It looks OK? on the master side but as we can see there is no answer..
Could I destroy RFM219 when I connect without leveltranslator at the beggining ?
I got another chip, connect directly to ESP32 with 3v3 levels and I have same situation...
I think that my SPI setup for 3WIRE SPI could be insufficient ?
When I swap chip to working wheather station we can see comunication:
I'm not sure I understand. The first picture in your post shows 4 channels. I can see the SPI Chip Select on the top trace, the SPI clock on the 2nd trace, MOSI on the 3rd and presumably MISO on the 4th?
It looks like the device is either not responding or is sending out 0xFF on the MISO trace. I'm not sure what is happening when the Chip Select line goes high briefly as there is activity on MOSI without any associated change in clock.
The second picture doesn't show the 4th trace, but the transfer of 4 bytes looks ok.
SPI is a 4 wire bus: SS (or CS), CLK, MOSI, MISO. Sometimes it's 3 wire, if the device only receives data so there's no need for a MISO line.
In your annotated LA picture, the bit where you wrote "there should be the answer", the answer comes back on the MISO (Master In Slave Out) line, not on the MOSI (Master Out Slave In) line.
Maybe you're thinking of an I2C bus where the data is transferred bidirectionally on the same wire?
EDIT: I took a quick look at the RFM219 manual that you provided in the initial post. para 5.6 on the SPI interface says:
The SDA is a bi-directional data pin. Address and data is always sent starting from the MSB.
Interestingly, the doc also refers to SPI as the Serial Port Interface. That is starting to suggest that HopeRF SPI is not the same as Arduino SPI.