Starterware/TDA3: Input interrupt test application for GPIO on Starterware doesn't work properly

Jho Zoo20

Part Number: TDA3

Tool/software: Starterware

I have a problem with Input interrupt test application for GPIO.

It's based on interrupts and I need to provoke hardware interrupt (to connect GPIO pin that I'm testing with a ground) to get corresponding output in the console. My GPIO pin is on J28 connector and I tried to connect it to the ground directly - using wire, after I loaded app on M4 core using Code Composer Studio. It doesn't work and my Code Composer Studio crashes down with 'power loss' message in the same moment when I connect those two dots on the board with a wire.

I think that I am not supposed to that from hardware point of view, so my question is: does anyone know any smarter way of testing this app on TDA3x board?

Best regards,

Jho.

over 8 years ago

0 Mariya Petkova over 8 years ago

TI__Expert 4885 points

Hi, Jho Zoo20,

Starterware experts are notified about your query and will comment directly here.

Regards,
Mariya

0 Rishabh Garg over 8 years ago

TI__Guru 55685 points

Hi Jho Zoo20 ,

GPIO interrupt application doesn't require you to connect any wire. You can refer to StarterWare user guide on steps to run this example application.

Regards,
Rishabh

0 Jho Zoo20 over 8 years ago in reply to Rishabh Garg

Prodigy 200 points

Dear Rishabh,

The problem is that I need to use test application not for the default pins but for the wanted ones.
I have changed pins in app code and some parameters in corresponding functions, but when I choose any of four options in the console I don't get any response.

Regards,
Jho.

0 Prasad Konnur over 8 years ago in reply to Jho Zoo20

TI__Expert 6065 points

Hi Jho,

The GPIO input interrupt test configures one of the pad from soc as gpio input pin.

interrupt is registered for one of the 4 options high/low level voltage, rising/falling edge.

So based on the option you should provide the voltage/edge on the pin for the interrupt to be generated.

Make sure you are connecting only the required pin to ground, and not shorting the adjacent pins. This example was added just for demo.

If you are doing for some other pin make sure the pad configuration is done properly before connecting pin to ground.

Regards,
Prasad

0 Jho Zoo20 over 8 years ago in reply to Prasad Konnur

Prodigy 200 points

Thank you Prasad, for your help.

I am pretty sure that my pad configuration is done properly. I have just one doubt. When I change pin in the code I should change interrupt number too. I am working on GPIO3 and there is not mapping for it's interrupt on IPU, only DSP. Since default example was for GPIO4 and they used interrupt number 51 (which is mapped to GPIO1 based on user guide) I used it too. Might that be a problem?

Regards,
Jho.

0 Stanislav Stilyanov over 8 years ago in reply to Jho Zoo20

TI__Genius 17740 points

Hi Jho,

Can you please share which J28 pin you are short-circuiting to gnd?

Thanks,

Stan

0 Rishabh Garg over 8 years ago in reply to Jho Zoo20

TI__Guru 55685 points

Hi Jho,

Yes you need to change the interrupt configuration. You can map the GPIO3 interrupt via the IRQ crossbar on any core in the SoC (particular interrupt line should be connected to crossbar).

You can use the same interrupt line i.e. 51.
For this particular example you should modify IRQ_XBAR_INDEX for GPIO3.

Regards,
Rishabh

0 Jho Zoo20 over 8 years ago in reply to Rishabh Garg

Prodigy 200 points

Hi Rishabh,

I switched to the IRQ XBAR mapping (instead of default one) and succeeded to run application properly. Thank you very much.

Regards,
Jho.

0 Rishabh Garg over 8 years ago in reply to Jho Zoo20

TI__Guru 55685 points

Hi Jho,

Glad to hear that the issue is resolved. Thanks.

Regards,
Rishabh

0 Stanislav Stilyanov over 8 years ago in reply to Rishabh Garg

TI__Genius 17740 points

I'm still confused why change in interrupt would solve the 'Power fail' condition.

0 Jho Zoo20 over 8 years ago in reply to Stanislav Stilyanov

Prodigy 200 points

Hi Stanislav,

Power fail condition isn't solved, I still get that output in CCS, the difference is that now my interrupt is detected before that. So, my software problem is solved, but hardware isn't.
The pin that I'm connecting to the ground is MDIO_MCLK, next to the R9055 resistor.

Regards,
Jho.

0 Rishabh Garg over 8 years ago in reply to Stanislav Stilyanov

TI__Guru 55685 points

Hi Stan,

I am also not sure about the power fail part.
I saw it as an issue with the software where Jho wanted to use different GPIO pin as compared to the one used in starterware application.

Regards,
Rishabh

0 Stanislav Stilyanov over 8 years ago in reply to Rishabh Garg

TI__Genius 17740 points

Hi Rishabh and Jho and thanks for your replies.

Jho,
I cannot find anything disturbing to short circuit the MDIO_MCLK to gnd, providing it is properly configured as input GPIO.
I can suggest you a couple of thing but they are all not related to power fail.
1. Turn on the internal pullup resistor
2. More appropriate would be to mount an external pullup of few kohms.
3. Turn on the GPIO debounce period if already not done so.

One thing that can help debug however would be:
Perform '1' to '0' switch not by using a wire, but switching the internal pullup to pulldown in software or in command prompt. This should isolate hardware and work as expected (or not :) ).

Regards,
Stan

Processors

Processors forum

Starterware/TDA3: Input interrupt test application for GPIO on Starterware doesn't work properly