1 of 1 people found this helpful
Thanks for pointing out this.
There is no errata/fixed version currently. You are right, the table is nothing with the Period Register, it is just to describe the PWM interrupt logic config(which is part of Register(29)), it should be moved under the "Config (29h)" section. The Period Register(2Ah) does not need a table to explain anything.
The latest version datasheet is available at here, but the table is not changed. We will fix this in the later version of the datasheet.
Thanks for the info and link.
I am working on to build a library for our system. GPIO is working in production, PWM is the next task.
Anyway, one question, could you imagine a situation, when I read back a port (as input) and I got back an invalid (0xFF) value and I have to reread again to get the real value?
I suspect it is not a software issue, but some electrical bug in the hardware. It is not so frequent (relative, as I read every 100msec and I have time to reread in case of problem) and I can handle it with a workaround, no other problem found in the last 1 year (we use 35 pieces of the 40 port version).
An other question, could you recommend an alternative product instead of this one? As I read, it is obsolete now.
- 400kbit - 1Mbit could be better
- POR, watchdog, user programmable settings (using for safe output on system down)