EHX Deluxe Electric
Mistress — Squealing
on Color

What came in

An Electro-Harmonix Deluxe Electric Mistress, the classic flanger/chorus pedal built around a bucket-brigade delay chip. Cosmetically fine, and not in true bypass — this revision runs a buffered output, so the pedal is always somewhere in the signal path whether the effect is engaged or not.

The reported fault: squealing that changed pitch as the Color knob was adjusted, only appearing in the last 10–20% of its travel toward maximum.

What we found

Safety checks first, as always — earth continuity and insulation resistance both came back clean, so the bench work could proceed.

The squeal itself pointed straight at the feedback path. Color controls feedback in this circuit, and instability that only appears near maximum feedback is a classic sign of a loop that's right on the edge of oscillating. The detail that confirmed it: turning the Range knob changed the pitch of the squeal. Range sets the clock frequency that drives the delay chip — so whatever was misbehaving had the clock circuit coupled into it.

Working schematics for this exact era of Mistress are hard to pin down — Electro-Harmonix revised the design more than once, and at least one schematic we tried didn't match the board in front of us (no Direct/Flanged dual-output arrangement, for one). Worth saying plainly: the board is the ground truth, the schematic is a reference, and when they disagree, you trust what's actually in front of you. Tracking down the right one at all was only possible thanks to fan-built schematic archives — the kind of unglamorous, decades-deep documentation effort that keeps old gear like this repairable in the first place.

Tracing the Color pot turned up the real story. The stamped markings on the can read CTS 250V — a 250K audio-taper pot. The original spec calls for 10K linear.

Someone had been in this pedal before, at some point, and fitted the wrong part — wrong value, wrong taper, both. Out-of-circuit measurement confirmed it at roughly 150K, well off the 10K the circuit was designed around.

That was fault one. But swapping in a correct 10K pot on its own didn't fully clear the squeal — which meant something else was off too. A genuine factory alignment procedure for this exact production run turned up, and it spelled out a feedback-trim adjustment as part of the normal setup process — meaning a degree of oscillation right at the top of the Color range can be a known characteristic, tamed by trimming, not necessarily a fault on its own.

Following it through properly turned up fault two: the internal clock, which should run at roughly 35kHz, was sitting at around 13kHz. Badly out of spec, and exactly the kind of thing that would couple into the feedback path in a way that tracks with the Range control — which is what set the clock speed in the first place.

Two faults stacked on top of each other, with the worse one disguising itself as a quirk of the design.

What was done

A 10K linear pot was fitted in place of the incorrect 250K audio unit, properly D-shaft and matched to the original 24mm body height. The clock was brought back up to the correct ~35kHz. With both corrected, the feedback trim was adjusted per the original factory procedure — dialled to sit just below the threshold where the loop wants to oscillate.

The pedal also turned out to be missing its Color knob entirely — a separate, unrelated problem, but one that needed sorting before it could go back to its owner.

How it left

Quiet through the full range of the Color control, with the feedback trim now sitting correctly relative to a properly valued pot. Working the way it would have left the factory in 1978 — the flange and chorus tones the Mistress is known for, with nothing fighting underneath them.

A close match for the missing knob was tracked down, a Davies 1410, measured carefully against the panel's remaining knobs for shaft fit and dimensions before fitting.