The MCP3564 is a 24-bit sigma-delta ADC with support for up to eight single-ended, or four differential channels. It supports extensive oversampling depths, flexible MUX front-end, circuit burnout detection, adjustable gain, Vref from [0v - 3.6v], and optional internal oscillator.

The board is available with or without high-accuracy external oscillator. If included, the oscillator is a Microchip DSC1001BI1-019.6608 MEMS oscillator at a fixed frequency of 19.6608 MHz (+/-50 ppm).

To ease confusion on the breadboard, the No-OSC version is OSH Park's trademark purple, and the OSC version is the After Dark color scheme.

Each board is tested prior to shipment.

Note regarding the newer "R" variant of this part: Microchip released a new version of this silicon in 2020. Only a handful of customers have the old silicon. All orders since June 2020 have been fulfilled with the MCP3564R. The driver supports both variants.

Note regarding the analog filtering on the front-end: This breakout was designed according to the app note in Microchip's datasheet. That includes the low-pass RC filtering on the front-end, which has a high-end cutoff frequency near 16KHz. So if you are using the part to sample frequencies higher than this, you should alter the filter by stacking an extra resistor on the channels you want extended.

Driver support: Enough customers have had issues sitting in github for >1 month, that I will restate in the product listing: I will respond to github issues (eventually). But you will have more timely support by going direct to the driver support chat room which I have created for this purpose: https://hackaday.io/messages/room/298450

Notes regarding oscillator selection: It should be noted that the noise characteristics of the ADC itself begin to fall-off somewhere after 12MHz. So for applications that require tight noise figures, it might be better to consult the ADC's datasheet and possibly order the version without an OSC in expectation of feeding it your own tailored clock.

Similarly, the channels are read sequentially. Not all at once. So higher clock rates will decrease "phase-shear" between channels. If your application requires that samples be taken synchronously, I would advise not ordering the oscillator so that you can feed two boards with a single synchronized clock.