Delay Feedback Limiter pt-80 pt2399

[thedefog]

Hey guys,

I've been trying to figured out an ultra-simple volume limiter for my pt-80 delay. I have a momentary switch wired to the repeat pot that temporarily sets the repeats at nearly maximum (threw in a small resistor in line to limit it to just below max), but the volume gets out of control quickly. I've toyed around with the idea of using an LED/LDR combo, and haven't found a simple enough solution. I wouldn't be opposed to clipping the signal with diodes.

Just picking your brains. I'd love to hear some ideas for this, as I'm better at just building and repairing electronics rather than designing them.

[zyxwyvu]

The Echo Base (http://www.diystompboxes.com/smfforum/index.php?topic=60662.0) uses two diodes to ground right after the feedback pot.

[sean k]

I've thought about this myself and figured out that the idea would be to split off the output and rectify it then send it to a 555 set up in astable mode which fires a 4017 which then controls the gates on a 4066 set up with trimmers across the feedback. Therefore at 4 or 8 bit counts you can have the trimmers set so the delay builds then is cut after four or eight repeats.

I found the astable on the Paia Ducker and it answered my need to have a fixed output to the 4017.

[chi_boy]

Does it go into that wierd run away mode where it gets into a feedback thing that gets louder and doesn't want to stop?

[JKowalski]

Perhaps you can try taming the diode clipping method with a tube-screamer type log amp. That way it actually sounds okay (not tooo grating) when it goes out of control?

[Taylor]

I actually like the way the Echo base sounds with the clipped feedback. Maybe I'm weird. Doesn't sound too harsh to me.

[igor12]

2 Si Diodes (1N4148) in parrallel and opposite directions (look up schematic for for MXR Distiortion plus for example app). Germanium Diodes will cut the volume more.

[thedefog]

Thanks for your input guys. Greatly appreciated. I really actually dig the way some of the older analog delays get wildly out of control as the repeats continue (Thinking like the space echo at the end of Karma Police). So as long as the volume limiting method is musical, I don't mind distorting 

[Taylor]

Well, I'm not an analog fetishist, so you can take my opinion as just that, but I think the Echo Base can sound pretty analogish when it feeds back, and it gets more like that the longer the delays get.

[earthtonesaudio]

You could send the feedback path through a separate circuit which drives an LED as the level increases, then place an LDR in the voltage divider going into the feedback pin, so the signal is attenuated rather than clipped.  Basically an optical compressor.

[jacobyjd]

You could send the feedback path through a separate circuit which drives an LED as the level increases, then place an LDR in the voltage divider going into the feedback pin, so the signal is attenuated rather than clipped.  Basically an optical compressor.

Better yet, add clipping diodes in parallel w/ a blend pot--then you can have both!

[thedefog]

Hey Guys,

I did the diode limiter like on the Echo Base. That pretty much works perfectly. It has just enough volume from the continuous feedback without the extreme volume increase. I highly recommend this option as it is extremely easy and doesn't compromise tonality of the original design.

Now, moving on, I want to add an LFO to this for delay modulation. I built the LFO section from the Magnus Modulus and am looking to implement that into this pedal. I think there was a post somewhere on here from someone that did this already.

[Mark Hammer]

Um, the PT-80 already does limit the feedback signal; it passes through the compressor half of the 571 compander.  What it seems people want to do is to achieve endless feedback/repeats that do not increase in amplitude.  Well, the answer, I'm afraid, is that achieving a flawless unity-gain feedback loop is near impossible.  There will be some loss of signal along the delay path, if only because of the lowpass filtering.  So, to maintain apparent level in spite of the filtering, you'd need to increase feedback beyond unity.  Do that, and the content below the filter cutoff will start to howl.

If it was an entirely digital circuit such that what came back through the feeback loop was "data", pure and simple, you'd have no problem.  But at the re-entry point to the compander, what you have is "signal", not data.

Those BBD-based devices that have no companding will often use diode-based soft limiters in the feedback return path, but they don't work all THAT well.  Moreover, where a diode-based limiter will add the sort of grit to a flanger signal that gives the comb-filtering more bandwidth to apply notches to, it does little to "sweeten" up discrete repeats.

The PT-80 is a clever and decent delay, but if you want endless repeats that don't get all weird on you, you need to go fully digital.

[cpm]

My problem with the "infinite repeats" is that it usually grows up in volume, reaching the clipping rails. Such huge volume increase can be really annoying.

The diodes in the feedback loop will tame the feedback level until its clipping threshold, which should be at a higher level that normal guitar level, or else the "controlled" feedback would clip too. This way going into infinite feedback still there is a big volume increase, limited only by the diodes Vf.
Also, diodes may leak, so long feedback settings may get further repeats some distortion becasue of this leakage.

I've got the solution though
When using a compander, we usually make the feedback loop on the outside (from the expanding output into the compressor input)
If the feedback signal is tapped before the expanding, this is, going back into the compressor without getting expanded, there you have a infinite feedback chaos at a controlled volume (with feedback resistor)