CDXL Response

benny_profane

Well-known member
My CDXL build (corrected board) is having a number of issues.
  1. With ‘Level’ maxed out, I get my clean signal with a fizz mixed in, but no autowah.
  2. The response is sometimes extremely slow, and only kicks it on the decay of the note. This happens at various points of the sweeps.
  3. The wah doesn’t fall smoothly. It peters out in a very wobbly (think super fast phaser) way.
  4. Overall, it’s very weak (when it’s catching the input signal). Not at all as extreme as a wah sweep should be.
I used the Xvive VTL5C4/2 from SBE. I’ve tried various values at R9 (i.e., 2k2, 4k7, and 6k8). 4k7 seemed to get the closest. I’ve also tried various dual op amps (i.e., TL07, TL062, and TL022). The TL072 has the worst response.

I’ve tested the circuit both with and without a buffer before it.

I’m wondering if others have had good luck with the Xvive reissues or not. I know there’s one successful build report, but that’s it. In any case, any ideas?
 
Haven't breadboarded it yet (got a PlimSoul mod going on the breadboard), but I did run some simulations. Are you absolutely sure that C6 & C7 are the right values? You can read them both?

Try this: temporarily install (tack solder) a resistor in parallel with R5. Anything around 470K will work. See if that makes any difference. This test will at least prove or disprove that the filter Q is too high in your build. I remain skeptical that there is anything wrong with the Xvive part. Still possible, just no convincing evidence.
 
Sorry to be resurrecting an old thread, but I really wanted to build a CDXL classic. After waiting a couple months and never getting notified for the pcb, I figured I'll make myself a PCB based on PedalPCB schematic and layout as a way to get familiar with EAGLE. At the time I didn't know people were having issues with this thing, so I went and got myself the xvive reissue vactrol from SBE. Had this same exact problem but after messing around, I actually ended up lowering the R9 to 870ohm, and swapped the C5 to 10uF. Not sure how well it performs compared to the original but it seems to be working for me.
 
Sharing my experience with building a 440 envelope filter. Maybe it will help others understand problems with their build and save some frustration. As always, YMMV but this was my experience.

Using an Xvive VTl5c4/2 from SBE, the tip from RajahMoore about changing C5 to 10uf is mandatory to get anything useful out of the circuit. I would recommend substituting a 5k trim pot and a 500ohm resister in series as a replacement for R9. That will allow tuning to get the best response. The best value for you is going to be different depending on your pickups.

In my build there was too much distortion; part is due to the diode to ground in the envelope section (sounds like a maxed out distortion+) somehow interacting with the filter section. From other build reports I’ve read from others more knowledgeable than I, this appears to be from some sort of ‘cross talk’ within the chip. If you use a separate chip for the envelope filter this distortion is greatly reduced.

There is also a significant amount of ripple that sounds like bad distortion. With C5 at any value less than 10uf the ripple/distortion is unacceptable IMHO.

I was able to get some usable ‘quack’ from the circuit, but with C5 at 10uf the decay slope is way too long and has a very rapid, stuttering and not pleasing end to the decay.

Although NOS vintage VTL5c4/2’s are becoming exceedingly difficult to find, after a 30 minute search on Google I was able to locate an electronics parts supplier (not a broker) that had four NOS VACTEC brand VTL54c/2’s on hand for less than the price of the Xvives from SBE.

Received the NOS parts today. Put R5 & C5 back to stock, swapped in the NOS VTL5c4/2 and I can report the Xvive does not work the same as the originals. With the VACTEC brand in place, the ripple distortion is significantly reduced (the circuit even worked ok with C5 omitted!), the decay is much more ‘natural’ BUT the overall effect is much more subtle than what I got with the tweaked circuit using the Xvive. The Xvive and VACTOL are not even close in performance.

I’m not an EE so can’t test/verify but I’m inclined to believe that the spec sheet from Xvive for their VTL5c4/2 is made up, or the mfg is stamping some other product as VTL5c4/2. I’m can’t say that all Xvive are like this, but the 3 I tried all gave same poor results compared to the NOS parts.

For the reults described above I used an 1983 Ibanez with stock humbuckers. When I plugged in a 2024 Fender ‘American professional 2’ Stratocaster with stock (I.e. weak) single coil pickups I had no effect (just clean tone) with the Xvive unless I really beat on the strings and at that there was only brief stuttering. With the NOS VACTEC I had barely any noticeable effect but at least it didn’t stutter. Putting a preamp in front helped.

Bottom line - if want an original 440 effect sound, you’re going to have to track down an original VTL5c4/2. I assume Perkins, EG&G or Excelitas brand of VTL5c4/2 will give similar results to the VACTOL brand but I cannot confirm that.

If you like a challenge, use the Xvive part and see if you can modify the envelope shaping section of the circuit to make it work better. If you’re just looking for a decent envelope filter then I’d recommend building something different (I built the mxr120 and am pleased with it but that is also another circuit where the manufacturer of the chips is the difference between success and failure - if you go that route read all the build reports you can find for what chips to use and what mods will get you unity gain)

-Rob
 
The DOD 440 is a clever circuit. Dare I say's it's a bit "too clever."

The filter (IC1-2, C6, C7, R12, R13 & the dual LDR) form a Wien Bridge bandpass filter. For the filter to work properly, the two LDRs need to track closely. If they don't , then the Q is all over the place. It could even oscillate. If the Q is too high, then IC1-2 saturates and we get distortion. It's not enough for the LDRs to have the same dark and light resistances, they need to track at all light levels and respond to changes in light level at the same speed. It all comes down to how well the LDRs are matched by the mfgr. The MuTron III has a similar weakness in that the LDRs need to track, but that's another story.

The envelope detector can be fixed with one resistor: 220Ω in series with C4. Opamps do not like having their outputs shorted to ground, and that's effectively what happens when either D1 or D2 turn on. Yes, there can be cross-talk inside a dual opamps because sometimes they share a bias circuit. Upset one and you might upset the other.

OTA-based filters, such as the Maestro FSH-1, do not have problems with tracking.
 
Back
Top