CC1200DK: RF Range Improvement

Please rad through  

Hi Team

Thank you. It is really helpful.

In bench level analysis, we are able to communicate with Sensor. I tried at -105dBm sensitivity also. But in field it is not working.Is there any proper method or process to tune the RF registers like, starting for carrier sense, RSSI analysis using Preamble, then AFC and AGC?

Or these setting will be automatically taken care by Stack itself?

Regards

Sreejith

As covered in the app note: Have you calculated the theoretical range with your setup?

" care by Stack itself", which stack are you referring to? 

Hi Team

RF calculation is meeting the range. Attached the image FYR.

Stack means, software/Register settings already given by TI.

Could you please share the register settings for 1.3KM range so that we can directly burn the program and test it using Eval board? If you can share a Software code , it will be really grateful.

Thanks.

The picture is missing in the post. Please use the "insert/ Edit Media" button to insert pictures. 

For the range estimation, have you measured the actual noise floor where you do the test? Due to increased radio traffic the noise floor is most often higher than the thermal noise floor and has to be considered. 

Unfortunately we don't have a database covering exactly what is sent to which customer. If possible, could you give a summary of what you have received here "Stack means, software/Register settings already given by TI"? 

Hi Team

I have attached the range calculation. I used SMART RF studio and was mentioning about the standard settings given for 920 MHz, 2FSK. In RSSI window of continuous Tx at -105dBm also we are able to detect the data in bench level data transfer test.

Thanks.

I see that you have used -110 dBm as sensitivty. The sensitivity for your settings will be different due to datarate and different RX BW. Have you measured the actual sensitivity? 

Do you mean that you are able to receive packet with -105 dBm? 

Do you use a PCB antenna or whip antenna? When using the range estimation tool it's  a requirement that you use the numbers that reflect the HW you are using. Is he TX power actually 14 dBm or something else for this module? 

Hi

I have received packet at -105dBm. Antenna is Monopole Whip antenna available in CC1200DK Kit. The transmitter power is 14 dBm.

Regards

Still I'm missing some parameters to understand your results.

- How long packets do you use?

- Which PER limit do you set for the range tests?

Which PER did you see @ -105 dBm. Could you confirm that you manage to receive at this level where you did your range test?

- What is the noise floor measured with SmartRF Studio using cont RX and no known TX on?

- Could you post a map of where you have tested? 

Hi Team

Could you please share a firmware which is proven at 1.3 Km?

Regards

For range tests it could be a fairly large difference between the theoretical range/ estimated range and the actual range measured in the field. The actual range is both time and place dependent.

- Time: I have seen cases where the background noise could vary more than 10 dB over time.

- Place: To get similar result as the range estimator the RX and TX board has to be 1 m above the ground and the the ground has to be flat (basically a football field or similar)

Hence I can do a test here with some parameters that give 1.3 km range but it's not given that those settings will give the same result in your end. 

--

That is why I'm trying to understand if the result you got is expected or not. 

Important parameters are:

- Sensitivity: I don't have this for the phy you are using.

- Background noise: I haven't received a number here.

I would like to get a feel on how it looks like where you do the test.

Thanks for the inputs.

Following are the parameters: Tx Frequency - 916MHz, Rx Filter- 200 KHz, Deviation, 55 KHz PER =10%, 2-FSK, Sensitivity expected --105 dBm, No background noise, Test condition- We are testing at Highway road condition,

RSSI graph level was coming around -107dBm at this condition.

I see that the link margin is set to 0 which will give a very optimistic estimate. If you set it to "no antenna diversity" you will get a estimate that is close to what you are actually seeing. If you want to have longer range you have to select a datarate that allows a lower RX BW (lower noise floor). Or increase the output power.

Thanks for the input. We will not be able to change the Rx filter band width, data rate or deviation as it is fixed by transmitter which is not in our control. Is there any other way you can suggest with register settings, for ex. AGC, RSSI etc.?

I am trying to avoid an additional PA on the board. Will this antenna help S1551AH-915S? It has a gain of 2dB and characteristics are better than TI Eval board antenna.

Hello Sreejith,

The antenna is passive and has 2dBi of directive gain in the perpendicular direction of the antenna compared to an theoretical isotropic antenna (the same energy in all directions). The following kit antenna 868-930 MHz: Pulse W5017 has the same directional gain (2 dBi). IF you used the 420-470 MHz: Pulse SPWH24433TI antenna, then yes it may help.

Hi Eirik

Thanks for the input. I am interested to know the link margin, the Path Loss calculator is considering for No antenna diversity. Also, can you please share a few inputs on how the tool is calculating the realistic distance(Without interference)?

Also I tried the range with Silicon Labs calculator. It is showing a different range estimation? Wondering why each calculator is showing different ranges, not able to correlate?

Any other suggestions to improve the range like reducing the Carrier Sense Threshold etc?

Thanks

Sreejith

Hello Sreejith,

The tool is explained here: 

• https://e2e.ti.com/support/wireless-connectivity/sub-1-ghz/f/156/t/375556 
• https://www.ti.com/tool/RF-RANGE-ESTIMATOR 
• https://www.ti.com/lit/pdf/swra479 

I do not know the dynamics behind the Silicon Labs calculator. But the TI excel range estimator use Friis line of sight and 2-Ray Ground Reflection Model. The No antenna diversity means adding a 15 dB link margin to get closer to a realistic achievable range and thereby reducing the range result as if the link budget was 15 dB lower (For a realistic link budget and range expectation, additional losses must be taken into account than the losses predicted by Friis).  The level of the guard band depends on the level of margin that is required. Theoretically, this can still be 0 dB and the radio link will still work. However, a certain guard band should be taken and this is normally in the range of 10 dB to 20 dB. For a system that requires a strong and reliable “fail safe” RF link then the margin could be increased furthermore.

There is also a Range Debug Check List tab in the excel tool.

I have to check the if Carrier Sense Threshold can be changed and could be useful or not.