Okay, this is a royal mess!!! I'm trying to make D-DAQ as robust as possible. Since I'm having 3 SPI signals go off-board, I know I need to terminate them. However, line termination is a royal pain to read about, especially when you're on a time crunch. I'll make this log short, too btw.

So, there are 4 basic types of terminations: Series/source, parallel, Thevenin, and AC. So, technically, it is possible for me to go back into the main board design, move traces around and series terminate each of my lines. However, if you've seen the board or it's graphical layout, it's hard to find space for 6 more resistors...

This leads to parallel termination which is basically a pull-up or a pull-down resistor at the end of the line. The issue here is that there is always current flow through the resistor whether it is sourced or sunk from the I/O pin. I'm dealing with a limited-to-no airflow environment and more power means more heat. Scratch that one.

When I read the word "best" I try to take note as to why the word is used. This was the case in the article on stack exchange for AC terminations. However, upon (PDF) further (website) reading (TI PDF) it's apparent that AC termination is tricky. Based on what I'm reading, I would not use this for long distance signaling and I'd have to set it up the capacitor for the frequency I'm operating at. D-DAQ is being designed to run the SPI bus at a specific speed but this doesn't mean that it will always be used at this speed. Capacitive reactance loading limits 1 of 2 things: bus speed, bus length. I can taylor the capacitor for a 20 MHz top end signal, but I'm limited to a 3 ft (4 ft really) cable. I'm sure someone will need a 6 footer at some point.

That leaves Thevenin. Basically it's a pull-up and a pull-down resistor pair. There is constant current, but due to the parallel nature of the resistors in line with the transmission line, the current draw is much less than parallel termination. As there is no capacitor in the circuit (ignoring transmission line capacitance), I don't have to worry about a limit to cable length or operating frequency.

If my understanding is accurate, I'll have an additional average current draw of ~9.2 mA per line on the 3.3V rail. For all 6 lines, that puts me at ~55 mA. Ironically, this is the current draw for 1 parallel termination.