Improving a DM-3

[Mark Hammer]

I made a Boss DM-3 clone a ways back, from this PCB layout:  http://i7.photobucket.com/albums/y263/GreggPics/BossDM-3.gif

As with most analog delays, when you aim for more than 2 or 3 repeats, the sound degrades pretty quickly, and sounds pretty harsh.  The runaway feedback ain't pretty.

The problem is that audio quality in the upper mids and higher suffers most as the signal passes through more stages (which it does when you feed it back).  So what I did to fix it was to add some shallow lowpass filtering on the feedback path.

There is a 22k resistor in the path between the "Intensity" (i.e., feedback amount) pot and pin 5 on the NE570 compander chip.  I swapped that for a 10 and 12k in series, installed like an arch, soldered at their junction, to yield the same series resistance.  From their junction, I ran a wire to a SPDT on-off-on switch that either left the resistor pair working like a 22k fixed resistor or connected a .01uf or .022uf cap to ground from the junction, forming a single-pole lowpass filter in the feedback path.

The single-pole aspect is important, since it permits progressive filtering.  That is, there has to be some signal left to filter on each repeat, and the shallowness of the filter assures that only some of the mids and highs are attenuated each pass through.

I don't really notice it all that much with the .01uf cap, but the .022uf cap setting is wonderful.  You can dime the Intensity control, and get lots of repeats but they never clutter the soundscape, and they never get dirty-sounding or harsh.  If you want bright repeats (and they sound pretty decent at short delays with little feedback), just flick the switch.  And if you want warm velvety chocolate pudding repeats, flick it again.

Recommended.

[oldschoolanalog]

Thanks for the tip Mark.  Think I'll give it a try on the next madbean Aquaboy (DM2) I have on the "build soon" list.