Hardware-specific questions

Solved
fletch's picture
fletch
Moderator(0)

If you have questions specific to the hardware or circuit design in the AT&T Cellular IoT Starter Kit, please post them here.

Bryan

antonio's picture
antonio
Junior(2)

Hello everybody,

 

I'm working with:

 

1) AT&T IoT Starter Kit 2nd Generation (AES-ATT-M18Q2FG-M1-G) connected to

 

2) IoT Breakout (AES-ATT-M18Q-CAR-G  LTE) with the software

 

3) reference project and API (both found at https://github.com/Avnet)

 

and done some preliminary tests to interface custom hardware.

 

I've successfully connected and triggered an Interrupt Service Routine (ISR) to the GPIO93_3V3 (pin 7 of J8 of AES-ATT-M18Q2FG-M1-G board) using software functions of the API AvnetWNCSDK (gpio_init, gpio_dir, and gpio_irq_request).

However, I failed in doing the same thing using the INT input signal (pin 15 of P2 of Breakout board).

 

Looking into the Hardware User Guide (M18Q2FG-ATT-LTE-Kit-HW-UG-v1-2.pdf) Tables 3 and 4 (pages 18 and 19), I read that the GPIO pins of the IoT processor M18Q2FG (with exceptions of a few, including GPIO93) are internally pulled up or down and I suspect that these internal resistors interfere with the operation of the 1.8/3.3V translators (FXLA108BQX) forcing them permanently in output direction.

 

The API seems to not offer a mean to disable these resistors even if the processor has the capability to enable/disable them.

 

Any idea how to workaround this issue?

 

Thank you!

Antonio

 

PeterF's picture
PeterF
Moderator(16)

Antonio

Thanks for taking the time to detail the issue you encountered with the INT signal (pin 15 of Click socket P2). What should be noted however regarding this signal (which after level-shifting is connected to GPIO07 on the WNC module) is that it is also routed to the INT2 output of the onboard accelerometer!

 

ie. To use this GPIO07 as an INT signal from Click socket P2, you must disable the INT2 output of the LIS2DW12 accelerometer, ie. either in your application software, or by removing R49 on the LTE SOM board. See comments regarding GPIO07 on pages 28 and 33 of the Hardware User Guide (M18Q2FG-ATT-LTE-Kit-HW-UG-v1-2.pdf )

 

To prevent the accelerometer's INT2 output (by default configured for push-pull operation) from interfering with GPIO07 being controlled by a different board, we suggest that you configure INT2 of the LIS2DW12 device as an open-drain output. To do this, set to '1' the PP_OD bit in the CTRL3 register of the LIS2DW12 device

antonio's picture
antonio
Junior(2)

Hi Peter,

Thank you very much for this detailed and clear explanation and make me aware that GPIO07 has contention with the onboard accelerometer.  We are using this IoT Breakout board only to test our custom board with the LTE IoT board (AES-ATT-M18Q2FG-M1-G).  Eventually, we'll plug our board directly to the LTE/IoT through its expansion connector. Thus I'm thinking to reroute the interrupt signal to the  INT_M1_3V3  instead of  INT_M2_3V3, changing the related ISR setting.

The reason I'm thinking of doing so is that looking at Table 15, page 33 of the Hardware User Guide (M18Q2FG-ATT-LTE-Kit-HW-UG-v1-2.pdf ) I read that GPIO94 is not connected to any other device but pin 15 of Click socket P1 (INT_M1_3V3) of the IoT Breakout (AES-ATT-M18Q-CAR-G  LTE). 

Doing in this way, I should be able to maintain the usability of the onboard accelerometer. Would it work?
    
Grazie!
Antonio

PeterF's picture
PeterF
Moderator(16)

Yes, use of GPIO94 is much easier. The only connection of GPIO94 is via level-shifter on the LTE Breakout Carrier to the P1 Click socket (INT_M1_3V3) 

 

 

antonio's picture
antonio
Junior(2)

Hi Peter,

I tried using GPIO94 as discussed, but does not work. 

Please note that I've successfully connected and triggered an Interrupt Service Routine (ISR) to the GPIO93_3V3 (pin 7 of J8 of AES-ATT-M18Q2FG-M1-G board) using software functions of the API AvnetWNCSDK (gpio_init, gpio_dir, and gpio_irq_request).

However, as I said, I failed in doing the same thing using the INT input signal (pin 15 of P1 of Breakout board), that is GPIO94. 

As you noted, the GPIO94 is not connected to any other device, so there is no contention there. However, it does not work and it is frustracting.

As I wrote, I noticed that GPIO94 is pulled down internally in the processor M18Q2FG (see Table 3 and 4), but GPIO93 is not pulled up nor down. I believe that this resistor forces the voltage translator in the output direction, that's why it does not sense the signal that I apply on the INT pin.

If this is correct, could you provide a software upgrade/mean to disable the pull up/down resistor of the GPIO pins?

Thank you,

Antonio

PeterF's picture
PeterF
Moderator(16)

Antonio

 

Value of the internal GPIO94 pulldown resistor is in the range of 55 kohm to 390 kohm.

 

As a quick work-around / test for this, try adding an external pullup resistor from P1 INT pin to 3V3 (ie. solder a 10 kohm resistor to rear of Click P1 socket between pin 15 and pin 7)   

  

I am working on a design using the WNC IMS2 cellular module. The specifications say that the minimum voltage is 3.3V, however all test current (Amps) ratings are given for 3.8V and 3.8V seems to be the recommended/nominal supply.

 

My question is: will the IMS2 be able to join the LTE-M network properly and transmit/receive data if the supply is 3.3V? Will it have a more limited range than at 3.8V?

 

It would be great if we didn't need to include a DC boost converter from 3.3V to 3.8V