DOD 440 Envelope Filter build report

[RickL]

I finally got this one working after several months of off and on frustration. I used the layout at GGG and tried to use a homemade LDR/LED combo which didn't work. When I broke down and bought the spec'd VTL5C4/2 it worked properly. BTW I used a TL072 for the opamp.

Like a lot of EF's it needs the sensitivity (called 'range' on the layout) turned all the way up to get a decent sweep. I modded this for better response by adding another 470k resistor in series with R5. If you build it you might just want to use a 1M resistor for R5.

I also used a dpdt centre-off off switch to choose three different sets of caps for C6/C7. Using 0.01u, 0.01u and 0.033u I got choices of 0.01u, 0.02u and 0.043u. The 0.043u isn't too useful (too muddy) so I'll probably change the 0.033u's to 0.022u's.

While it works just fine, I wouldn't say it's outstanding in any way. Worth building but I wouldn't put it at the top of the list.

Mark, I'm sure you have a bunch of ideas for improving it. Let me know and I'll try them and post the results.

[Brian Marshall]

It is definitely subtle, and for my gear, not a funk machine..... it actually works best for playing leads in front of a distortion.

[jsleep]

I wouldn't call it subtle.  Mine seems to quack loud and clear.  It is basic and certainly not one of the best EFs around, but I believe it has it's place in the DIY world.

This project would work well for someone who has never build a EF and wants to learn from it and end up with a somewhat useful box.  I know I learned a lot when I built mine.  The fact that its a fairly simple circuit that uses the LED/LDR makes it fun.  You can also easily (obviously) add a manual control by switching in a pot for the LDR (can't remember the value that worked best).  Something similar to the "Q" pedal or Wild Mouse when you switch to manual mode.

Rick's point about the VTL5C4/2 should be emphasized.  I too couldn't figure out the right tweaks to get a sub part to work.  Small Bear has the VTL5C4/2 of course :-)

JD

[Aharon]

The VTL5C4/2 has a dark resistance of about 400K/on resistance of 150 ohms.An LDR approaching those specs should work.The problem is that most LDR I've measured are more like 5 or 10M.
For  tweaking LDRs, putting a resistor in parallel to taylor the resistance should work.Mounting the two (LED and LDR)in sockets should help also.
Aharon

[RickL]

When I first built it I tried replacing the 220k resistors that are in parallel with the LDRs with a dual 500k pot to tweak the response. The best I ever got was a very subtle sweep if I had the guitar volume, sensitivity and range pots in just the right place.

I've had good luck with homemade LDR/LED combos in other pedals but not this one.

FWIW I swapped the 0.033uF caps for 0.022u's. That position of the switch (0.032uF) is still pretty dark but it's at least a little more usable. If you're going to use switchable range caps I'd either go smaller (i.e. 0.005u, 0.001u and 0.022u) or just have two choices (0.01u and 0.02u).

Mark where are you? Usually you have suggestions for 3 or 4 more pots and switches.  :wink:

[Mark Hammer]

Oh I'm here all right.

Using this schem as reference point - http://www.generalguitargadgets.com/diagrams/ef440_sc.gif - there are the usual recommendations:

1) If you don't get enough sweep, make R5 bigger (e.g., 680k).  Sensitivity is determined by the gain of non-inverting stage IC1a.  When gain is altered by changing the pin 2 to ground resistance, bass rolloff changes as gain is increased, such that at max sensitivity there is less response to lower notes.  As with the MXR Dist+, loss of low end in this stage can be fixed by increasing the value of the cap.  Let me emphasize, though, this does not change the bass output of the pedal/circuit, just how much the envelope follower responds to lower notes.

2) If you want the sweep to linger, make C4 bigger (e.g., 4.7 or even 10 uf)

3) If you want to slow the attack, stick a small-value resistor (<1k) between D2 and C4

4) If you want to use if with bass, stick a small-to-medium value (say, 10k) resistor in series after C8, and run another similar-value resistor from the R1/R2 junction to the free end of the resistor after C8.  Now run a 100k terminating resistor to ground.  That will mix in some straight signal with the swept filter so you don't lose bottom during the sweep.  Note that since the dry signal is tapped before processing, and since the filter build around IC1B is inverting, the filter will be subtracted from the dry signal at the output, yielding a swept notch.  Very funky in its own right.  You may have to tinker with mixing resistor values to get a decent balance or simply stick in a trimpot and tweak.

5) R15 injects some DC from the supply into the LED to adjust the default brightness and the starting point of the sweep.  Not having used one I can't say how much useful variation this provides.  However....the min and max DC offset available at the R6/LED junction is set by resistors R13/R14.  To move the range downward so that initial starting point is lower, make R14 smaller by the same amount you make R13 bigger.  To move the starting point upwards, do the opposite.  To allow finer start-point adjustments, make R15 a 47k pot and divide the "missing" 50k between R13 and R14.

The ways of Wien Bridge filters are a mystery to me.  I have no sense of what needs doing to adjust Q.  It may be worth making any switching of filter caps independent, though, to see what that does to the passband.  So, use two toggles, one for C6 and another for C7, for rangeswitching purposes.  Certainly tinkering with the passband would have implications not only for filter-inly sounds but for dry+wet mixtures.

There.  Let it not be said the central agencies don't help the line departments!  :wink: (little civil service joke there, folks)

(BTW, are you a PSAC person, or are you unaffected by current goings on?)

[RickL]

I'll give those mods a try and report back tomorrow (already did the R5 mod, see first post).

Yep, I'm PSAC. Apparently I get to deprive Canada of accurate measurement in a couple of weeks. That'll show 'em.

[RickL]

1) first part of this suggestion already done, see OP. Second part (increased response from bass notes) didn't seem to have any effect. I tried as high as a 0.47uF cap and didn't hear any difference in the sweep on lower notes. I don't actually hear much difference in response between high and low notes stock so maybe there's not much to improve. I didn't try it with a bass, maybe you need to get lower than a guitar's low E to hear a difference.

2) This works good. I used a spdt centre-off switch to parallel 4.7 or 10uF with the stock 1uF and I heard a definate difference in release time. I'll keep this one.

3) This works too but it's not as noticable. I wired a 1k pot between D2 and C4 and I could hear differences in the sweep between one end and the other. I'll probably just use a switch to choose between 0 and 1k or maybe 0, 500 ohms and 1k.

4) This didn't work at all. I ran 6.8k resistors from each of the points to the outer lugs of a 5k pot. The wiper of the pot went to the 100k terminating resistor and the effect out. It didn't matter where I had the pot set, the volume was very low and the sound was muddy. If I lifted the wet sound or the dry sound from one end of the pot I got the other sound clear and loud. I think an active mixer might be needed for this to work correctly. Or my interpretation of the mod is wrong. I figured with the pot in the middle it would be the same as dry and wet each having 9.3k (6.8k + 2.5k) going to the 100k and I could tweak the pot for the best mix.

5) I didn't try this one. I'm happy with the sweep range stock.

6) I'm not so sure this is a good idea. When I tried it I found that if C6 was much smaller than C7 I got uncontrollable self oscillation, pitch controlled by the range (marked level on schematic) pot. The oscillation didn't happen the other way, with C7 lower than C6. If I set the range pot low enough with C6 at 0.01uF and C7 at 0.02uF I could get an interesting triggered burst of oscillation that sounded almost like a sub-octave was following the envelope. Something to play with maybe.

[Mark Hammer]

Well, a couple of swings and misses, but it appears there was a base hit out of it.  Let us know what else does and doesn't work.