xefned

Well-known member
I very carefully built "The Thing" from DeadEndFX, and the moment I plug in 9v, the LEDs light up. They're supposed to pulse with the guitar attack—adjustable with the trimpot. And they do at certain settings, but they never go fully dim.

The 3 on the left are brighter than the 3 on the right. That might be normal since they're 2 pair of LEDs in series (not parallel.)



1684709593004.png


The only fucky thing I did is sub the 2k rate pot with a 5k pot w/ 3k3 resistor on the outside lugs.
The rate pot seems to have no effect. So perhaps that resistor trick doesn't work in this configuration.




1684715700115.png

The drag part is turning the full 25 rotation trimpot, I never reach the 6.2v specified. I can adjust only from 4.38vDC to 5.37vDC

1684720309327.png

I'll replace the rate pot (currently 5k +3k3Ω resistor) with a proper 2k pot but I'm not sure it'll make any difference. As is, the rate pot does f**k-all. So I might as well try.

If anyone has any thoughts or suggestions, then thank you. Otherwise this is just a report from the field.
 
Last edited:
I've got a similar issue with mine- the LEDs stay lit no matter what. I am able to bias as low as 6.39, and no further.

I don't recall who, but there's another user here with the same situation

I haven't been too fussed with fixing it, as its basically just a fancied up Prunes and Custard/ Curds and Whey with a bunch of bells n whistles I don't particularly care about

Does yours sound close to the demos you've seen?
 
Does yours sound close to the demos you've seen?

Actually, yeah. It's noisy, and a bit of a one-trick pony. But at the right setting, it's morphing nicely and spitting out all kinds of fun, moving harmonic interplay.

I don't think I built it wrong. Perhaps it’s just a weird beast where a different vactrol and different gain of BJT throws off the stock adjustments.

On mine, the mode switch does nothing audible. The envelope effect is only present when morph is on. That might be normal. I dunno. Hopefully, swapping out the rate pot with a proper 20k will fix the rate pot problem.

It's fun. I'm glad I got to DIY it.
 
Last edited:
After more dicking with the trimpot, there's a sweet spot where the LEDs pulse in time with your picking, but it's at a lower voltage. Very weird. So perhaps play around if 6.2v doesn't work for your build.

Lots of self-oscillation? Turn the trimpot up (higher voltage) and it'll minimize. I found a narrow window where the LEDs pulse and the desert winds of oscillation calm down.

Now the threshold knob acts more like a noise gate.

The left LEDs are from Tayda. The dimmer LEDs on the right are Vishay. I could order more Vishay to balance them out, but does it even matter? Do those LEDs affect the audio? Probably not.
 
Last edited:
Also, you gotta wonder what that DNL resistor was for. Looks like some voltage divider in the negative feedback loop of one-half of the TL072 driving the clipping diodes. (I'm making shit up.)

Maybe I'll tack a pot in there just to see what it does. :) It's the only way I learn.


1684723017523.png
 
@DGWVI that was me. But yeah, same issue, but to be honest mine sounds ab like the demos. I spent a bunch of time troubleshooting and reflowing some joints which I think may have jacked it up more since I was getting some weirder voltages off one of the ICs, so I decided to leave well enough alone. For what it’s worth, I believe those are for aesthetic purposes only and since you did like I did and installed them IN the box, it doesn’t really matter. Which vactrol did you use?
 
After more dicking with the trimpot, there's a sweet spot where the LEDs pulse in time with your picking, but it's at a lower voltage. Very weird. So perhaps play around if 6.2v doesn't work for your build.

Lots of self-oscillation? Turn the trimpot up (higher voltage) and it'll minimize.

Now the threshold knob acts more like a noise gate.

The left LEDs are from Tayda. The dimmer LEDs are Vishay. I could order more Vishay to balance them out, but does it even matter? Do those LEDs affect the audio? Probably not.

Perhaps I'll pull mine out and tinker more. It gets most of the sounds I've seen in Andy's demo, so I'm guessing it's not too far off. The LEDs aren't in the audio path, they just indicate the input signal level.

If you're digging what you hear with this one, and haven't already, I suggest building a Prunes and Custard. It's smaller, and just has the bits you need for the squelchy wavefolding

Also, you gotta wonder what that DNL resistor was for. Looks like some voltage divider in the negative feedback loop of one-half of the TL072 driving the clipping diodes. (I'm making shit up.)

Maybe I'll tack a pot in there just to see what it does. :) It's the only way I learn.


View attachment 49058

With everything else going on with that particular feedback, I'm not sure that the 1M is even doing much- it runs in parallel to all this -
Screenshot_20230521-194538.png
 
Cool. Thanks for the tip on the Prunes and Custard. Definitely has the same harmonic characteristic, but in a smaller box with simpler controls.

I'll probably stick with the Dynamorph for my situation bc I envision using it on only a few songs for recording overdubs, but probably won't go on a live pedalboard. It also seems to have a little more ghost octave than what I'm hearing in Youtube demos and a more chirpy envelope effect.
 
Which vactrol did you use?

I had a leftover VTL5C3 desoldered from the Meatsphere, so I just used that from my junk box.

BTW, I had a perfectly working Meatsphere until the internet brainz convinced me it would sound better with a roll-your-own vactrol. It's envelope response got worse. And I broke one of the legs desoldering 1 of the 2 VTL5C3s. Arg!

So the best thing about building a failed Meatsphere is that I ended up with a desoldered VTL5C3 I could use in the Dynamorph. LOL.
It was an older one from Perkins Elmer, not the xvive which reportedly has slightly different light reaction times from the originals.

What Vactrol did you use?
If starting from scratch, I'd go with the NSL-32SR3 for consistency with DeadEndFX's values. I don't really trust those xvive things for reliable, repeatable results.
 
Last edited:
I have an Xvive vtl5c9 in mine, but with a Cordyceps board on its way, I may harvest that and try the NSL-32SR3. Mostly because I’m sorta cheap and don’t really want to drop $11 on a vactrol right now. Thankfully I socketed mine in The Thing specifically for experimentation purposss.
 
I have an Xvive vtl5c9 in mine, but with a Cordyceps board on its way, I may harvest that and try the NSL-32SR3. Mostly because I’m sorta cheap and don’t really want to drop $11 on a vactrol right now. Thankfully I socketed mine in The Thing specifically for experimentation purposss.
The Cordyceps really doesn't need a fancy vactrol. Both of mine use cheap LCR0203s from Amazon, and work flawlessly
 
Now I have to …

Any value in the DNL spot kills the envelope effect for me, even low values under 100Ω.

I swapped the fake 2k rate pot with an actual 2k rate pot with no improvement.

Was planning to run it at 12v, just to see if I could get up to 6.2v at the test point, but then remembered the giant axial electrolytics: they're rated for 10v.

So, I'm just going to be happy with it as is. Every pot does something except the rate pot. And the Threshold pot is boolean.

Maybe I'll wire the expression jack before I box it up just for the heck of it. [EDIT: a week or so later, I wired the expression jack with no improvement.]
 
Last edited:
A new thought:

The ONLY knobs not working are THRESHOLD and the RATE. And both knobs are related to the vactrol response. I'm guessing the Vactol or vactrol resistor is the problem. (Mode switch seems ineffectual too.)
All my NSL optos were purchased about 18 years ago from Allied. I'm a habitual overstocker, so I bought several VTL variants (VTL5C3, VTL5C1, etc.) back when they were made by Perkins Elmer.
Since DeadEndFX's project is relatively recent, they're probably using the xVive approximations, so their recommended value for R60 (220Ω) might not work with vintage VTL5C3.

Should I try an NSL-32SR3 instead?


build docs said:
The original uses an NSL32 type vactrol. Once tuned, resistance across the LDR side
of the vactrol in Morph mode will vary between 20K and 200K (resistance is limited to 200K by R48). The
harder the attack, the brighter the LED's will get, and vactrol resistance will drop, raising the drive level. We
have tried The NSL-32, NSL-32SR2, NSL-32SR3, VTL5C1, and the VTL5C9. All react differently, some faster,
some slower. Some have a wider resistance range, some are narrower. Our personal favorite is the VLT5C9,
with the VTL5C1 in a close second place. If using a VTL5C1, use of a 220r (R60) limiting resistor will provide
performance close to the VTL5C9.

I'm unsure. It sounds like both you folks (@Diynot and @DGWVI) used the Xvive VTL5C9, which is actually what worked best for DeadEnd, but did not work for you, correct?

Also, the build docs don't recommend a resistor value for the NSL-32SR3. :unsure:
 
Last edited:
Any value in the DNL spot kills the envelope effect for me, even low values under 100Ω.

If DNL is a theoretically infinite resistance (since no resistor means no connection) then small values under 100 Ω approach the exact opposite. I’d expect large values (1M-10M Ω?) to be more suitable in a DNL spot than a small value would be.
 
Any value in the DNL spot kills the envelope effect for me, even low values under 100Ω.

If DNL is a theoretically infinite resistance (since no resistor means no connection) then small values under 100 Ω approach the exact opposite. I’d expect large values (1M-10M Ω?) to be more suitable in a DNL spot than a small value would be.

Thanks man. I don't even know what that resistor is supposed to do. I'm just in the mood to try everything! :)
 
Yes, I did use a modern Xvive. I have toyed with throwing in an NSL R3, just never cared enough to pursue. Let us know what your experiments yield

I will experiment and report back.

The spec sheets show significantly different ON and OFF resistances.


Xvive VTL5C9
Xvive VTL5C1
Perkins VTL5C1
NSL-32R3
RON
1k5
25k
200Ω
60Ω
ROFF
2M
20M
50M
25M
TRISE
4ms
2.5ms
2.5ms
5ms
TDECAY
50ms
35ms
35ms
10ms

Resistance ON time changes with current, a value I can't know. For example, the PE VTL5C1 lists a R-on in the range of 200Ω - 20k, depending on current. I'm assuming low current? Forward voltage also differs (1.65v typical on PE 5C1 but 2.5v on the Xvive version.)

At the very least, I need to bump my R60 if I wanna stick with the VTL5C1. I suspect my LED is always on inside the opto.
 
Back
Top