Since you are only shifting 16 bits, I’ve to doublecheck: you’ve only connected 2 74HC165 SRs and 2 74HC595 SRs together, right? Or do you have more than 2 SRs per module?
How exactly do you want to send the clock and latch/load signals?
And why are you not using a common core module, running with MIDIO128 or similar, for the test?
This has the advantage, that you would test the whole system.
These are the MBHP_DINx4 and MBHP_DOUTx4, so yes, they have 16 bits each one, and will test first a OUT module and 16 LEDs, then a IN plugged on that OUT, and check the output of the IN.
Why I cannot test with the core? Please take a bucket of pop-corn : because I destroyed the optocoupler yesterday (yikes, I punched it and broke 1 pin, it was very late, and I was nervous…), but I ordered one other. So, the MBHP_CORE seems working (at least it doesn’t burn, and voltages are OK at the control points), but with the bare bootloader.
Meanwhile, I will solder the IN and OUT modules and test them with an Arduino.