SPIM Problem, SDK 15.3 S132

RE: SPIM Problem, SDK 15.3 S132

cemerick — Fri, 12 Apr 2019 04:07:46 GMT

Looks like we will resolve this by making the spi read/write be blocking functions that wait for the spi transfer to complete. Then it looks like the ti CC3100 needs another 300us delay between transfers in order to work ok.

RE: SPIM Problem, SDK 15.3 S132

cemerick — Thu, 11 Apr 2019 15:13:15 GMT

In the TI porting spi diagnostics, 4 bytes are being either written or read for each transfer. So I am still puzzled as to why there are thousands of spim events in a few seconds when the code is not doing any transfers. I will see if I can work around issues by aborting or stopping the spi function.

RE: SPIM Problem, SDK 15.3 S132

Jørgen Holmefjord — Thu, 11 Apr 2019 14:15:34 GMT

The examples in the SDK use the (nrfx) SPI(M) drivers, which only trigger STOP task in abort/uninit functions. Typically a transfer is allowed to complete, which will leave the peripheral in idle/stopped state.

RE: SPIM Problem, SDK 15.3 S132

Jørgen Holmefjord — Thu, 11 Apr 2019 14:15:33 GMT

RE: SPIM Problem, SDK 15.3 S132

cemerick — Wed, 10 Apr 2019 15:51:53 GMT

Thank you I will see if that stops the spim handler interrupts. Are there any examples that show/use STOP tasks and related events to demonstrate starting, stopping or pausing functionality ? That would help communicate what is required a bit better.

RE: SPIM Problem, SDK 15.3 S132

Jørgen Holmefjord — Wed, 10 Apr 2019 15:33:57 GMT

Both should be fine. You should not get any events/interrupts from the SPIM peripheral after you trigger the STOP task and received the STOPPED event.

RE: SPIM Problem, SDK 15.3 S132

cemerick — Tue, 09 Apr 2019 14:34:40 GMT

What is the proper way to stop the spim events ? I can't seem to find a way to stop the spim events. Do you have to uninit the spim or can you disable the spi interrupt ?

RE: SPIM Problem, SDK 15.3 S132

cemerick — Tue, 09 Apr 2019 13:58:37 GMT

The dual for sections had already been removed. The TestSpi function was not being called more than expected . There is only one write and then three reads of 4 bytes so I don't see what is triggering all of the spim events.

RE: SPIM Problem, SDK 15.3 S132

Jørgen Holmefjord — Tue, 09 Apr 2019 08:36:18 GMT

I'm not sure what you are trying to do in this section, but it does not look like a good way of doing this. You should avoid using the same variable name for both loops, and the brackets seems to be misplaced.

 for (int i = 1; i < 11; i++)
 for (int i = 1; i < 11; i++)
 {
  while (!spi_xfer_done) {__NOP;}
    {
      NRF_LOG_INFO("*****Starting Pass : %d \r\n", i);
    //  CC3100_InterruptEnable();
      irgcnt = 0;
      TestSpi();
      // spi_Close(1);
      __NOP;
      }
 }

You should also make sure to add a endless while loop at the end of the main() function, the application might not work as expected if the main loop is exited. Have you checked it TestSpi() function is called more times than expected? The SPI peripheral will give you events when it is done transferring data, regardless of the state of the connected device.

RE: SPIM Problem, SDK 15.3 S132

cemerick — Mon, 08 Apr 2019 17:24:44 GMT

The spim event handler is being called. Multiple times even after the wifi chip is reset/disabled.

RE: SPIM Problem, SDK 15.3 S132

Jørgen Holmefjord — Mon, 08 Apr 2019 12:18:30 GMT

Hi,

[quote user=""]either get an assert or a flood of interrupts[/quote]

What kind of assert or interrupts are you receiving? Is this interrupts from the WiFi chip, or are you getting interrupts from the SPI interface?

Best regards,
Jørgen