Doomsdelay Device

[Brett Sinclair]

Maybe i'm overambitious with my limited electronic skillz, but here goes... 

I thought it would be a kewl to make a PT2399-based delay with an external loop in de feedback path. I read about the PT-80 mods by moosapotamus that allow processing of only the "dry" or the "wet" signal, but i'd like an effects loop that processes the signal on every repeat. Put an EQ in there with some rolloff frequencies and some slight chorus... or a whammy (pitching up every repeat one octave)... or a reverb... you get the idea 

The new Moog MF-104Z Analog Delay seems to allow exactly that, so i borrowed some ideas there 
I studied the PT-80 and Robote delays and read the Geo and AMZ articles over and over again looking for clues... and this is what i came up with...

Controls:
* Delay Time
* Feedback
* Dry/Wet Mix
* Input Level
* Output Level
* switches to select Delay Time Range short / long
* switch between External and Internal feedback loop

Connections:
* Input/Output/Send/Return (doh)
* Expression (for modulating Delay Time)



I "borrowed" most pieces from other schematics and examples.
There is probably a lot of stuff i'm not considering that i should... i'm learning.

If someone wants to help improve or donate a Moog Delay... shoot. 
All comments welcome!

[troubledtom]

dang, it's a pretty bad ass start. 'gonna have to look at this for a while.
               i like the name too
                             - tom

[varialbender]

I'm not very tallented, so I'll just ask... did you remove any filtering that used to be in the feedback loop? (I'm assuming there was filtering in there) I prefer nothing in there so that I can have full control over the repeats with what I put in the loop, but some prefer filtering to make it sound nice without anything in the loop anyway... either's fine... which way are you going about it?

[moosapotamus]

Totally cool ideas. I've always wanted to add that effect loop to the regen/feedback (repeats) path in my PT-80, too. It looks like you are trying to create a whole new circuit (I haven't had a chance to fully absorb it, yet, tho). Alternatively, my preference would be to start with either one of the Rebote delays at tonepad or the PT-80. They are already great sounding circuits. So, no need to reinvent the wheel, IMO. Just add the effect loop to the regen path, along with whatever other cool mods you might want to try.

Referring to any of those schematics, for the effect loop in the regen path, it looks to me like the send & return should cut in just after the wiper of the repeats pot that's tied to pin 14. That would allow you to control the amount of effected regen. Or, if you wanted to control the amounts of straight regen and effected regen separately, you could add a separate pot that taps off at the same point as the existing repeats pot and goes to the send jack. Then cut the return in to the line coming off of the wiper of the repeats pot. You might get some signal feedback, crosstalk, or something unpredictable going on without mixing properly. But, it might sound cool, too... dunno.

Someone please correct me if there is some reason why any of that wouldn't work, or if there is a better way... I haven't actually tried it myself, yet.

~ Charlie

[Brett Sinclair]

did you remove any filtering that used to be in the feedback loop?

Yup, to some degree... I removed the additional filters after the delay that are used in the Robote and PT-80.
I will redraw the delay in and out part in a more readable way, tonepad-style.

Referring to any of those schematics, for the effect loop in the regen path, it looks to me like the send & return should cut in just after the wiper of the repeats pot that's tied to pin 14. That would allow you to control the amount of effected regen. Or, if you wanted to control the amounts of straight regen and effected regen separately, you could add a separate pot that taps off at the same point as the existing repeats pot and goes to the send jack. Then cut the return in to the line coming off of the wiper of the repeats pot. You might get some signal feedback, crosstalk, or something unpredictable going on without mixing properly.

Well i started looking at ways to mod the PT-80 and Robote, but ended up redrawing/redesigning the circuit...

On the input buffer i wanted hi-Z, but with the possibility to feed the delay a hotter line-level signal, so there's an attenuator there.
Signal then goes to both the output and feedback mixer. If the external loop is not used, you'll just get a buffered (but inverted) version of the input on the send. If it is used, the input signal is first processed and then delayed (instead of processing after the delay - how it would work as you described).
The return is phase-inverted again.

After the delay, the delayed signal is mixed with the input in the feedback mixer and goes back in the loop.
Instead of an output mixer where you can specify the level of the delay signal, i wanted a dry/wet balance control like on the Boss RV-3. (And now i've read the "panning for fun" article i'll change those resistor values to R1=R2=R3=R4=15K, R9=10K and R5=51K... )

I dunno what is better... processing before or after the delay. (I figured before was best which lead me to the scheme above, so that the signal hits the effect loop unaffected by the delay first time... but would it make much difference?)

[Processaurus]

Very cool, I've tried this idea with a mixer & a ton of patch cords and its the best, I'm suprised Moog is the only company doing this.  It sounds awsome with an EQ, you can make almost a parody of an analog delay.  Anyway, heres a few thoughts I had looking this design over:

I like the master volume.  If you wanted to get a bit of boost at the top end of the volume you could make the feedback resistor of U3 larger (like 250K, to give 5x boost), and avoid the helpless feeling you get when things don't "go to 11" .

You also might want to think about a way of gently clipping run away feedback, like maybe two back to back LEDs in the feedback loop of U4. 

I don't feel like it would make much of a difference if the signal was delayed pre or post loop, if whats coming out of the delay is a faithful reproduction of what went in.  Purists might want the loop after the delay, to "warm it up" or something.

You could use a one TL074 for all the opamp duties. 2 ICs total, not bad.

Anyway, thanks for sharing this design.

[Brett Sinclair]

Thanks for the suggestions! Going to try out that soft-clipping...

I've updated the scheme, but i think i need to check on the delay section again (redrawn detail at bottom)
Guess i'll start putting this together this weekend to see how it turns out... 

[Processaurus]

Just curious if anythings happened with this, Brett.  Its a brilliant idea.

[Pedro Freitas]

Hello!

I did this a few weeks ago. Installed jacks so the feedback path would go through an external
effect.
I installed the jacks between the 1uF and 22Kresist. in the feedback loop with no impedance
matching opamp or whatever. Worked great 

Pedro

[sodapop]

Is this similar to what Robert Kelley does with Boss DD-3? 

http://www.robertkeeley.com/product.php?id=41

[Brett Sinclair]

Due to lack of time i still didn't get round building / experimenting with this... 
All parts eagerly waiting though...

I don't know if the Keeley mod is a fx loop in the feedback path or just an insert before the mix with the dry signal?

Pedro, you used the Robote2.5 or the PT-80 circuit?

cheers & happy new year lads! 

[Processaurus]

Is this similar to what Robert Kelley does with Boss DD-3? 

http://www.robertkeeley.com/product.php?id=41

No, thats an effects loop for the total delay signal, rather than the regeneration of the delay having to go through the effects loop with every repeat.  The second option is more appealing to me personally, no offense to Mr. Keeley.

Pedro, thats great news, I'll try it sometime soon with a rebote delay I have laying around, if it was the Rebote you tried it on?

[Transmogrifox]

Is this similar to what Robert Kelley does with Boss DD-3? 

http://www.robertkeeley.com/product.php?id=41

It think for something like the DD-3, you couldn't do like what is suggested for the PT-80 type delay.  I would suspect that all the dry signal mixing & feedback is done digitally.

You could make a bufffer/opamp mixer with send/return and turn the dry to 0 on most delays and pull this off.  I did this with my Digitech RP-20 but it didn't turn out great since I hadn't included a pot to control the feedback level so to limit the repeats.  It was very interesting what was going on when I used a wah pedal in the delay loop.  I was using it to control the delays:  When it got to regenning real hard on the low frequencies, I'd sweep the pedal forward so to start attenuating the low freq's, then the highs would come up and start to get out of control.  If I stayed on top of it, i could keep it under control to a certain extent, but it would never completely die so things would really start to sound like crap in short order since the signal got so distorted each time through the wah pedal and D/A A/D process with each repeat.

[Pedro Freitas]

It was a normal PT80 from GGG.
If you look at the schematic there you'll quickly notice the 22K/1uF pair in the loop.


Pedro

[caress]

can anyone confirm this schematic?  it looks really interesting...

[Jaicen_solo]

Am I alone in thinking this: ... or a whammy (pitching up every repeat one octave), is the best idea i've heard all year?

[John Lyons]

This looks very cool  ( granted, a  2 year old post...) and would be a nice thing to get modulation, distorted and or treble cut or bass cut repeats. Would put it in the reggae, dub catagory along with the space echo/echoplex and other characteristic delay devices.
Way above me head but I'd build it when/if it gets sorted out!!

John

[DWBH]

I posted a new topic regarding this delay some time ago...

I still drool myself when I see that schem, but seems like few people put themselves into doing this...
Perhaps I'll make a pcb for this...

[soggybag]

This is a very interesting idea. I remember seeing this posted a year or two ago.

This schematic is a little confusion there is a lot going on. It looks to me like it uses a non inverting input but has an inverting buffer in the output mixer.

[theblackman]

someone on experimentalistsanonymous.com did it with a pt80, albeit without all the mixing/buffers etc.