DOD 440 reissue - how'd they do that?

[Mark Hammer]

I made myself a DOD 440 Envelope Filter with a home-rolled 1-LED/2-LDR optoisolator.  I foolishly went with a white LED, which was likely a poor choice.  It gets the sound, but the sweep is somewhat all-or-none, rather than the sort of responsive quality one hears in the demos.  I did stick a 220R resistor in series between the diode and the 1uf cap to ground just to get a slightly more human-quality attack time, but I'm pretty confident that's not causing the problem. 

I figure I probably need to bias up the LED, via the Range control (over on the right hand side in this drawing), by replacing the 22k resistor with 24k, and sticking a low-to-medium value resistor between the pot wiper and LED to keep the current modest.  That way the Range control can feed a low voltage bias signal to the LED to keep it just near the point of illumination.  What comes from the rectifier section over on the left will simply add to that default bias voltage to make the LED brighter.  Use of the stock parts would probably work just fine with the circuit as is.  I'm just looking to make a few simple changes to allow a homebrew opto to behave better.

I followed David Difrancesco's info from an old issue of Polyphony and installed a 3-position Q control for the filter, which turned out to be a good choice*.  But when I was poking around on-line for some Youtube demos of the pedal to do some A/B comparisons, I was surprised to find out that the 2014 reissue of the pedal has an up/down switch, which the original didn't.

Anybody know how it's being done?

(* Replace the 430k feedback resistor with a 390k unit, and use a toggle to add in 47k and 100k on top of that.  Or simply stick a 100k pot in series with the 390k.)

[Govmnt_Lacky]

Don't know if this will help but, from the info I have read about LDR response times, the best light source to go with is either a Yellow or Green LED. Most LEDs respond best between 500 and 600nm wavelengths.

[PRR]

> went with a white LED

White LEDs need quite high voltage, nearly 4V.

This rectifier is one-peak, from a 9V supply, so it can hardly make 4V onto the LED.

Is it really too late to use a red LED?

[PRR]

The rectifier is *peak-to-peak* (oops).

Still in all, asking almost 4V from 9V through a lossy opamp and two Si diodes may be asking too much.

[Mark Hammer]

Thanks for the replies.  The LED and LDRs were glued together, though I suppose I can always make another.  Thankfully, both are cheap these days.  I figured I could conserve envelope voltage by using a schottky for the series diode, and simply feed the diode a couple volts from the Range pot to bring it in range of lighting.  But a more suitable  LED may simply be easier.

I'm still curious about how they do downward drive and whether there are any changes to the filter as a result.

[samhay]

The schematic above has a couple of errors. The schematic below looks more plausible.

As far as adding a down sweep switch, you can flip the orientation of D1 and D2 with a DPDT. The range control won't work very well like this though, so I wonder if they redesigned the follower.


(view for larger version)

[Mark Hammer]

I'm not so sure that would work.  Keep in mind that to sweep downward, the LED has to start out bright, and then the envelope signal subtracts from the DC voltage illuminating it.

[daeg]

Old thread, but to finally answer the original question; the DOD FX440 Reissue is not based on the DOD FX440, but is rather a modded DOD FX25.

There is an extra opamp stage in the envelope section and while there are some other tweaks it just looks like simple inversion of the envelope / control signal.

[Rob Strand]

There is an extra opamp stage in the envelope section and while there are some other tweaks it just looks like simple inversion of the envelope / control signal.

I've implemented a down-mode envelope filter in DSP and it's not as simple as just doing the reverse of the up-mode case.   The best example as a template I could recommend would be the Mutron unit.   For the down mode, they increase the detector gain and add a bias to set the starting point for the (upper) filter cut-off.

[Mark Hammer]

There is an extra opamp stage in the envelope section and while there are some other tweaks it just looks like simple inversion of the envelope / control signal.

I've implemented a down-mode envelope filter in DSP and it's not as simple as just doing the reverse of the up-mode case.   The best example as a template I could recommend would be the Mutron unit.   For the down mode, they increase the detector gain and add a bias to set the starting point for the (upper) filter cut-off.

Yup.  Down-sweep is NOT simply the mirror image of up-sweep.

Picked/plucked strings start out with lots of harmonic content and then lose it as the initial pick attack decays.  When you sweep upwards, you hold off that harmonic content and introduce what's left as the sweep moves up.  When you sweep down, you start out with all that harmonic content and then accentuate the fading out of that harmonic content as the sweep descends.  The result is that it sounds like more "abrupt" and severe filtering with downsweep, compared to upsweep, which feels a little more relaxed.

The "solution" to a more pleasing downsweep is twofold.   First, the sweep needs to start just a little lower than the maximum high point of the upward sweep, so that the contrast between initial and final harmonic content is not so great.  Second, the speed of the sweep  has to be slowed down just a bit.  If one implements the sweep reverse in the MXR Envelope Filter, which has variable Attack time, you'll find that slowing down the Attack time helps a lot.

[POTL]

There is an extra opamp stage in the envelope section and while there are some other tweaks it just looks like simple inversion of the envelope / control signal.

I've implemented a down-mode envelope filter in DSP and it's not as simple as just doing the reverse of the up-mode case.   The best example as a template I could recommend would be the Mutron unit.   For the down mode, they increase the detector gain and add a bias to set the starting point for the (upper) filter cut-off.

Hi, always wondering why is gain being added to the detector when the phase is switched in the detector?

[Rob Strand]

Hi, always wondering why is gain being added to the detector when the phase is switched in the detector?

The simple answer is without it, it doesn't sound right at all.   The more techy reason is it pushes the filter down into low frequencies quicker.  Also you want the filter to stay low during the note decay so you need more gain to hold it there.

I suppose the flaw is trying to keep the envelope touch sensitive by having the envelope tied to the guitar signal level.   A more synthy solution might be to use a trigger (which fires on the note) but then it follows a fixed trajectory (more like the higher end noise gates).   For up-mode the touch sensitive thing is nice for dynamics and shaping what comes out with your fingers but for down-mode maybe the whole idea is wrong.