Radio Shack "Electronic Reverb"

[Mark Hammer]

While assisting in the cleanup of former forum member Peter Snow's basement, I came upon a pair of these.

Here is the circuit they contain.


As you can see, the "reverb" is a simple MN3207-based short delay, with 8 poles of lowpass filtering on the BBD output; the "repeat"/feedback signal tapped after 4 poles of lowpass.  As something from the dawn of karaoke, I gather the mic preamp was for making amateur singers sound impressive.

Any point in subbing NE5532 or LM833 dual op-amps for the AN6552 chips it comes with?  Anything you would do with this unit, apart from give it to Goodwill or cannibalize it for parts?

[Rodgre]

I've had one of these since they first came out in the 80s. I actually used to plug a speaker into the RCA output and crank the input level with a guitar plugged in. I wasn't that smart, but I also used it as a distortion into an amp by doing the same thing, but turning the master volume down on my amp.

It's a pretty lo-fi delay effect and the longest delay time is about 75ms, so okay for slapback and manual flanging effects. Not bad for $39 in 1985.

Roger

[Mark Hammer]

Yeah, I was thinking about making it a sort of chorus, but the filtering perplexes me.  Would a chorus with THAT much lowpass filtering sound any good?  And if one left off the 4 poles built around Q4 and Q5, taking the wet signal from the same point where the repeats are obtained, would that sound acceptable?

[Rob Strand]

Yeah, I was thinking about making it a sort of chorus, but the filtering perplexes me.  Would a chorus with THAT much lowpass filtering sound any good?  And if one left off the 4 poles built around Q4 and Q5, taking the wet signal from the same point where the repeats are obtained, would that sound acceptable?

With a BBD the required filter cut-off frequency depends on the delay time.  As you increase the delay you need to decrease the cut-off frequency.   You can see some recommendations in the old BBD datasheets.

A delay units typically requires more delay than a chorus.  So most delays units have lower cut-off frequency.

It's the cut-off frequency more than the order that impacts the sound.  The order affects the high frequency noise and aliasing.   If you use a high order filter you can usually push the cut-off up a bit.  Since delay units are usually pushing the BBD to the limits you might see higher order filters used.  The cut-off on that unit is about 2.7kHz with a large peak in the response.

What you really need to do for a chorus would be to push the cut-off frequency up a bit.  For best sound you would be better off replacing the filters altogether.  Copy the filter design from a known good chorus.

If you don't want to change the filters you could try:
- take output from output of stage 1
- feed BBD into input of stage 3 and take output at stage 4
- feed BBD into input of stage 2 and take output at stage 3

The cut-off is little low in all these cases.  There's also some peaking.   The simplest case of the first option might be OK.

FYI, there's also a filter in the input side.  Not too bad as is.   Overall sound depends on both.

[Mark Hammer]

I'm used to seeing a lot more anti-aliasing ahead of the BBD, followed up with 3 or 4 poles of LPF, post-BBD.  As you correctly noted, there IS some pre-BBD filtering, but it pales in comparison to the post-BB wet path.  Weird.

IF I was to squander some time trying to chorus-ify the unit, I'd likely reduce C22 to something that would move the delay range lower, tap the wet signal at C27, and up the corner frequency of the Q2/Q3 filter sections, conceivably by reassigning the C28/C30 caps to where C23/C26 currently are.

But all of that is assuming I have the time.

[Rob Strand]

I'm used to seeing a lot more anti-aliasing ahead of the BBD,

I was just talking generally.  Anti-aliasing is before the BBD.  Reconstruction after the BBD.

followed up with 3 or 4 poles of LPF, post-BBD.  As you correctly noted, there IS some pre-BBD filtering, but it pales in comparison to the post-BB wet path.  Weird.

IF I was to squander some time trying to chorus-ify the unit, I'd likely reduce C22 to something that would move the delay range lower, tap the wet signal at C27, and up the corner frequency of the Q2/Q3 filter sections, conceivably by reassigning the C28/C30 caps to where C23/C26 currently are.

But all of that is assuming I have the time.

Here's a good starting point for some reasonable mods:
- take the output from Q3 - so lift C32+ and wire C32+ to Q3 emitter,
then make the following changes to the filtesr:
- C23, C25 1n8
- C24, C26 180p

Leave the BBD input side filters as is.

Two identical filters stages to simplify things.  With those few part changes the overall response, including the effect of the input filters, is a very close approximation to the overall response of the Small Clone.

You can tinker with the oscillator cap to set the delay.
For CE 2 response use filter caps 1n and 220p (instead of 1n8 and 180p).

[Rodgre]

Before modding the filtering of the circuit, I'd be curious to hear what would happen with simply applying a control voltage from an LFO to the delay time as-is, just to see if there's anything interesting to be had from the get-go.

You folks are smarter than me about this stuff, though. I love reading your posts like this!

Roger