QSPI CS problem

[Ran_Talbott]

Last year, I built a system around an SB72 that (among other things) uses a TPIC2603 (basically an SPI shift-register with high-power outputs) to drive solid state relays to control some equipment.

Now I'm trying to duplicate it, and running into problems: the outputs of the TPIC2603 don't match the data I'm sending. They start out all off at reset, but they all go on, and stay on, the first time I send data over the QSPI.

I'm using exactly the same software as last year, and I think the hardware is identical (it's just one chip hand-wired on a protoboard, and I've verified the small number of connections to the SB72).

I've confirmed with my low-bandwidth analog scope that correct-looking data is going out to the TPIC2603 (though its limitations mean I can only be about 95% certain).

One thing I've noticed is that the active-low Chip Select is going low between QSPI operations. I never checked that on the old hardware, so I'm not sure whether it's doing that, too (the old system is still in service, so it's a hassle to check). It's not clear from the MCF5272 User's Manual how to force it to stay high except when it's active. It's definitely being actively driven low at the moment: putting a pullup on the line didn't help.

Any suggestions?

Thanks,

Ran

[Ran_Talbott]

I went back and checked my connections one more time, and found a solder bridge that was screwing up the chip select signal.

Sorry if anyone wasted any time trying to figure this out.

Ran