CC1310: Frequency hopping problems

From the information given I can't really say why it behaves like this.

- Could you reduce the number of frequency hopping channels to 2-3 and ideally set up a sniffer for all channels to see what is going on? 

- When you turned off frequency hopping, did you get the RSSI you would expect? 

Hi TER, 

1-I thought it would be some kind of RF filter problems in my custom board, so I decided to reduce the channel number to 4 and ideally separating them from each other as follows

#define CONFIG_FH_CHANNEL_MASK {0x81, 0x40, 0x20, 0x00, 0x00, 0x00, \
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, \
0x00, 0x00, 0x00, 0x00, 0x00 }

So here I am using channel 0 7 13 and 20. This is the same configuration for the FH_ASYNC_CHANNEL_MASK. I have noticed that my custom board works very well when in debugging in CCS. Once I remove from debug I get that strange behaviour.

2-When I turnoff frequency hopping I get a good RSSI, it is more or less the same as the launchpad board RSSI. 

To my opinion it is related to the macACKWaitDuration. According to the specification macACKWaitDuration=54 symbol periods for the 50kbps configuration. Since it is not able to receive an ACK from the coordinator the device is trying to send and AssocRequest to the other channel, and it does this for all the configured channels.

Let me know what you can suggest.

Regards

Sokol

" I have noticed that my custom board works very well when in debugging in CCS. Once I remove from debug I get that strange behaviour.": In the cases that things works with the debugger and not without it's typically a issue with the 32 kHz xtal. If the  debugger is connected the chip is prevented from going to standby.

First, test the empty example and see if this run. Also to to change to the 32 kHz RCOSC in the CCFG.c to see if that gives a different result.

Hi TER,

I tested the empty example and it works very well. My custom board does not have an external 32KHz, so I am using the internal RC oscillator( changed the CCFG.c file). I am using the oscilloscope to measure the sleep time on the empty project and that seems pretty fine.

Let me know what you think about that.

Thank you 

It sounds like a timing issue due to the use of RCOSC. Not sure if we have tested running the stack without 32 kHz xtal with frequency hopping.

If you have sniffers on the hopping channels used you could be able to see from the log when the hopping happens.

Hi TER,

To prove that this is not related to the RCOSC I did the following:

I removed the XOCS from the lauchpad and then I changed the CCFG.c file configuration to #define SET_CCFG_MODE_CONF_SCLK_LF_OPTION            0x3        // LF RCOSC . Even with this configuration the launchpad works pretty fine. To me it seems like a RF filter problems but I am not sure.

Thank you

Interesting test. I can take a quick review of your schematic and layout to see if I see something that is strange. If you don't want the design public on the forum you can send me a friend request with the gerber files for layout and the schematic as pdf.

Any update? 

Hi Ter sorry for the late response, I was out of the office for a couple of days. Please find attached the pdf files.

Let me know when you have a look at it.
Note that the antenna has been designed to be an external antenna not an integrated on PCB.

Regards
VALVE_1.0.pdf

Hi Ter,

Did you have time to have a look at it?

I could not see anything strange on the schematic. Have you checked that the center frequencies are correct? Since the 24 Mhz is clock master this has to be correct to avoid any drift. 

Have you looked into this: If you have sniffers on the hopping channels used you could be able to see from the log when the hopping happens.

I have checked the frequencies using HACKRF and analyzing them by inspectrum.I did the tuning of the internal cap array in order to put center frequencies as close as possible.

I have not tried it yet to have multiple sniffers on the hopping channels. I will update you once I give it a try.

Thanks 

Sokol

Any update?