Håkon Alseth was helping me with this: https://devzone.nordicsemi.com/f/nordic-q-a/83126/hard-fault-using-twi_scanner-example
I accidentally clicked "verify answer" when trying to reply, and I can't open it back up again.
Allowing the topic author to undo a "verify answer" would seem to be fundamentally important. Please consider adding this feature to your help board.
To continue that thread:
scath said: "However, I now detect a valid TWI device at 0x00 -"
Hakon said: "you should start scanning from address=1."
Fair, but I now detect a device at 0x01.
scath said:
Example:
Device 1) SCL:P0.21, SDA:P0.20 - in this instance I expect an address between 0x2a or 0x2b
Device 2) SCL:P0.26, SDA:P0.27 - in this instance I expect one of six addresses, none of them are 0x00
Hakon said: "Have you scoped the lines to see what happens?"
I don't have a scope. However, checking each line with a multimeter see's the voltages go up and down appropriately on all SCL and SDA lines: the SCL gets pulled high to ~3V, the SDA line finds a value of roughly ~1.3V-1.7V.
Hakon said: "Did you remove the pca10056 system files and add system_nrf52833 and nrf52833-startup files?"
No. I'm using the pca10040 example - the pca10056 project causes a hard fault as per my last ticket. I have not altered the build otherwise for fear this would break what little progress I've made.
Can you tell me precisely which files to change, and where? Preferably without referring me to a ticket with an analogous issue, since "fixes" for other boards often don't work with the 833.
Thanks so much.
S.
