Which digital controlled Potmeter to choose for Arduino controlled Analog dev?

[The Red Rooster]

Hi All,

I am thinking on getting my feet wet by experimenting with Digital controlled Potmeters in analog tonestack and/or gainstages.

My intention is to create a setup letting an Arduino control several Digital controlled Potmeters.

Do you have any experience with such a setup?

Which control method is best (less noisy in audio chain) SPI or I2C?

I have been looking at:

https://docs.rs-online.com/51da/0900766b81380ca2.pdf

Which seems to be an ok when looking at both price and control steps.

Kind Regards

Kurt

[ElectricDruid]

I've used the Microchip MCP42xxx chips for the sort of thing you describe. In fact, I built a programmable overdrive with them. To keep the rest of the signals in the same range as the pots, I also used the MCP6002 5V op-amps. I never got very far with the project. It worked, or at least *I think* it did (behaved mostly as expected) but I found it all rather underwhelming, so I lost interest and did something more exciting instead.

As far as the pots go, they do what they say, they're relatively easy to get running, and the 256-steps is good. 5V headroom is something that takes some getting used to, and I think I probably need to learn some better techniques to get the best out of it.

I haven't compared I2C pots with SPI pots, so I can't answer your question. In general, you want to make sure that fast digital signals are kept apart from the audio, and you probably want the processor that's talking to all this stuff on a different 5V supply to the pots and the rest of the audio path. *Don't* try and use a single 5V supply! That's just *asking* for trouble. My hat is off to you if you can pull it off and keep it quiet, but I'd be very cautious about that for starters. At the very least, the digital side and the analog side need to fed separately and only meet back at the 5V input, probably with extra filtering on the analog side voltage.

Good luck and keep us posted!

[Sweetalk]

I'm using the SPI MCP4251 with really good results, I cannot hear any communication noises because the SPI works at 400KHz so it's way out of the sound specter. The 256 steps are good as ElectroDruid says.

As for SPI vs I2C: in SPI you need MOSI, SCK and one Chip Select pin per digital pot you use. In I2C you need SDA and SCL, the selection of the device to be controlled is made by the address set in that device (can't remember if the I2C MCP have external pins for addressing).

[ElectricDruid]

I cannot hear any communication noises because the SPI works at 400KHz so it's way out of the sound specter.

...and at that, the SPI is not being pushed at all (In fact, even the serial UART on a PIC will go that fast). Most SPI devices will do several MHz comfortably. Whether that reduces problems or increases them probably depends on what you're worrying about!

It's a good point though - comms noise is going to be very high frequency and should largely be amenable to being filtered out. I've had issues where I was doing DAC updates at some regular interval (say every 1 msec) because then you get a burst of high frequency noise, but *that noise appears at a 1KHz frequency* so althoguh the noise itself might not be too bad, the 1KHz update rate certainly is!