Punch-in: a useful option

[Mark Hammer]

I started finishing off a number of circuits pedals I'd half built over the past year or so.  One of them is a build of Scott Swartz' PT-80 that I built into a Bare Box #2 that Steve Daniels had generously sent me.

The PT-80 is yet another PT2399-based 3-knob delay, that I had to add a few mods to...simply because nature compelled me.  You can see the schematic here:  http://www.generalguitargadgets.com/pdf/ggg_pt80_sc.pdf  Scott included a compander, and did his best to emulate the analog lowpass filtering of the AD-80 and DM-2.  Decent-sounding delay.

The nice "front porch" on the Bare Box (I'll post pics later tonight) can easily accommodate several stompswitches, so I installed two: one a standard 3PDT bypass, and the other a momentary SPST to serve as my "punch-in" option.  What do I mean by "punch-in"?  On traditional open-reel tape machines, it was not uncommon to find a punch-in feature, that let the user listen to the recorded tracks, wait for just the right moment, and insert a riff or whatever over top of what was already there.  Sometimes the punch-in would be a latching remote switch operating a solenoid/relay to turn the record head on, and other times it might be a momentary switch.

I happen to like momentary switches and momentary effects.  One of the things they let you do with a delay is introduce delay for specific passages and remove delay for anything after that in a very seamless way.

I implemented this feature in the PT-80 as follows.  The schematic shows a 1uf cap feeding the input signal to the 571 compander chip at pin 11.  I simply inserted a 1M resistor just ahead of that 1uf cap, and ran wires from each side of the 1M resistor to the momentary footswitch.  All the footswitch does is bridge/shunt the resistor.  With the added resistance, whatever is already in the delay path and regen loop reaches the output, but nothing new can get in. 

That's sort of the same thing as the tails function found on more and more delays, but in general, one has to hit engage, and then hit bypass to have evidence of tails.  The way I've implemented it, I keep the effect "on" (engaged), and when I want to insert delay on anything, I hold the momentary down for the relevant portion, and then release the momentary when I don't want to add any more delay.  Since the effect circuit is engaged whatever is inserted into the delay path while holding down the momentary continues to spool out, depending on delay time and feedback/repeats selected.  But for all intents and purposes, as soon as that stuff is done, it's like the pedal has been put back to bypass....until I step on the punch-in switch again.

So how is this "better" than tails?  You'll find the single action of holding down a momentary and releasing it somewhat more fluid than the double action of hitting the bypass switch twice.  Much more of a performance feature.  To defeat the punch-in feature, I simply bridge the 1M resistor with a SPST toggle (although I intend to replace it with a DPDT to use an indicator LED for status).

The same sort of performance feature could also be implemented in a variety of other effects, like chorus, phaser, and flanger.  Since none of these continue on in a "tails" fashion, the momentary connection is to be inserted at the mixer stage, where dry and wet are combined.  So, for example, one could stick a 1M resistor just before R22 in the CE-2 below and get chorus only as long as you hold the additional momentary down.  The momentary would need to be mounted externally, but you get the idea.  If you think of a player like John Scofield, who occasionally inserts drastic chorus sounds, that's just the sort of tool he needs.

I'm sure you clever folks can think of other applications.  I just wanted to get your minds wandering.  So don't just stand there.  START WANDERING!

[stallik]

Nice idea Mark. Iron warming up...

[notnews32]

Super interesting Mark... awesome work!
Any way you could upload a clip of how this effect sounds in the PT80?

[Craiz]

This is an absolutely lovely idea. But what does the compander 571 chip do that warrants the placement of the resistor there? Or rather, where would the resistor go in a pt2399 delay that doesn't have one?

[garcho]

i use the punch in on my Echo Base extensively, i guess from doing audio engineering i couldn't imagine building any delay without it; i certainly wouldn't want to play one. i started using "punch in" for other pedals too, sometimes you just want extra volume and distortion for one phrase, or want a crazy fuzz or envelope filter or ring mod for one quick jazz-noise but don't want the hiss or bleed. the punch in is useful even without "tails" effects, do it!!!

[Mark Hammer]

This is an absolutely lovely idea. But what does the commander 571 chip do that warrants the placement of the resistor there? Or rather, where would the resistor go in a pt2399 delay that doesn't have one?

Two very good questions.

The compander chip serves several functions.  First it manages headroom on the delay chip, just as it does with analog delays.  Second, it keeps noise levels low.  Not there there is all that much noise from many 2399-based delays, although sometimes folks aim for a little more delay time, and that can result in a bit of noise.  Scott Swartz also included the 571 because he was attempting to emulate the Ibanez AD-80 as closely as he could...and his tech info indicates that it is VERY hard to tell the audible difference between the PT-80 and AD-80.

The input to the 571 is essentially the very beginning of the delay path.  To have a delay sound close to the input (albeit with some treble loss resulting from the filtering) in terms of dynamics, the signal has to pass through BOTH the compression and expansion halves of the 571.  Passing through only one half will get you some odd-sounding sounds.  Consequently, the point where the input to the delay path is disrupted needs to be the input to the compander chip, and not a point farther along in the delay path.

Make sense?

Super interesting Mark... awesome work!
Any way you could upload a clip of how this effect sounds in the PT80?

I'll see what I can do.  I'll let you know whenI've put something up on Youtube.

[Craiz]

That makes great sense, thanks Mark. If I'm getting it,  you want it as early in the delay line as possible so you get the 'tails' but no new delayed signal.

I'm still a little lost as to where it would go in a 2399 reverb that doesn't have the compander - say, a Rebote, Neptune, or a Small Time (but this question is more about my trying to get my head around delay circuits - how to ID where the start of the delay line is, I suppose).

[Mark Hammer]

My bad.  I got distracted and forgot to add that part.

Let's use the Tonepad Rebote 2.5 as an illustration.  There is no compander ahead of the pre-delay lowpass filtering, as there is in the PT-80, but we can identify the critical path between the input to the pedal overall (IC2b), and the start of the delay path (the output of pin 7 in IC2b to the 1uf cap leading to the 12k resistor.

Stick the 1M resistor between pin 7 and that 1uf cap.  When the 1M resistor is bridged (i.e., a straight wire connection from one end to the other), the unit behaves as normal.  When one lifts the momentary switch and lets the 1M resistor do its thing, you should hear barely anything new in the delay output.

[Mark Hammer]

Promised pics and here are some.  You can see that the Bare Box allows for a nice spacing between the bypass (on the right) and the momentary (on the left).  I probably could have spaced them even a little farther apart. 

I installed something like this on an Echobase built into a 1590BB (the wide way), and also included another momentary X-wing SPDT, that allows one to momentarily insert LFO modulation, or momentarily suppress modulation, as selected via a toggle (that selects normally open or normally closed).  Even using the wide side, stuffing three switches in that space was tricky.  What I did was mount the bypass in the middle, seated as high as I could get it (skip the inside nut), and the two momentaries at the corners, seated as low as I could get them.  That way, if you aimed for the bypass, that's what you were most likely to hit.  The momentaries in the corners were a reasonably easy target. 

The usability depends on the accessabilty and the reliability of that accessability when the user is in motion.  The little "porch" on the Bare Box 2 provides a nice separation between the switches and any controls.

I have to confess stealing this idea from the Line 6 Tonecore pedals that use two switches under one foot treadle.  Press lightly and you do tap tempo with the higher-mounted micro-switch.  Press harder and you reach the lower-seated micro-switch for bypass.

[Craiz]

That clears it up enormously. I'm going to try and implement this into my multiplex build (thanks so much for your help on that too, by the way) - I'm  sketching out a momentary switch for the punch in but also a toggle to select whether it affect one delay line, the other, or both. This is going to be quite a lot of fun to play around with

Seriously, Mark, thanks. I started reading your posts five or six years ago when I got started with pedals and you've been a huge source of info and a pillar of the community ever since. It's really cool to be chatting about this stuff with you - it's too bad my friends would look at me pretty blankly if I showed up tomorrow and started gushing that Mark Hammer helped me out with my delay build.

[bloxstompboxes]

Interesting Mark. You built the pedal in sort of a backwards way, not a bad backwards way either, for the enclosure. I would expect the stomps to be on the other end and the knobs vice-versa. And the LEDs on the angled part. I like it.

[Mark Hammer]

It's a very flexible form-factor.  Maybe not as comfortingly solid as the cast-aluminum Bare Box 1, but still very flexible.  Not visible in the pics but the power and signal jacks are mounted along the rear skirt...or at least what I treated as the rear skirt.

Much like some municipal expressways, it doesn't really matter whether the controls are safely out of the way by being lower, or by being higher.  What matters is that they are at a different level so that one activity (stomping, setting controls) doesn't interfere with, or compromise, the other activity.  You don't want "worlds to collide".

[Kipper4]

defo thanks Mark

[Mark Hammer]

Addendum

So, I turned the volume and the delay mix up this evening, and found out that 1M may not be enough.  OR it may be just what you want, depending.  I can still hear some delay leaking through when not stepping on the punch-in switch.

So, perhaps 1M5 or 2M2 is a better value to use for more fully suppressing the delay signal until you want it.  Alternatively, a value like 470k or 680k may be appropriate for keeping just a touch of delay in the back, and using the punch-in to goose the delay on demand.  I supppose a person might use a 2M2 resistor, and a 3-position toggle to select normal delay (bridge the 2M2), a smidgen of delay until punch-in (stick a 1M in parallel with the 2M2), or no audible delay until punch-in (2M2 and no thing bridging).