Digitech RDS-3600 MIDIfication project with Arduino

[damien]

Hi I just bought this very hard to find Digitech RDS-3600 digital delay and I really like it so far!
It's an unbalanced rip off the Lexicon PCM42, with the 7 segment display with the time in ms, Hold/freeze feature and modulation on the clock (speed, depth).



Here's my project.
I want to add the famous Lexicon 42 double opto-limiting feature on the Input path and on the feedback path with something like a CLM50 chips, to replicate the behaviour, at least a little, of the Lexicon.
I also will implement a midi sync feature to calculate the delay time and control the timing of the unit from a Arduino board. The Arduino will also control digital potentiometers to control the audio input gain, feedback level, audio output gain, mix level, and the 3 delay parameters, Delay time, mod speed and mod depth.

The MIDIfication should be a smooth process, it's just receiving midi CC from the computer via midi and translating the CC value to the resistance value via MCP41100 digital pots (SPI) or equivalent.

I'll post about my journey on this thread and share the required mods, schematics, arduino files, pictures and more.

All pots on the analog path are 100K single or dble deckers so no problem but...
The first thing I'll have to deal with is the 500K pots to control the speed modulation of the LM324 responsable for the PWM to clock.
500K digital pots don't exist unfortunately and I think the maximum R would be 100K.

1.Is there a way to produce the same timing with a 100K digital potentiometer by replacing the value of the capacitor in the RC circuit? Or maybe a total different way?





Thanks for your help about this! I'll keep you posted along my journey to hack this fun DDL

Damien

 

[EthanSpaulding]

Hey Damien,
Cool project. I don't see how this circuit functions. Is U48 even an oscillator? Or is it part of creating the VC that controls a VCO? Someone with a big brain would be needed. Post the full schematic, if you don't mind!

 

[damien]

Hi Ethan, U48 is a LM324 Single Supply Quad Operational Amplifiers IC.
I saw this chip many times on PWM motor/fan regulator, it's pretty common.It's known for its very low slew rate.
I automatically thought it would be used to deliver the pulses for the timing clock. But it's just a guess. Maybe it's just a VCO LFO like in old synths, many use LM324 like the old korg mono.
https://3.bp.blogspot.com/_rScBRKlTdoE/TNQraMhovzI/AAAAAAABsz8/PJkljExPMXE/s1600/monotron_sch.jpg
https://www.onsemi.com/pdf/datasheet/lm324-d.pdf
Here is the LM324 part of the circuit :




Here is the full schematics for this series of vintage Digitech digital delays :

https://music-electronics-forum.com/attachment.php?attachmentid=50276&d=1535492462
I am waiting to receive my digital pots and arduino to dive into it. I'll check the signal all along the path from the VCO jack in to the clock line. (U41 output)

I never came across this kind of circuit, it's going to be interesting to see how it really works.
From what I can see, once the delay time is set witout any modulation the modulation part is added on top of the steady carrier.
For exemple if the delay is set to 1000ms, if I add modulation, it never goes under 1000ms but adds the oscillation upon it. 1000, 1100, 1200, 1100, 1000, ...
1000ms would not be the center line of the oscillator but the baseline.

It's a cool effect, very grainy especially when multiplying the sample time x2 or x3 it's grungy, good grungy.

I'll post more once I get my parts to start the project.

 

[damien]

I decided to go with a pair of AD5206BRZ100 (6 Channel 8-Bit Digital Potentiometer 100K) so that I can bridge 5 of them to make a 500K pot and the rest of taps for the 6 other 100K pots. I'm waiting to receive SOC24 to DIP adapter PCBs and I'll be good to go with this project. I'll post as soon as everything is available.

 

[EthanSpaulding]

Will you stack 5 potentiometers, then multiplex between their outputs?

Or s there a way to stack them , then use SHUTDOWN to select the "wiper"?

 

[damien]

Not sure now, I never worked with the SPI arduino library, I'll need to explore what's possible and if there's some sync posssible between wipers.

The ad5206 data sheet is there : https://www.mouser.ca/datasheet/2/609/ad5204_5206-3118604.pdf

Worst cas scenario, I was thinking just to hardwire all the different poles in series and adjust their respective wiper position from 0 to 100k on the same busses to give 0 to 500k. I need to try on a breadboard beforehand to see if it's going to work. I'll get into the details when Santa will deliver the remaining components

 

[EthanSpaulding]

I can see how you could easily add digital potentiometers in parallel for lower value pots, but the series connected digital pots involves switching from wiper1 after the first 100k is passed and from wiper2 after 200k etc. I think it would be well worth your efforts to reanalyze and scale any function that doesn't already use 100k controls. The "speed" control is the prime example. I still dont understand it. It doesn't look like an oscillator to me. It seems like it should find equilibrium...

Todays challenge: What is U48 on the schematic doing? Specifically the two sections of U48 that make up the "Speed" control.

 

[Voyager10]

The first half of U48 ought to be a relaxation oscillator (Relaxation Oscillator using Op-amp) which produces a square-wave output. The second half is then an integrator (Op amp integrator - Wikipedia) which turns it into a triangle wave to modulate the delay time.

The 'speed' pot is being clever, in that it increases the current drive to the integrator as the speed increases, so the triangle wave amplitude stays the same.

So AFAICT the circuit shown there is wrong. It will be DC stable with pin 6 pulled close to the +12V rail, pins 7 and 5 at the -ve rail (or 0V, it's not shown). The integrator will eventually give up when pin 1 hits the supply rail, so it'll end up with pin 2 pulled negative and pin 3 mid-rail.

It's possible they're doing something weird by having pin 6 super-close to the supply rail, and feeding pin 5 from a very high impedance, but if so it's beyond my understanding...

 

[Hubbub]

I decided to go with a pair of AD5206BRZ100 (6 Channel 8-Bit Digital Potentiometer 100K)

Note that your LM324 supplies are 12V and ground, so expect the outputs to go to 0 to +10V or more. Page 6 of the AD5206 datasheet has an "Absolute Maximum" on the pot terminals of "VSS, VDD" and also a "VDD to VSS" maximum of 7V. Trying to use a 7V maximum rated part on a 12V supply will result in the part failing quite prematurely. May even let the magic smoke out.

 

[Voyager10]

Do you have a scope? Can you measure what the signal is at the R129/R127/speed pot junction? My guess is that it's a square wave, which increases in amplitude as the 'speed' increases (so amplitude * signal period is a constant), centred around the 'mid rail' voltage seen on U48 pin 3.

If so - and this might be considered heresy - it's just a low frequency square wave, there's nothing special about it. Connect a DAC to the output from the Arduino and synthesise it in software.

Even wilder suggestion: remove U48 altogether and drive the R119 / pin 14 junction direct from a DAC. Then, all of delay time, speed and width can be driven from software. You could do funky things like synchronising the modulation rate with the MIDI clock...

 

[damien]

Thanks @Voyager10 and @Hubbub . I'll check at my friends' lab whith a scope to see what's the signal type at this point.
I'm pretty busy with other projects right now on my freetime, but it's next up on my list so as soon as I can dive into it, I'll try your suggestion and hopefully I'll be able to generate unavailable shapes from the arduino, that would be so nice!
Syncronizing to the midi clock was my first intent and I thought about being able to control everything else via midi and this is were I am
I would really like to implement an opto limiter on the input and feedback path to make it behave like a Lexicon PCM84. It shouldn't be too hard to modify it for that.
I'll check all voltages and signal before implementing the AD5206 because it's about 10$. I don't want to smoke them
Thanks for all your informations, I've never heard about a relaxation oscillator before. I like projects like this because it always brings new knowledge.
Happy new year, all the best in all your audio projects!
Damien

 

[damien]

I just ordered a "toy" scope on aliexpress for that kind of purpose .

 

[Hubbub]

FYI its one thing to measure the signals in operation, but you also need to look at possible swings at power-on and power-off. A safer guide is device supply voltage(s). When crossing a boundary from one supply to another it's best to assume either might be zero while the other is 100%. So some sort of current limiting is needed to prevent inputs being damaged. (Outputs tend to be beefier and I speculate they will probably survive about 10x the current an input will tolerate.)

 

[damien]

Do you have a scope? Can you measure what the signal is at the R129/R127/speed pot junction? My guess is that it's a square wave, which increases in amplitude as the 'speed' increases (so amplitude * signal period is a constant), centred around the 'mid rail' voltage seen on U48 pin 3.

If so - and this might be considered heresy - it's just a low frequency square wave, there's nothing special about it. Connect a DAC to the output from the Arduino and synthesise it in software.

Even wilder suggestion: remove U48 altogether and drive the R119 / pin 14 junction direct from a DAC. Then, all of delay time, speed and width can be driven from software. You could do funky things like synchronising the modulation rate with the MIDI clock...

Hi,
I checked the signal at R129/R127/Speed pot with my baby scope and I did not get a square wave but the signal was swinging around 0v Positive and negative with its amplitude and speed increasing together proportionally.
It started very slow like 1 oscillation per 2 sec and could go around 5Hz max.
Here is the screen when doing tests :


If I could bypass the oscillator and just use a DAC converter from the arduino to send the oscillatior signal for modulation that would be awesome!

Let me know what you think about that scoping, and if you need other measurements ;-)
Thanks for your help!
Damien

 

[damien]

I scoped further on the path to the clock signal before U47.
I got a good old triangle signal whose frenquency changes with the speed knob and its amplitude increases with the width knob.




I could probably generate a triangle from the arduino at this point or a bit later on the path to the clock INs (before U41).

 

[broklee]

Can you post some pictures of the boards inside of the unit? Could better assist with some closeups of the internals