Ultra simple delay / phase shifter help

[isophase]

Hi all,

I did search around but can't seem to find this specific device and I need some help.
I am looking to build an ultra simple delay pedal that would be capable of delaying the signal up to about 1 second (more if possible) 100% wet, and no feedback.
In other terms, what I am looking for is a sort of « phase shifter » or kind of all pass filter that can shift in time up to about 1 or 2 seconds. I would like the device to be super simple with only one control knob to alter the delay time. A transparent unit with regard to sound quality would be perfect but Sound quality is not very important.
So how can I do this?
Is it possible to do this with BBD chip or best to do it digital?
Thank you in advance for your help!

[ElectricDruid]

If you need delay up to 1 or 2 seconds, you're probably going to need a digital solution. It *could* be done using multiple long BBDs like the MN3205 (Coolaudio and Xvive still make clones) but noise gets to be a problem, and the circuit is large.

Something like the FV-1 would be a possibility. That chip can produce delays up to 1 second at decent quality (32KHz sample rate ) and doesn't need a lot of other parts. There are kits and PCBs available for FV-1 delay projects that you could use, and then all you'd have to do would be to disconnect the dry signal to give yourself a wet-only output.

The other common chip used for delay is the PT2399, but that won't manage a full second of delay (roughly 0.5sec max) and the quality gets very poor at longer delay settings.

[isophase]

Thank you Electric Druid,
You answered all my questions
I will have a look at the FV1 chip and project kits see if I can adapt one for my special pedal.
Thank you!

[PRR]

> you're probably going to need a digital solution.

Or magnetic tape. I used to rig delays from 0.2sec to 40 minutes. Not to mention storing sound for 30 years.

{checks calendar} Oh, sorry, this is the 21th century.

[Digital Larry]

Or magnetic tape. I used to rig delays from 0.2sec to 40 minutes.

Hot on the heels of Frippertronics, Frippertronics volume 2!

Seriously though, 40 minutes!  Wondering how you managed the tape spooling.  7 1/2 IPS * 2400 seconds = 18,000 inches/12 = 1,500 feet.

[Ripthorn]

If you use the ES56033 you can get 1 second delay time, but I don't remember how clean it is. Much cleaner than a PT2399, for sure. A digital solution will do a better job, but at much higher cost and complexity.

[ElectricDruid]

ES56033 is very similar to the PT2399, except it only has 32K RAM instead of 44K, so I'd be surprised if it manages to do much better either in terms of time or quality. I've never tried that chip though, so I defer to those who have.

[puretube]

PT2395 would be similar to the `99, however with external memory for longer delay:
http://experimentalistsanonymous.com/diy/Datasheets/PT2395.pdf

If someone manages to address a multiple of 256k DRAMS, even longer delaytimes than the 800ms mentioned in the datasheet are possible (or higher clock-speeds for better quality).
[haven`t been there ...]

But for "simple" and "easy", I`d prefer the FV-1 (although *digital*).

Interestingly, *Belling* used to make a similar set of "echo"-chips as *Princeton*, about 20 years ago: the BL0305 &  the BL0306 (gotta go down my dusty basement, to see if they`re still there ...).

https://www.datasheetarchive.com/pdf/download.php?id=67361b30c2cec23165d6c79f74d8f2a99fd5b8&type=M&term=bl0305

https://www.datasheetarchive.com/pdf/download.php?id=f0fed5b73b46b895be71ed22d6a2b5540543af&type=M&term=BL0306

[ElectricDruid]

PT2395 would be similar to the `99, however with external memory for longer delay:
http://experimentalistsanonymous.com/diy/Datasheets/PT2395.pdf

If someone manages to address a multiple of 256k DRAMS, even longer delaytimes than the 800ms mentioned in the datasheet are possible (or higher clock-speeds for better quality).
[haven`t been there ...]

It'd be easy enough to add further flip-flops to provide higher bits to the address counter, if it were a simple incrementing count. The problem starts when the internal control logic tries to jump to some other location and can't address the extra bits ("page" bits, essentially) that we've just added. So whether such a thing is possible would completely depend on what the Main Control Logic part of that internal diagram does with the address counter. You'd have to record some of the bus state and have a good dig through it - not impossible or even difficult with a modern logic analyser.

I don't have such a tool, or I might be tempted to give it a try, just for curiosity.

[puretube]

Found an old thread on addressing extra-DRAMS, but that lead nowhere, iirc ...
https://www.diystompboxes.com/smfforum/index.php?topic=25250.20

And a "compatible" *Mitsubishi* chip datasheet:
http://www.experimentalistsanonymous.com/diy/Datasheets/M50195P.pdf