Best practice for encoding and decoding DUT/UUT registers? Class? Cluster? Strings? Bitbanging?

[CaseyH]

I am architectecting a LabVIEW system to charactarize silicon devices.  I am trying to decide the best way to enqueue, encode, and decode device commands executed by my test system.

For example, a ADC or DAC device might come in both I2C and SPI flavors (same part, different interface) and have a large register map which can be represented as register names or the actual binrary value of it's address. 

I would like my data structure to

*) be protocol agnostic

*) have the flexibility to program using either the memonics or hard coded addresses

*) be agnostic to the hardware which executes the command. (

*) I would like to enqueue mulitple commands in a row.

I am thinking a detailed class is my best bet, but are there are examples or best practices already established?

[Yamaeda]

I agree on the detailed class inherited from a general DUT-class. Especially if you want to mix interfaces you need to keep those as far away from your top vi as possible. 🙂

As to the 4th point i'd implement command-vi's as enque and have a in-class command-queue (or possibly just an array of commands).

/Y