spi1 does not work
BNNorman opened this issue · 11 comments
I'm trying to use an ILABS CHALLENGER RP2040 Lora 868M module. This has an RFM95 on-board with SX1276 chip and uses spi1 not spi0.
I'm trying to get the otaa_temperature_led example working on it
in main.c I have this.
// pin configuration for SX12xx radio module
const struct lorawan_sx12xx_settings sx12xx_settings = {
.spi = {
.inst = spi1,
.mosi = 11,
.miso = 12,
.sck = 10,
.nss = 3
},
.reset = 15,
.busy = 2,
// sx127x would use dio0 pin, and sx126x dont use it
// .dio0 = 7,
.dio1 = 20
};
These pins work fine using a simple SPI program to read the sx1276 Version (0x12).
However when using pico-lorawan your code blocks and never returns a value. The reason is in spi-board.c here:-
uint16_t SpiInOut( Spi_t *obj, uint16_t outData )
{
const uint8_t outDataB = (outData & 0xff);
uint8_t inDataB = 0x00;
spi_write_read_blocking((obj->SpiId == 0) ? spi0 : spi1, &outDataB, &inDataB, 1);
return inDataB;
}
obj->SpiId returned zero instead of 1 when I added print statements. It seems that the 'obj' hasn't been updated somewhere along the line. (I have to break off now but will hunt for that later)
Regards
Brian
Adding this at the start of SpiInit() in src/boards/rp2040/spi-board.c works
obj->SpiId=spiId;
Hi @BNNorman, I'm happy to see someone using this board as I have just purchased one and could not get it to work with this repository's examples.
Where did you get the pin assignments? I am referencing https://ilabs.se/challenger-rp2040-lora-datasheet/. The nss, reset, and dio1 pins on the data sheet differ from the pins specified in your code.
The pins are mentioned in that document if you scroll past the pinout diagram. The lora pins are not exposed. I also found the schematic which confirmed those gpio pin assignments. It was the spi1 implementation in the software which was causing the problem on this board.
@BNNorman The data sheet lists
GPIO14, GPIO15 and GPIO18 is connected to DIO0 – DIO2 respectively of the RFM95W.
but your code says
.dio1 = 20
Pin 20 is broken out for use on the board, physically labeled "SDI".
Also, the datasheet says
"GPIO9 – LORA_CS"
while your code says
.nss = 3
Pin 3 is also broken out on the board.
The sheet says
GPIO13 is connected to the reset input of the RFM95W.
Your code lists
.reset = 15
I thought GPIO15 should be the DIO1.
This is why I'm confused.
Please advise. I will try your updated spi-board.c code when I have time, hopefully tomorrow.
Ok, my bad re pins BUT, in my defence, m'lud, I was raising the issue of spi1 not working and copied the setup from another board I was working with. The attached schematic shows the correct LORA_xxx pins which are not available on the breakout connectors.
.
So ....
NSS=GPIO9
SCK=GPIO10
MOSI=GPIO11
MISO=GPIO12
RST=GPIO13
DIO0=GPIO14
DIO1=GPIO15
DIO2=GPIO18 (not used in the code)
@BNNorman Thank you for clarifying. I'm sorry, I did not intend to sound rude. I just wanted us to have the same starting point.
Cheers
@BNNorman would you be able to open a pull request for the change you suggested in #24 (comment) ?
I have forked your repository with the aim of trying to get it to work with the sx126x radio since Waveshare's pico-lorawan
code changes ignore my TTN keys even though it is based on your repo. (Which works fine on other HDW using sx1272 thatI have)
I also changed the default REGION to EU868 in the OTAA_temperature_LED example config.h.
If I create a pull request I think it will include both changes.
As you can see from the screen shot, there is just one addition after line 14. Is it worth the effort of a pull request just for that?
As you can see from the screen shot, there is just one addition after line 14. Is it worth the effort of a pull request just for that?
It's up to you, I was just thinking it would be nice for you to get credit for the fix.
I'm not that bothered. But thanks for the thought.