'Space and Time' reverb with pre-delay and tails

[samhay]

Last year I spent some time getting to know the Belton/Accutronics BTDR-2 reverb 'brick' and decided to design something that wasn't closely based on the data sheet examples. One aspect of the design was a mix control that panned between fully wet and dry signal.
http://www.diystompboxes.com/smfforum/index.php?topic=104667.msg939946
I never boxed this up, as I couldn't get it to do tails bypass nicely, and I got distracted by other things. I got it back on the breadboard recently, and after a few major tweaks and a little mission creep involving a PT2399 for pre-delay, I present the 'Space and Time' reverb with a fully featured PT2399 (pre-)delay.



A cheap and cheerful, and somewhat boring, 5 min clip of some ambient noodling while fiddling with various settings.
https://dl.dropboxusercontent.com/u/11996927/SAT_Reverb.mp3

It has 5 knobs:
Mix: pans between fully dry and wet signal.
Reverb: pans between the output from the PT2399 and the reverb brick, which is directly fed from the PT2399. ccw you have a normal delay, while cw you have reverb fed with a pre-delay.
Delay: sets the PT2399 delay - this is the 'time' control.
Regen: is your delay 'repeats' pot, which also feeds back the reverb - this is the 'space' control.
Tone: is a Sallen-Key LP filter, which works on the wet signal after the delay/reverb are mixed. This can give a pretty extreme treble cut and when used on the delay can give a nice low-fi tape echo sound.

Why did I add pre-delay? Until I started looking, I hadn't seen it done before, but I had come across fancy reverb software that has this feature. I have also seen a number of multi-fx builds with a delay and reverb in the same enclosure, which seems like a pretty good combo. I am not the first to think of using a PT2399/Belton brick combo to do pre-delay either - I subsequently discovered the EQD Ghost Echo, which is pretty similar, and I hear there is a similar project in the Madbean skunkworks.

I haven't built this yet, and it is still on the breadboard. Comments most welcome.

[bluebunny]

Bl00dy h3ll, Sam!  Have I got to go and buy yet another brick??   

[samhay]

No, just build this one instead (edit - I am teasing of course).

For reference, I have been using a long brick here. However, I wonder if it might work better/different with a short one - looks like I might need to buy another one too.

[deadastronaut]

what, no modulation?...

sounds great man, cool idea, sounds like a good allround ambient machine...nice.

cool name too.

[samhay]

The pre-delay adds some nice amplitude modulation when you turn the repeats up.
I tried to directly modulate the brick, but it wasn't entirely pleasant - as far as I can tell you have to modulate the V+, which is a bit akward and the brick makes some horrible noises if it is starved too much. In the end, I figured the brick is modulated anyway. Once the BTDR-3's are out, it will be relatively simple to modulate the delay pins, so we can have another round of designs then.

A question - is 5 knobs on a 1590B going to be too crowded?

[deadastronaut]

5 knobs, not really if you use davies knobs  its fine..

the brick however can take up space,  if it were me i'd go 1590bb for a more comfy build...

[bluebunny]

No, just build this one instead (edit - I am teasing of course).

Ha ha!  I'll let you and Rob fight it out - it's his Chasm that's next on the list...

Edit: Who am I kidding?  I'll end up doing both...  

[Mark Hammer]

That shore is right Purdy, Sam.

I covet!

[midwayfair]

Why switch 1B as part of the bypass? Doesn't that setup drop the dry level at lower mixes?

[samhay]

That shore is right Purdy, Sam.

I covet!

Thanks Mark

[samhay]

Why switch 1B as part of the bypass? Doesn't that setup drop the dry level at lower mixes?

The switching probably needs a little more explanation, and I am open to suggestions, as it isn't perfect.
The mix is different to most delay/reverb mix pots, which typically mix in wet signal while maintain dry signal at unity. Here, the mix pot pans between fully wet and fully dry signal. If you have it hard cw, then there is no dry signal hitting IC1B. Consequently if you then lift/ground the op-amp end of R8 and leave the mix pot connected when fully cw, you will get no output once the wet signal has decayed.
If we switch in dry signal via SW1B, IC1B can now pass dry signal at unity gain. If the mix pot was still connected, the dry gain would be anywhere from 1-2 depending on the mix pot setting.
What's the down side? You will get a jump in dry signal (to unity gain) when switching to bypass. This will become more pronounced the further cw the mix pot is set. The workaround is to set the mix pot up in a conventional manner, but then you miss out on the really lush reverb and weirdness.

[Kipper4]

Yippee. Just ordered my first brick.
Looking forward to this one.
Expect lots of simpleton questions Sam.

[deadastronaut]

really lush reverb and weirdness.

that's what we want.......beep.....................beep..................well i do anyway..


@marc: i fight to the death ...but ive been killed 23,179,002. times, and counting..  

cool project sam..

[slacker]

Very nice Sam.

You will get a jump in dry signal (to unity gain) when switching to bypass. This will become more pronounced the further cw the mix pot is set. The workaround is to set the mix pot up in a conventional manner, but then you miss out on the really lush reverb and weirdness.

I don't think you can do the switching any better than it is, any delay, reverb or modulation pedal that lets you have 100% wet is going to do the same thing, it's just the nature of the beast. If anything having tails makes the transition nicer than using true bypass or other non tails switching.

[samhay]

Thanks Ian.
That's mostly the conclusion I have come to about the switching too.
However, while one advantage of the switching now is that it can be done with a DPDT rather than electronically, if I used a FET that switched in R11, I could set it to turn on quite slowly to fade the dry signal in. I will have a bit of a play with this, but I am not sure if it will warrant the extra parts count.

[samhay]

Still tinkering with this on the breadboard while I cobble a layout together.
I added a LED from pin 7 of the PT2399 to ground (schematic updated). This is Jon's (midwayfair) trick as used in his 'Hamlet' delay to stop the PT2399 from distorting (which is not pleasant). This works beautifully here too and also keeps the reverb brick from clipping* when the regen is turned up.
Using an ultrabright red LED (blue had too high a Vf to prevent chip distortion), this will make a nice overload indicator on the box.

Is it known why this works? The slightly better datasheet below says pin 7 is for current control of the demodulator. Pin 8 is the current control for the modulator and if your put the LED here it has the opposite effect - would make a 'nice' chaos switch.

http://www.musicfromouterspace.com/analogsynth_new/ECHOFXXX/pdf/PrincetonTechnologiesPT2399EchoChip.pdf

What's worse the 1 PT2399 clipping? 3 PT2399s clipping in a reverb brick.

[kingswayguitar]

cool

[samhay]

Another quick clip in response to a query about how to make a reverb sound like a flying saucer.

https://dl.dropboxusercontent.com/u/11996927/SaT_UFO.mp3

Turn the feedback up, then back the tone off

[samhay]

Measured the current draw today with 2 different BTDR-2s. Was pleasantly surprised - only about 80-90 mA. The vast majority of this was the brick and the PT2399, but could probably get away with using a 78L05 at a pinch.

I think I will use a relay to do the bypass switching. This will only cost another 10 mA or so, and will make the wiring a lot tidier.
If I can find space, a PIC to do the switching and a second relay for true bypass would be nice - nothing like a bit more mission creep.

[deadastronaut]

a PIC..

your getting posh..