PT2399-based "Space Echo" idea? (in progress)

[diablochris6]

A crazy idea popped into my head; create a PT2399-based delay that resembles the Roland Space Echo. I know it won't sound close, but it could be a fun exercise. I am trying to stand on the shoulders of giants. I basically pulled components of delays from Madbean, Josh of 1776, and Dead Astronaut. I am indebted to each one of you.

Each PT2399 acts as one of the playback heads of the echo. The individual delay controls allows for the playback stagger of the Space Echo but also allows for other configurations. I originally wanted to use switches to activate each delay chip like the Space Echo, but after seeing Dead Astronaut's Abductor Delay, using level pots could create some interesting sounds. All the delay signals go into the tape filtering section of MB's Zero Point Super Deluxe. I added a Baxandall tone stack for individual bass and treble control of the repeats, like the Space Echo (I took the values from ROG's Ginger pedal). I also figure that the second part of the op amp could recover any volume lost. The signal then goes through a feedback loop and through the mix control. I also added a dry killswitch, tails, and a pad to add the simple modulation board that 1776 sells. The only other thing I might attempt is a way to patch in a reverb pedal and a switch to put the reverb at the beginning of the circuit or after the delays only.

But first things first. Would anyone be willing to check out my schematic and tell me if everything looks good? I need an extra set of more experienced eyes to tell me if I am making a major error somewhere. Would it make more sense to cascade the signal off of pin 15 rather than 14 to get some of the dry signal in the next delay chip? Thanks!

[wavley]

Did you see the Multiplex that 1776 sells?  That actually gets you most of the way to what you're going for.  If I remember correctly there was a really big thread here while it was being developed, you might check it out for ideas.

As the owner of an RE-501 and a RE-101 I can tell you that your idea is pretty cool, folks will tell you that you'll never get the true sound, and that's true, but all of the fun of a space echo is all the complex delay patterns, pitch shifting, and self oscillation that make a space echo.  Sound quality+Fun=Awesome, but Fun=Awesome too.

[anotherjim]

Looks perfectly feasible to me.

Tape playback heads each get the same signal off the tape. Here, you have some degeneration by daisy chaining several delays, so all but the first is an echo of an echo. That said, at reasonable delay times, the PT2399 is reasonably clean, so it may well not be a problem.

You have some compression on the output (IC5a), whereas tape compresses at recording - but I can't think why it would make any difference doing it that way around, and you have the feedback compressed too.

Stocks of DIL PT2399's are now critical 

[slacker]

Looks good to me, just a couple of things you might want to look at. You probably want some power supply filtering caps on the 5 Volt line, either at the output of the regulator or on  the vcc pins of the PT2399s. I'd also add a protection diode to the 7805, cathode to the output anode to the input, this protect the regulator in case the input gets shorted.
I'm not sure your volume recovery opamp will work wired up like that, you might need some resistance between the wiper of the treble control and the input.

[diablochris6]

Thanks for taking a look at this.

I actually made the Multiplex Jr. (I put it in my Earthquaker Devices Disaster Transport Sr. clone). It sounds great. I also studied the schematic of the standard Multiplex to help me with this design, which is why there might be a little similarity. I also looked at Madbean's Zero Point Super Deluxe and Deadastronaut's Abductor for guidance and inspiration.

Jim, as far as the signal chain goes, I am thinking about cascading the delay chips off of pin 15 of the previous chip to try and get a little more of the clean signal running into it. I will test off of both pin 14 and 15 to see what I like the best. I thought about tying all the delay times together like the Abductor, but I think that by having individually controlled delay times, you can stagger them to get the Space Echo timing or think of it as having adjustable tape heads (kind of like having 3 Echoplexes).

Thanks for the tips, slacker!

Thanks again for the support.

[deadastronaut]

but I think that by having individually controlled delay times, you can stagger them to get the Space Echo timing or think of it as having adjustable tape heads (kind of like having 3 Echoplexes).

or both..

might be a switching headache though..hmmmm...

i like the idea of the baxandall too....that'll be nice.

[Luke51411]

Interesting circuit, I'll be following the progress

[Luke51411]

It looks like the price of pt2399s is going up in anticipation... I bought from mklec last time for like $.50 (I think... could have been $.90) in quantities of 10 or more now they are at $1.80 in quantities of 100 or more...  
looks like its tayda for me now... I've done alright with the pt2399 I've bought from them in the past.

[deadastronaut]

oops!..

[diablochris6]

Rob, if I tweaked the Voltage control of the PT2399s like in the picture I drew up, could I use a simple 3PDT switch to determine whether the delays are in sync or have their individual controls? Even more devilish of an idea...use a 3P4T rotary to get individual control, synched delays, PT2399 #1&2 synched only, or PT2399 #1&3 synched only.  If this is feasible, it would be madness.

[deadastronaut]

hmmmmm...

looks like its worth breadboarding...

i thought the switching would be a nightmare..


gonna need a bigger boat...

[~arph]

I have used a max395 and a simple uC to do the switching of four PT2399. You could select every combination of 4 playback and 4 feedback heads.

[~arph]

You can also look into the CD4053 three bits to select one of eight.

[diablochris6]

I updated the schematic to add the new features bouncing in my head.

I built this up using some delay module pcbs and a breadboard. I had to tweak the gain on the opamp, but delays are audible now. After jacking up some PT2399s by putting them in backwards, I got it up and running with a few issues...

1. The delay times are very short. The effect doesn't venture very far from a slapback echo. I need to extend the delay times.
2. There is too much of a volume drop between each delay stage. This might have to do with the filtering as well. The delay doesn't have much of a stagger between the PT2399s. This could be a problem with me not discerning between the short delay times.
3. Even though I am not sure why, the 2nd and 3rd delay ICs seem to depend on the Level pot of the 1st delay. It shouldn't, because I have the signal piggybacking before the Level pot.
4. When I try to wire the delays so that each PT2399 has its own individual Delay time pot, I don't hear any delay. Each PT2399 is receiving the ~5v from the regulator.

I am still tinkering around, but it anyone has any helpful ideas, I am all ears.

[diablochris6]

Remedied a few things. Found out I had a bad 5v regulator. It was causing a bunch of weird things to happen, including some oscillations in the delay lines. It sounded almost like a bad tape, which would be cool to simulate, but this is already too difficult of a project. Upon further inspection, one of the PT2399 legs wasn't seated correctly in the socket. I found some decent working PT2399s and have reduced my problems down to the following:

1. The delay times are way too short on each PT2399. The most I can get is a slap back effect.
2. The 3rd delay sounds pretty distorted compared to the other two. Could be a bad IC, could need work on the filtering?
3. I'm not sure, but the level pots on the 3 delay lines seem to be fighting against each other at times.
4. The feedback pot goes into self-oscillation extremely fast after a small turn of the pot, especially if the volume of delay 3 is turned up.

I'm getting closer...

[diablochris6]

Here is an update:

The delay sync rotary works really well, as you can see me noodle with the controls in the video above. You can get some interesting rhythms. I slapped together a Rubadub reverb and placed it in the delay send/return, and it sounds phenomenal. I still need to acquire a modulation board to test that, but it will come soon.

The only two problems I still have are:
1. The feedback goes into crazy oscillation way too quickly. I will see if a different value pot would do the trick. A log pot could tame it a bit.
2. When turning the bass pot, things get all scratchy. When the pot is full CCW, there is no delay sound at all. You can hear this at about the 2 minute mark in the video. Usually, the presence of the scratchy pot has to do with DC issues, correct? I will measure the Baxandal circuit with my multimeter, but is there something wrong with my design of the tone circuit?

[diablochris6]

I forgot to update this project for those of you who were interested.

Here is a demo of a prototype board with an arpeggiator going through it. There were a few design errors, but they were fixed before production of more PCBs.

A guitar demo of the fully operational circuit:
https://soundcloud.com/mr-beans/sagan-demo

A few pics of my build:



The build documentation: https://drive.google.com/file/d/0B7QeV_7r70dTUHY1VHhMS3BkaFVuZjJGdVMzOXVyYTI0NG5V/view?usp=sharing

And the crazy vero layout: http://tagboardeffects.blogspot.com/2015/10/sagan-delay.html

I did have PCBs made, but I just sold out. I might have some more made soon, if anybody is interested.

[anotherjim]

Good work.
Did you fix problems 1 & 2 mentioned in your previous post?

2 first - return R5 to Vcc (5v reference) instead of ground.

1 - Feedback scheme looks a bit messy. When the clamping diodes conduct, the impedance seen by C24 in the output of the 3rd delay is very low and forms a high pass filter passing over 50kHz with the pot at max! A high pass that raises as feedback increases isn't a bad idea as we usually don't like bass howl, but it probably needs scaling so the corner frequency is nearer 300-500Hz at maximum feedback. I'd suggest some resistance in series with this path. 10k (ish) after C24 could solve it and maybe reduce the tendency to "sudden death" howl-round.

Having the low impedance of the diodes off the pot wiper will play hell with the pots taper, probably making it excessively non-linear giving an "all or nothing" response, so some resistance should maybe be placed in series, after the pot wiper?

[diablochris6]

Yes, the issues are resolved. I attached the EQ section to some voltage rather than ground. As for the feedback loop, I used a log pot and even removed the clamping diodes in my build. The soft clipping after the Pt2399s helps get the same sound, and the Baxandall can be used to round out the tone of the oscillating feedback.