One-Way SPI to multiple devices on long ribbon cable

[T3sl4co1l]

How are you wiring between each buffer and display?

Tim

[nctnico]

Got some new boards in today which incorporate source termination resistors and a SN74HC367N buffer chip between each display...  Was able to reliably drive 5 (I don't have the rest wired up to test yet) at 30MHz, but if I up the frequency any more the screens at the end of the chain start having issues (as opposed to the ones at the beginning).  If I lessen the amount of screens I can increase the Hz more.  So there's still something that's getting lost over the distance despite having repeaters every foot or so

I'd measure the signals using an oscilloscope. See if the timing is still OK.

[zacaj]

How are you wiring between each buffer and display?

I've hand-crimped some small wiring harnesses (each display's mounting position is unique)

[gnuarm]

Got some new boards in today which incorporate source termination resistors and a SN74HC367N buffer chip between each display...  Was able to reliably drive 5 (I don't have the rest wired up to test yet) at 30MHz, but if I up the frequency any more the screens at the end of the chain start having issues (as opposed to the ones at the beginning).  If I lessen the amount of screens I can increase the Hz more.  So there's still something that's getting lost over the distance despite having repeaters every foot or so

I don't know that so many buffers are a good thing.  The signal needs to be restored.  Buffer chips can have an asymmetrical drive that changes the duty cycle of the clock.  If your pulses are 15 ns wide and you lose a ns on each one they don't need to be buffered very many times before the pulse gets too thin to work well.