Improved SEEED RP2040 Buffer Board--Mutable Instruments Homage

Hi there, continuing to work on the AD9833 volt/VCO, it's raining endlessly here, I am in a crappy mood, and man I am getting tired of this damn VCO build. 

Ah yes. The RP2040-AD9833 audio VCO. Is it done yet? can we pretend it is?  Update 3-6-23: It's done and came out rel good! Thanks to these guys for support; more about the finished + working AD9833 based V/octave VCO here.

For the ADC and analog buffer portion of the VCO I lifted design basics from Mutable Instruments, who make fantastic open source synth modules:

• MCP6004 low cost quad rail to rail op amp for analog input buffering and clamping
• Unipolar operation for the op amp
• Inverted input from buffer to the MCU (why not--saves parts--fix this in software--gotta love Mutable!). Everycircuit simulation of the idea is here.
• Zeners for most of the voltage regulation--cheap and simple
• a fifth 12-bit inexpensive ADC in addition to 4 provided by the RP2040 MCU (I'm not counting the on board ADC used to measure MCU temperature which isn't needed for the VCO)
• other general improvement and fixes to my initial design stupidity 

I did a post for version 1 of this same idea (here) but it sucked, this version irons out a lot of kinks.

Get this buffer board--a dev board for a dev board?--from my sponsor's PCBWAY's project page, here.  You can also get gerbers for a lot of other AudioDIWHY projects from their community pages.

You will need the most excellent and very tiny SEEED XAIO RP2040 board to make it go.

 

Some improvements to the previous version of this buffer board:

• Jumpers to enable digital or analog (JD0, JD1, JD2 on the schematic) for the the D0,1,2 MCU pins.  
• Ground pour for bottom PCB copper (imagine that)
• a 10K timmer and 96.7K resistor in series, along with a 33K resistor for the opamp's negative feedback, to get us very close to the necessary 3:1 ratio for incoming voltage to the dedicated ADC IC

Using the jumpers:

• If you want to use a given op amp buffer stage, connect its two corresponding jumper pins.
• If a jumper is disconnected: to prevent the unused op amp stage's output permanently slammed to its V+ rail, put a modest voltage (say 1/2V) at its non-inverting input.

Let's build it:

Happiness is a white box o' boards from this blog's trusty sponsor, PCBWAY:

Thanks as always to PCBWAY for sponsoring this blog....please help this blog and check them out.

Running the default blink sketch the whole thing worked first time and drew 40mA for the positive rail. The current draw from the older version was more like 150mA. Wow!! There must have been a stupid mistake I missed on the last iteration of this board, making it draw too much current.

Se habla blink? Claro que si.

We are ready to use this board for other projects!

More more MORE pcbs for the AD9933 VCO. Again: Are we done, yet?

Using the board should be straightforward.  If you are using ADC 0-1-2 (on the RP2040) you need to jumber JD0, 1, or 2 accordingly.  Otherwise, to use a digital in instead of buffered analog, leave JP0, 1 or 2 off. Also, if you are not using a given op amp stage you should leave C6/R10 (ADC0);  C7/R11 (ADC1); and/or C8/R12 (ADC2) off the PCB.  In this configuration, the unused buffer will get the -10V input signal which will keep the unused stage from flopping around at its output.

Onward....

Being a glutton for punishment I laid out pretty much the same board with a 14 bit ADC. Will it work more and better? We will see in the next few posts...but hopefully this damn VCO this board is going into will be done before the next global extinction event.  Update: With the VCO working the 12 bit MCP3201 ADC cuts the mustard. This was unexpected? So....for now the 14 bit version remains  unpopulated, but it will see life--assuming it works--in a future project.

Until then: Don't breathe the fumes--always bite off more than you can chew--stay in vegetables--eat your school.