CC1310: PER varies with time - same location - 865Mhz to 867MHz

Hi,

you write that you have done a lest using the CC13-90 Launchpad but what was the result? Did you measure at the same time with the original board and saw a difference between the board with a SAW filter and a board without a SAW filter? 

Hi again,

We are still not able to repeat the same PER test at the field site with the CC13-90 boards.

But we did some testing in our lab to kind of simulating the mobile phone interference and compare the PER performance of the CC1310 Launchpad and CC13-90 Launchpad with SAW filter (866.5MHz)

Test Setup:

Two launchpad boards, one as Tx and other as Rx.

Tx board is surrounded by 10 mobile phones and Rx board is surounded by 10 mobile phones.

All the mobile phones were connected in a single conference voice call.

This test was done separately for CC1310 launchpads and then for CC13-90 launchpads (same frequency 866.5MHz and same power of 12dBm)

Not sure whether this is the correct way to check. But we tried.

Observation:

We noticed that the PER in the CC13-90 board is lesser than the CC1310 launchpads.

i.e. PER for CC1310 boards was 38% and PER for CC13-90 boards was 10.8% with mobile phones around.

And also the PER was 0 to 1.5% in both cases without mobile phones around.

I will share the detailed report soon.

So thats the reason, I was asking whether this SAW filter could play a part in reducing the interference.

We suspect only this mobile interference at the site as there are no other noticeable source for noise.

//Siva

Interesting test. 

To cover the main question first: Yes, a SAW filter will help with out of band interference. The chip has a given blocking performance @ 10 MHz and higher offset from the wanted frequency. This blocking will be dependent on the settings you use and you can find these numbers in the datasheet. Cell phones and base stations can operate on relatively high powers. Normally the cell phone power is adaptive meaning that the power is higher if the phone is far from the tower. Since the power is high, some of the energy will also be present on frequencies outside the band used by the cell and could be relatively high. Hence, in some cases it's required to add extra blocking to avoid interference from cells or other transmitters and for this, a SAW filter is a good solution. 

Interesting if this is caused by cell phone traffic since it sounded k\like it was dependent on the time of the day.  

Hi,

I have attached the test details and observations of the PER test comparison for CC1310 vs CC13-90 boards (with cell phones around).

One other question, rather than interference from cell phones, can a cell phone tower nearby be a source of the disturbance?

We have identifed a cell phone tower somewhat close to the problematic site. Thats why the question.

Also, finally we have managed to send the CC13-90 boards (with SAW filters) to the same site as well.

I will share the result of the PER test with these CC13-90 LaunchXL boards soon.

If these boards perform better in the site, then could SAW filter be a solution?

//Siva

PER Test - LaunchXL CC1310 vs CC13-90.xlsx

A base station could definitely impact. Some of the bands used by cell phones are relatively close to the ISM bands and some of the energy sent on these bands will also be sent on the ISM bands due to high output power and phase noise. 

To reduce the impact of the cell tower, using SAW filter is a good solution since a SAW filter will significantly attenuate the out of band noise. I believe that adding a SAW filter is fairly normal for applications that need to be robust in environments with a lot of other radio traffic.   

Hi,

Thanks for the clarity. I will update the results of the PER test with CC13-90 boards (with SAW filter) at the problematic site soon. 

//Siva

Hi,

Finally we managed to perform some PER tests at the problematic site with both LaunchXL CC1310 and LaunchXL CC13-90 boards. Please find the results attached in the excel sheet below.

The results show that LaunchXL CC13-90 (with SAW filter) performs considerably well in this noisy environment. At the same time, the performance of LaunchXL CC1310 (without SAW filter) was severly affected by the external noise.

With these results, can we be sure that the SAW filter will solve the problem due to external disturbance? Can we proceed implementing the same in our design?

In this case it looks like a SAW filter clearly improves the performance.

SAW filters should typically have a 50 ohm interface and I have seen evaluation boards for SAW filers. Meaning that it should be possible to connect a SAW filter in front of you board given that you have a SMA connector. Then you can do Board - SAW filter - Antenna on the location you have issues to see if a SAW filter solves the issue in your location.

Note that a SAW filter gives 2.5 - 3 dB lower output power/ poorer sensitivity when added in between the chip and the antenna.  

Hi,

Thanks for the clarification.

Sure, we will try to get an eval board for a SAW filter. But unfortunately, our board doesn't have an SMA connector.

We will find a way to connect it to our board and check it out at the same site. I will let you know the results

And yes agreed that SAW filter lowers the output power due to insertion loss.

//Siva

Hi TER,

To my understanding i think the SAW filter improves the selectivity. Correct me if am wrong. Is it possible to improve the selectivity by changing settings of the internal receiver? will there be any improvement if we change the IF of the internal receiver in this case? 

Yes, SAW improves selectivity/ blocking for far out interferes (outside the pass band of the SAW filter) 

The blocking will be lower for the image frequency (2xIF). Changing the IF could help in those cases that you know you have an interferer at the image frequency.

But in this case I believe you have interferers far out meaning that the blocking can't be improved by changing settings. 

In short, I believe it's not possible to get better blocking by changing chip settings and if better blocking is needed an external SAW filter has to be added.

Hi,

Thanks for the information. To just give it a try, could you please help us in changing the IF in Smartrf studio. Would like to change the IF in smartrf and do the same test again. 

Meanwhile we are also working to implement the SAW in our design. 

You can test with different IF using the field above in SmartRF Studio. 

Hi,

Can you please explain how the given value in the could be converted into frequency. The result we got from the 4.12 Signed format conversion is not suitable here. 

"For CMD_PROP_RADIO_DIV_SETUP, the intermediate frequency can be specified through the intFreq parameter, which calculates the setting in the modem for RX and is written to the configuration parameter area. If this parameter is 0x8000 and for CMD_PROP_RADIO_SETUP, a default intermediate frequency as given in Table 23-147 is used."

In the above description from Technical Reference manual it is understood that 0X8000 is a value to get a default intermediate frequency stated in the table. So there is a hidden formula in that and how to get that logic?

Using the equation 0x8000 does not give the default IF. 0x8000 is just selected as a number to mean default.

For testing you can use .intFreq = dec2hex(IF frequency*4096/1e6)

Hi,

We tested with changing IF but the result was not as expected like SAW filter.

Below is the Circuit with SAW filter added for our design. I have just added the SAW filter after the Balun and LC filter. Please give your feedback on it. 

RF Schematic.pdf

RF Schematic.pdf

Hi,

A1, A2, A3 should theoretically not be needed but I would include the footprints for at least the first alpha series of the boards to see if the SAW filter has a 50 ohm interface or if you get better result by performing a matching. 

Other than that the schematic looks good.