Microcontroller SPI communications with Rutex R2020

[minerva]

I'm currently using a Rutex R2020 to control a motor. I've got everything wired up and running, and I've plugged the R2020 into a PC's parallel port and I can test everything out using the Rutex Windows test/setup software, and it basically all just works.

However, I'm working in a little bit of a specialized application - not a CNC machine controlled from a personal computer. I want to control the Rutex R2020 board from an AVR microcontroller (actually an Arduino).

I could use either SPI or direction/step commands to communicate with the R2020 from the microcontroller - but I think SPI is probably a more powerful means of control, which would allow me to control things more accurately without doing really fiddly sub-millisecond timing for the step pulses, and it will let me do things like read off the motor current from the SPI registers.

I've got the basic list of SPI register functions and their addresses, as given in the R20xx datasheet, which is a good basic place to start, but it's not quite yet everything that I need.

What I need to be able to do is spin the motor quickly, at a given speed, but stop after a certain amount of movement. The Windows software basically lets me do that, eg. run for 20,000 steps and then stop after 20,000 steps of motion. But I want to replicate that functionality on the microcontroller, without a PC.

The SPI motion command is simply a 16-bit number of steps per each particular servo loop cycle. OK, that basically makes sense. My motor encoder is 50 steps-per-revolution, so I'll just set the servo loop time to 500 microseconds (let's just assume 1X SPI interpolation at this stage), and I'll set the velocity word to 2. That will give me 4000 steps per second, or 80 revolutions per second, which is the sort of speed that I need.

What SPI communications mode do I need to talk to the R2020 board? Eg. mode 0, mode 1, mode 2 or 3, etc?

I need to be able to start at a given speed, let's say 2 steps per servo loop, and I want to continue spinning for a given fixed amount of rotation. So, I want to, for example, spin until exactly 625 encoder steps (12.5 motor revolutions) have passed and then stop.

Eg. suppose I want to set the servo loop time to 500 microseconds, (1 x SPI interpolation), then spin at a speed of 2 steps per loop, and spin until after 625 encoder steps have passed and then stop.

So, can I just set the velocity word, wait until a certain number of encoder steps have passed (Can I read out the encoder position from one of the SPI registers?) and then quickly just set the velocity word to 0?

Does that sound sensible? Will that work, or is there a more efficient way to do it?

I can just connect up the R2020's data lines (SPI data in, SPI data out and SPI clock) to the microcontroller's standard SPI interface pins (MOSI, MISO and SCK), instead of having them connected to the PC parallel port, of course, but normally, for SPI communications between devices, we also need the chip-select (CS) pin, to allow multiple different devices to talk to the SPI bus.

Which pin, if any, do I use on the R2020 for chip-select on the SPI interface?

The Mode select (SPI/dir-step) pin on the R2020 is connected to the PC parallel port when using the Rutex-specified parallel port wiring schematic. Will this pin work for chip-select? Will driving the mode-select pin high de-assert the R2020 board on the SPI bus, allowing other devices to take their turn for communications with the microcontroller on the SPI bus?

What other information should I review, so that I can write the microcontroller's SPI code to talk to the R2020?

The R2020 datasheet says "For programming details consult the STEPEX2 protocol documentation available from Rutex." What is this documentation, and where can I get it?

Any other information or support or advice anyone else would care to add?

Regards.