How to interface external TWI sensor (MAX30105) with Nordic Thingy 52 application source?

[Kartik]

I am trying to implement the MAX30105 pulse oximeter sensor along with the Nordic Thingy 52 application source.

I referred the sample ADXL375 code provided by Simon here https://devzone.nordicsemi.com/members/simoniversen and have integrated source for max30105 accordingly.

Now as I debug the code, it goes into infinite loop while executing the nrf_drv_twi_init() function invoked by twi_init().

Can you please help me understand the issue here?

I see that the handler function is missing; is it a mandatory function to have?

[awneil]

Kartik said:

it goes into infinite loop while executing the nrf_drv_twi_init() function

Do you mean it's going to the Error Handler ?

Jørgen Holmefjord said:

Have you called nrf_drv_twi_init/twi_init elsewhere in the code?

Kartik said:

Yes, the source code that I am using is invoking the twi_init()

That wasn't actually the question!

Jørgen Holmefjord asked if you are also nrf_drv_twi_init() elsewhere in the code.

If you call nrf_drv_twi_init() again when it has already been called before, you will probably get an error - and end up in the Error Handler.

This should be easy to see in the debugger ...

[Kartik]

No, I am not calling nrf_drv_twi_init()  elsewhere again. 

On line #375 while invoking nrf_drv_twi_init(), NULL is passed as the 3rd argument of event handler. 

From documentation I get that NULL in place of event handler means working in blocking mode. 

I see address 0x000000 in place of event handler in the watch window watching m_twi.

So is this the reason execution end up in infinite loop since event handler is pointing to NULL?

[awneil]

It's still unclear where, exactly, this "infinite loop" loop is.

Again: is it going to the Error Handler ?

If it is, you should be able to see the error code which causes that ...

[Kartik]

Ok awneil, I will check this and add my observations, currently I am not in my lab.

But I guess it is going to the error handler. When I step into the nrf_drv_twi_init() code condition 

if (p_cb->state != NRF_DRV_STATE_UNINITIALIZED) fails and thens when I step into the execution goes in some assembly code and does not seem to come out of that.

[awneil]

Around line 375 you have:

    err_code = nrf_drv_twi_init(&m_twi, &twi_config, NULL, NULL);
    APP_ERROR_CHECK(err_code);

The APP_ERROR_CHECK macro checks the err_code, and will go to the Error Handler if it is not  NRF_SUCCESS

So you need to step over the nrf_drv_twi_init(),  and examine the value of err_code that it has returned.

Then look that up to see what it means.

https://infocenter.nordicsemi.com/topic/com.nordic.infocenter.sdk5.v15.3.0/group__app__error.html

[Kartik]

Source code execution runs till different locations each time.

At times I have observed error "twi_manager   : ERROR: twi_manager_request: collision 1"

And then I have to restart the debugging back from main().

[Jørgen Holmefjord]

You will get this error if there is a TWI transfer ongoing on this TWI instance. Avoid passing the return code to APP_ERROR_CHECK and retry the transfer later.