CC1352R: Sub-1 GHz forum

Peter Blomqvist

Other Parts Discussed in Thread: CC1352R

I am debugging CC1352R range. Using 868 MHz in Contiki the range is lower than expected, maybe 320m with both antennas 1.70m above ground and in open terrain.
Verifying this using Simplelink SDK 6.10, NORTOS rfEcho example (IEEE 868Mhz, 50kbps 100khz RX bandwidth etc) gives much more range, 600m it seems.

Comparing differences in code the Contiki uses an old SDK version, but what seems relevant is that the RF patches (rf_patch_cpe_prop.*) are different in the newer SDK version.
I wanted to apply the old patches but the rfEcho examples does not re-build the driverlib.lib holding the RF patches.

Is there a way to trigger rebuild of the SDK libraries?
Also, is the older FW a plausible explanation in the different ranges i experience?

Best

Peter

over 3 years ago

0 TER over 3 years ago

TI__Guru**** 317180 points

Note: Contiki is not supported directly in this forum, for Contiki questions the Contiki forums should be used.

OS or SDK version should not affect range.

- Are you performing the range test using the same HW, same antennas and at the same place when you get 320 m vs 600 m?

- Did you use exactly the same settings/ phy (datarate etc) in both cases?

- For re-build: You can try to include rf_patch_cpe_prop.c from a new SDK in the project, that this will be compiled/ used instead of the precompiled in the .lib file.

Range is set by the link budget: Output power and sensitivity. The patch should not impact this.

0 Peter Blomqvist over 3 years ago in reply to TER

Prodigy 80 points

Thanks,

Right. I have exacly the same devices, antennas, in same (outdoor) location at same height and range is very different.
Regarding PHY settings: output power should be 13dB in both cases, however, the Contiki version do use "multi mode" instead
of "prop mode" used in simplelink example - could this affect the range?

Best

Peter

0 TER over 3 years ago in reply to Peter Blomqvist

TI__Guru**** 317180 points

Could you check which Core SDK version that is used in the Contiki version that is used?

Are you able to do conducted measurements (using a spectrum analyzator)?

When operating on 868 MHz, "multi mode" vs "prop mode" should give the same result.

0 Peter Blomqvist over 3 years ago in reply to TER

Prodigy 80 points

The SDK version used by Contiki is really old. 4.x.
I was thinking trying examples of older SDK, but not sure it's available and i can get CCS working with such old SDK?

Best

Peter

0 Peter Blomqvist over 3 years ago in reply to Peter Blomqvist

Prodigy 80 points

I should add that running Contiki communication breaks down at about -90dB RSSI:
Would it be possible to mesure RSSI in the simple link SDK?
Comparing PA of a board running Contiki to a board runnning SDK examples seems interesting.

Best

Peter

0 TER over 3 years ago in reply to Peter Blomqvist

TI__Guru**** 317180 points

Btw: Have you tried to test with SmartRF Studio? The RSSI is reported for all received packets.

RSSI for the SDK: https://dev.ti.com/tirex/content/simplelink_cc13xx_cc26xx_sdk_6_10_00_29/docs/proprietary-rf/proprietary-rf-users-guide/proprietary-rf/cc1101-migration.html#rssi

Does Contiki use some sort of carrier sense? Since the communication break down at fairly high RSSI level, could if be that a threshold is set and the sensitivity is limited this way? A 6 dB difference in the link budget gives 2x range.

0 Peter Blomqvist over 3 years ago in reply to TER

Prodigy 80 points

I modified the rfEcho example to output RSSI values and there i get much more reasonable values - the link works down to -100dB or even bit less.
The carrier sense is a good point, i will investigate that.

Best
Peter

0 TER over 3 years ago in reply to Peter Blomqvist

TI__Guru**** 317180 points

Did you manage to find out why you got poor results?

0 Peter Blomqvist over 3 years ago in reply to TER

Prodigy 80 points

Changing RSSI threshold for CCA seems to improve things.

A followup question: What is the default value? The rfEchoRx code does not set this threshold.

Best
Peter

0 TER over 3 years ago in reply to Peter Blomqvist

TI__Guru**** 317180 points

Depends. As one example ETSI define the threshold in the relevant standards. This is typically a value that has to be set according to the usage.