interfacing FPGA board

I've put mesa 7i90 firmware on an FPGA board i had lying around. I think i know how i'm going to do my 24v inputs and outputs, but how would you guys recommend interfacing the FPGA pins to a stepper driver or encoder? can i connect directly from the fpga to the driver, or does it need protection? I've been seeing a lot about differential outputs for stepper drivers, is that necessary?

What FPGA board and other hardware do you have?

Bare Spartan 3 and 6 I/O pins have an absolute maximum input voltage of 4V
5V signals will kill them so in general you need buffering or protection.

In addition the input protection diodes will be destroyed if you exceed about -0.8V
on the inputs (easy to do with the large amount of high current, high frequency
ground bumping in industrial equipment, just touching a FPGA pin to frame ground
will often kill it)

Differential drive is not normally needed for most step motor drives though
it does improve noise immunity. Its mainly used for step/dir servos
that may have MHz step rates

The FPGA is a no-name chinese one with a xilinx XC6SLx9,and I have dm556 stepper drivers.

this is what i'm looking at for the encoder
encoder
.

would a 74LVC245 be fast enough to shift an encoder signal down to 3.3v without missing pulses?
what should i do for the stepper drivers? TVS diode? could i just use a buffer?

So you tell, that you buy some FPGA "no-name" board that even doesn't have name (and I don't ask about it datasheet), and asking, if it will work? Am I write?

You can try it. And tell us what you will get.

I made an interface board for the DE10 nano running Machinekit (mksocfpga) which also ports hm2 firmware. I don't know what kind of I/O buffering the DE10-nano might have on board but the pins are listed as 3.3v tolerant. I just used differential transmitters (SN75174) for the stepgens and differential receivers (AM26LS32) for the encoders. The stepgens are straight forward, but for the encoders the AM26LS32 datasheet specifies a minimum of 4.75v on VCC which dictates the chips output voltage. I used 3.3v to VCC so the output to the FPGA is 3.3v and have not had any issues. Testing it I've spun small steppers with cheapo drives to 3000 RPMs and read back a 1000ppr encoder in quadrature with no detectable lost counts/steps. I don't have any further isolation or buffering on the board, didn't get around to figuring that out but I've never killed anything on it. I drag this board with me to work because with the test GUI I made it makes it super easy to just test steppers, encoders, I/O whatever. That's a bunch of dirty 90's era printing presses with half the wiring yanked out and noisy electronics all over the place.

Test
drive.google.com/open?id=16uMomvdHAVa41lQ45JzOAsTudEWS06MG

Link to the project
github.com/ShadeTechnik/socfpga-developement-OSHW

Looks like very nice job. But you should test your board in noisy environment to be sure. For example run asynchronouse motor with freaquency drive near your board.