Skip to main content
brianshankland9
Associate III
November 14, 2017
Question

STM32CubeExpansion_WIFI SPI actually works?

  • November 14, 2017
  • 1 reply
  • 548 views
Posted on November 14, 2017 at 14:52

I've seen a few posts on issues with the SPI interface to the SPWF04 WiFi module, and from the past weeks experience have yet to get any of the examples supplied with the middle ware to actually work with SPI.

Eventually I want to use it with a STM32F7 processor but as there were no examples I bought a SRM32L476 Nucleo board so I could use the examples straight from the supplier unchanged.

Using the UART I got the Client Socket example to work quite quickly, but the SPI just hangs and the data I see at the PC is missing the first block that was supposedly written.

looking through the middleware SPI code there are many commented out 'debug' sections which were obviously used during development and a few worrying comments to put in random delays to get things to work.

So as you don't often see messages from users who got it working I'm trying to see if there are a great majority out there or is everyone using the slower UART interface method?

Is the SPI version of the middle ware actually commercial grade quality or should I write my own?

    This topic has been closed for replies.

    1 reply

    Gerardo GALLUCCI
    ST Employee
    November 15, 2017
    Posted on November 15, 2017 at 07:20

    From SPWF04side, some issues have been solved into FW1.1.0, and this will be available on st.com within a few days. From X-Cube side, a new release (3.1.0) will be released too, still supporting F401, F411, and L476 (no plans to support F7 Discovery, since it's not a Nucleo board).

    Please be sure to have set jumpers on X-NUCLEO-IDW04A1 in the right position0690X0000060PEEQA2.png

    brianshankland9
    Associate III
    November 15, 2017
    Posted on November 15, 2017 at 10:17

    Many thanks for the heads up, are the release notes already available as I have a reasonable idea that the topics that are causing the issues are realted to not correctly linking up replies to the socket write commands since I can see from additional printf debugs that I’m being returned from the SPI socket write before the data has actually been sent over the SPI bus even though this should require an OK reply from the board.

    regards

    ===============

    Brian Shankland

    Software Engineer

    Icon Research Ltd

    3 Raw Holdings

    East Calder

    West Lothian

    EH53 0HY

    UK

    Tel: +44 (0)1506 885000 (Ext 226)

    Fax: +44 (0)1506 885501

    <http://www.iconresearch.co.uk> www.iconresearch.co.uk