Boss DD-20 Mods and Observations

[12bass]

Recently I became interested in improving the sound of my DD-20 Giga Delay. To that end I replaced the original cheap/low performance M5218 and NJM2100 op amps with modern low noise/high performance OPA1612 and OPA2365 respectively. Additionally, I added film bypass capacitors across the electrolytics which couple the various analog stages. Bypass capacitors help retain transparency in AC coupled circuits.

The end result is that the DD-20 now sounds very close to the unaffected signal (when using a true bypass A/B switch for comparison). Before the mods it seemed to compromise the tone somewhat. It still loses a very slight amount of clarity, but the difference is unlikely to be noticed in a gig situation.

I also did some experimentation with placing the DD-20 before the input of my Eden WT-500 and using it in the effects loop. When used before the amp, the DD-20 adds a noticeable amount of background hiss, even in bypass mode. This noise increases as you turn up the effect level. I was hoping to reduce the background noise with the low noise op amps, but I'm not sure how much effect they have had. At some point I would like to compare my modded DD-20 to another stock unit.

What I found, however, was that placing the DD-20 in the Eden's effects loop reduced the hiss considerably, such that it is hardly noticeable. Thus, for lowest noise, I would recommend using this particular pedal in an effects loop, if possible.

Another finding is that, when used in mono, the DD-20 sounds better using the "B" input and output. My suspicion is that the "A" input gets split and fed into the "B" input when used in mono, and that this reduces the input impedance by half, thus loading down the input and dampening the signal somewhat.

For those who are curious, the A/D-D/A converter is an AK4552 (IIRC), which is an AKM 24-bit/96 kHz CODEC with decent, but not stellar, specifications. No idea of the sampling frequencies/bit depth used in this application. My suspicion is that most of the "sound" of this pedal comes from the analog circuitry, not the digital sections.

[12bass]

After some more testing I found that the DD-20's -20 dBu setting is slightly overloaded on transients when placed in the Eden's effects loop. Switching to +4 dBu (to access: hold down on/off pedal at power up) cleans things up, with a slight increase in the noise floor.

[fretlessrock]

I'm a big fan of the DD-20, and it was the heart of my looping rig for a long time (paired with a RC-20XL). I can say that I just played it the way it came out of the box and didn't worry about tweaking any hardware. I don't think I ever put it in the fx loop, but the difference in signal levels and impedance are probably causing most of the difference. Interesting note on the overloading. I always had some kind of buffer device in front of it and noticed that it didn't like anything too hot on the input. That sensitivity switch trick came in handy a few times. What I'm not sure of is if it alters the input sensitivity and output, or just one or the other. I've seen the settings described as both.

Thanks for the post and keep posting any other findings as they come about.

[syndroam]

Had a DD20, had to let it go because it messed up my signal!!! When it was in my chain I had to turn up my bass volume twice as much as i normally would.

[B.C.]

I had one but sold it. Loved the sound, but just not the delay interface I was looking for. I think I'm better off with a DD-7 or something like it.

Nice mods my friend. Typically when I am modding a pedal, the input op amp(usually used as an impedance buffer or something of that nature) is the first I go for, with the input caps a close second.

Typically I would replace electrolytic with film if at all possible, for noise and lifetime improvement.

You probably could, if you are very careful, swap A/D converters. Now that would be really interesting

[12bass]

Quick update:

Recently I was finally able to get a hold of the DD-20 schematic. This makes it much easier to see what is going on in the circuit and where best to make modifications. After noticing that the modded DD-20 was using around 230 mA with the new op amps (stock DD-20 uses around 175 mA), along with finding that the bypassed tone was still a bit veiled, I thought it might be worth having another look. A friend and I tried to see if we could find any obvious oscillation using an oscilloscope, but nothing was immediately apparent.

In light of the high current draw, I replaced the three OPA1612s with OPA1642s, which have JFET inputs and use less quiescent current. The modded DD-20 now uses around 185 mA, or a little more than stock. Also, I think that the low noise FET op amps may have reduced the hiss slightly. Additionally, I reduced the values of the capacitors in the feedback loops of the op amps, which opened up the treble response somewhat. Now the pedal's buffered non-delayed audio path is quite flat up to 20 kHz and sounds only slightly different than true bypass. The delay path remains somewhat rolled off in the upper treble, but this appears to be due to design limitations.