I have just tried version 541b09gg, and the problem remains. Behaviour
is the same with and without OpenAT.
Which module do you use?
For how long do you pull the interrupt line low? Is it a pulse, or do you
keep it asserted until the Wismo responds?
I have changed the external uC to keep the interrupt low until the Wismo
responds (by doing an I2C read). This seems to work, but I will have to
wire up a logic analizer to check if the Wismo is actually responding to
the interrupt, or if it is only seeing and reporting it when it wakes up for a
different reason.
On your application, do you monitor CTS to check if the wismo is in sleep
mode? How often does it wake up (pulse CTS on/off)?
I use Q2400A and Q2406B modules with OpenAt 2.0 and OpenAt 2.10c.
I use a pulse with roundabout 30ms to low level to receive the WIND indikation.
No, I don’t monitor CTS.
Perhalps the direct coupling from the uC and the IRQ-pin could be the problem. I use an open collector transistor and no additional external pull up resistor to switch this IRQ-pin.
I hope you can find the problem, because in my applications I have no problems with this pin.
Perhalps the direct coupling from the uC and the IRQ-pin could be the >problem. I use an open collector transistor and no additional external >pull up resistor to switch this IRQ-pin.
The external uC is an AVR, and I’m using the IO pin as an open collector
driver. The output is never driven high. When not asserted it is configured
as an input without a pullup, and when asserted it is an output driving
low.
The problem was caused by the uC not asserting the interrupt line
for long enough.
When W32K=0 the interrupt latency is 3-4mS, and the Wismo will
respond to a pulse a few uS long.
When W32K=1 the interrupt latency is 7-8mS, and the interupt line
needs to be asserted for a few milliseconds.
I have changed the uC code to keep the interrupt asserted until the
Wismo responds, and so far this has worked perfectly.