Earthquaker Devices Rainbow Machine dissection

[brokenstarguitar]

Hey guys, how's it been? So, I've had a request to clone a EQD Rainbow Machine, thinking it shouldn't be too hard to reverse. Well, seems I'm wrong. I went out and got one, removed the board from the enclosure and while inspecting I noticed the main chip on the board has "Zido" written in silver marker. After some research and probing I found the chip to be a "Spin Semiconductors FV-1 DSP Reverb" chip. Now I'm thinking the guys at EQD might have just used the on board pitch shift program being that's the main effect of the Rainbow Machine, but I could be wrong.

Now I have a bunch of pictures I'll upload if anyone wants to help dissect this thing and built a working clone. Anyone on board?

[slacker]

The Rainbow machine does two pitch shifts and none of the built in programs do that so I'm pretty certain they're using their own program. You can tell for sure by seeing if the board has a 24LC32 chip or similar on it, if it has then that means it's using their own program stored in that chip.
You could make a similar effect using the built in program number 4 which has a pitch shift on one channel and a delay on the other by chaining the two channels in series and using analogue feedback, making a delay with pitch shifter in the feedback loop which is basically what the rainbow machine is. Or you can write your own program to do the same thing like I did here, skip to about 3:00, either way I doubt you could build a clone any cheaper than you could buy a real one.

[brokenstarguitar]

The Rainbow machine does two pitch shifts and none of the built in programs do that so I'm pretty certain they're using their own program. You can tell for sure by seeing if the board has a 24LC32 chip or similar on it, if it has then that means it's using their own program stored in that chip.
You could make a similar effect using the built in program number 4 which has a pitch shift on one channel and a delay on the other by chaining the two channels in series and using analogue feedback, making a delay with pitch shifter in the feedback loop which is basically what the rainbow machine is. Or you can write your own program to do the same thing like I did here, skip to about 3:00, either way I doubt you could build a clone any cheaper than you could buy a real one.

Slacker, there's only 4 chips (the FV-1, 2-tl074s, and a cd4049) and 1 3.3v regulator. That's the reason I thought they used the on board program. I'll post a few pictures in a few minutes when I get to my computer.

[brokenstarguitar]

[slacker]

Can you figure out or measure what pins 16,17 and 18 of the FV-1 are connected to, for each pin it will be either 0 or 3.3 Volts. That will tell us which program it's using. It could be that they have had the chip programmed with their own programs rather than the factory ones, in which case you're out of luck.

[cloudscapes]

If you order enough of them (thousands) you can have your own custom program put on it, no EEPROM needed.

[bean]

There's no crystal and the 4049/tracking pot seems to be interacting with pins 9 and 10 on the FV-1. That probably has some significance.

[slacker]

Oh yeah, I hadn't noticed that I did wonder what the CD4049 was doing but didn't look closely enough. They're using it clock the FV-1, probably with tracking pot controlling the clock speed, that sort of fits in with what they say the tracking control does.

[Ice-9]

Check pin 13 of the FV-1, if it is connected to 3.3v (direct or through a resistor) then it is using the FV-1 internal programs, if it connects to gnd (as above) then it has an external eeprom somewhere.

Yeah, 4049 is for clocking the Chip and as Ian said probably along with the track pot to control the clock.

[ichbindaz]

There's a good discussion on the Rainbow Machine over at FSB. Prevailing view seemed to be that there was no external EEPROM, and that there was a feedback from output to input to create the magic. Some say it's the standard FV-1 pitch shift program, others think there's more going on.

[bean]

I totally agree- just form the pic here I think the Magic pot as feedback is a very good guess.

[slacker]

Some say it's the standard FV-1 pitch shift program, others think there's more going on.

The pedal does two pitch shifts, none of the internal programs do that, so I don't see how they're using an internal program.
My guess is it's a custom program that does the two pitch shifts, one of the FV1 control pots does the Pitch -4th to +3rd, another control pot is the volume for that, and the third control pot is the volume/pitch for the +- 1 octave. Tracking controls the speed of the FV-1 slowing it down means the delay inherent in the way the pitch shift works becomes long enough to sound like a short delay. Magic is analogue feedback around the FV-1 and tone is tone control somewhere either after or in the feedback loop.

[brokenstarguitar]

Can you figure out or measure what pins 16,17 and 18 of the FV-1 are connected to, for each pin it will be either 0 or 3.3 Volts. That will tell us which program it's using. It could be that they have had the chip programmed with their own programs rather than the factory ones, in which case you're out of luck.

I haven't been able to check voltages yet but I have figured out a few things.

Check pin 13 of the FV-1, if it is connected to 3.3v (direct or through a resistor) then it is using the FV-1 internal programs, if it connects to gnd (as above) then it has an external eeprom somewhere.

Pin 13 of FV-1 to 22k then 3.3v regulator

Tracking 1&2 to 4049 pin 13&14

Pitch 1 to Grd
Pitch 2 to pin 20 on the FV-1
Pitch 3 to output of the 3.3v regulator

Secondary Lug 1 to pins 18 & 19 of FV-1
Secondary Lug 2 to Negative of 1uf electro (haven't figured out where the eclectro goes yet)
Secondary Lug 3 to 1k then 5k then 1uf then pin 24 & 25 of FV-1

Magic lug 1 to pin 1 of 074 and Cathode of  one 4148 and Anode of the other
Magic lug 2 to other Cathode and Anode of 4148s.

I'm still trying to trace the other pots with continuity and visual inspection. What do you guys think? Onboard program? Special program?

[slacker]

Pin 13 of FV-1 to 22k then 3.3v regulator

So, like Mick said it's using an internal program, probably a custom one though not a factory one. Voltages/connections for pins 16,17 and 18 will tell use which program and probably confirm it's not a factory one, see below.

Pitch 1 to Grd
Pitch 2 to pin 20 on the FV-1
Pitch 3 to output of the 3.3v regulator

Pin 20 is control pot0 input, which does control pitch in two of the factory programs.

Secondary Lug 1 to pins 18 & 19 of FV-1
Secondary Lug 2 to Negative of 1uf electro (haven't figured out where the eclectro goes yet)
Secondary Lug 3 to 1k then 5k then 1uf then pin 24 & 25 of FV-1

pins 19, 24 and 25 are all ground so lug 3 must go somewhere else as well. Pin 18 is one of the program selection pins, if it's connected to ground then the FV-1 can only be using programs 0 - 3, program 3 is the only one that does a pitch shift. Pins 21 and 22 are the other control pot inputs, see if any of the other pots are connected to them, if they are then it's definitely not using the factory built in programs because program 3 doesn't use those two control pots.

[Ice-9]

Program 3 of the internal programs has a pitch transpose on 1 output and a delay on the other output, looking at the pcb both in/outs are used so it may be possible that the internal program is the standard spinFV-1 programs.

It would be useful to work out which program is being used so if you can tell us what the S0-S2 pins are tied to, eg gnd or 3.3v then at least we know what program number is being selected.

I haven't heard what the rainbow machine really does with the pitch effects but I seriously doubt that it would be viable for a custom program to be programmed to the FV-1 unless you were ordering something like 10,000 units, maybe with modern day fabrication it is possible for small runs to be affordable, I just do not know.

Simple way to find out, remove the chip FV-1 and stick in another known production chip.

[brokenstarguitar]

Hey guys, sorry for the delay. Got a lot going on. I did manage to get the pn voltages though. They are as follows......

1. 1.68v
2. 1.68v
3. 1.68v
4. 0
5. 3.37v
6. 3.37v
7. 0
8. 3.37v
9. 1.7v
10. .94v
11. 0
12. 0
13. 0
14. 3.37v
15. 0
16. 0
17. 0
18. 0
19. 0
20. 3.6v
21. .71v
22. .71v
23. 3.37v
24. 0
25. 0
26. 3.35v
27. 1.72v
28. 1.72

[Ice-9]

With pin 16-17-18 at 0v that would select the Rev/chorus program from internal memory, so I must concede that the FV-1 must have a custom effects set directly in the FV-1 core.  It would be interesting to modify the pedal to select the other 7 effects locations. You may find something very exciting in there.

After all, if they have gone to the expense of having a custom effects set burned to the chip it would make sense to have more than just 1 custom effect in the core.

[slacker]

It would be interesting to modify the pedal to select the other 7 effects locations. You may find something very exciting in there.

I was thinking exactly the same.