CC1101-CC1190EM869RD: CC1101-CC1190 communication range issue

Hello,

Please find the snapshot of the CC1190 range extender portion above. 

Yes we have a spectrum analyser. I'll confirm on the frequency.

Until then you may have a look on the schematic above.

Hello,

I have tried changing the Saw filter in the schematic by SF2314E along with the 26Mhz crystal. The matching network elements were kept untouched.

However no much change in the RSSI values was observed. The frequency at which it transmits is 865.71 as confirmed with a spectrum analyser.

Our standalone CC1110 boards without range extender give a much better RSSI value of -30 to -35 dbm at the same distance (approx 1-1.5m apart) to avoid saturation.

What seems to go wrong with the Extender is not predictable. We have even changed register settings as per AN094 for setting maximum power output (in Dbm).

Section 7 of AN094 mentions about register changes particular to the range extender. However I could'nt find the FIFOTHR register in CC1110. Is that specific to CC1101 only? Is there any other register setting which needs to be changed in CC1110 ?

Also as per section 9.4 of AN094 , using a 27Mhz crystal would induce spur at 864 Mhz while using a 26Mhz crystal would induce spur at 871Mhz. As per the frequency band of my application (865-867Mhz) using 26 Mhz seems fruitful as per my understanding. Please confirm.

Regards,

Nidhi

Part Number: CC1101-CC1190EM869RD

Hello,

I have tried changing the Saw filter in the schematic by SF2314E along with the 26Mhz crystal. The matching network elements were kept untouched.

However no much change in the RSSI values was observed. The frequency at which it transmits is 865.71 as confirmed with a spectrum analyser.

Our standalone CC1110 boards without range extender give a much better RSSI value of -30 to -35 dbm at the same distance (approx 1-1.5m apart) to avoid saturation.

What seems to go wrong with the Extender is not predictable. We have even changed register settings as per AN094 for setting maximum power output (in Dbm).

Section 7 of AN094 mentions about register changes particular to the range extender. However I could'nt find the FIFOTHR register in CC1110. Is that specific to CC1101 only? Is there any other register setting which needs to be changed in CC1110 ?

Also as per section 9.4 of AN094 , using a 27Mhz crystal would induce spur at 864 Mhz while using a 26Mhz crystal would induce spur at 871Mhz. As per the frequency band of my application (865-867Mhz) using 26 Mhz seems fruitful as per my understanding. Please confirm.

Regards,

Nidhi

In the other thread you write that you have a spectrum. Which output power do you measure with the suggested power settings?

Have you tried to measure output power after C17 without U2 mounted to see that you get the expected output from CC1110?

Notes:
- L8/ C14 was used to get the correct impedance for the SAW filter in the ref design. For a different SAW filter other components may be required.
- L3/ C5, L4/ C6 form a notch on the 2. harmonic frequency. If you change the fundamental you may need to adjust the notches some.

Hello,

I've measured radiated power with RF explorer handheld spectrum analyser. 

I tried measuring radiated output power after isolating the CC1110 by connecting a wire antenna at the other end of C17 and breaking the path to saw filter. It gives RSSI values in the range of -35 to -40 dbm at about one meter distance or less approximately. 

Would there be any reference available to design the exact matching network for the saw filter and the notch filter values?

It seems matching those should be able to give much better RSSI values with CC1190.

Meaning that you are not able to do conducted measurements?

To debug a design with CC1190 you basically need to do conducted measurements. When you do radiated measurements like you do you have too many other factors that impact the results: Antenna tuning, you are not measuring TRP but the power in a random angle etc.

Hello,
Inspite of doing radiated tests, CC1110+CC1190 should be able to give atleast comparable result to standalone CC1110 if not better. However the RSSI values in the range of -80 dbm with CC1110+CC1190 indicate to me a design issue. Our standalone CC1110 boards give an RSSI value in the range of -35 to -40dbm tested with the same RF explorer spectrum analyser.. Probably is this due to the impedance matching network not being matched with the new filter as well as the notch filter values unchanged ? I would need support on designing on those filter networks and then i would want to perform conducted as well as radiated power tests. Possibly it should improve results.

Regards,
Nidhi lodha

Have you checked that the PA_EN, LNA_EN and HGM pins have the correct value? It could sounds like they are not set correctly.

You should not have to do anything with the CC1190 part if you follow the ref design, both schematic and layout.

Hello,

I forgot to mention we have set the PA_EN high, LNA_EN low and HGM high. 

Do these settings look fine?

We have hardcoded these on hardware end by unmounting the DNI resistors. Is that fine?

The schematic shows LNA_EN as high however we changed that to low while testing.

Also as you mentioned that we would have to change nothing if we follow the reference design, however B39871B3725U410 filter has a passband of only 2 Mhz while my application would be transmitting in the 865-867Mhz range. Hence as per my understanding I had changed the filter accordingly with 866.5Mhz center frequency.

Can I still use the same reference design without changing anything for my application?

Please have a look at the thread below I have mentioned all my settings here.

e2e.ti.com/.../722613

Regards,

Nidhi Lodha

Hello Sir,

Did you have a look at the schematics and all the above threads?

Could you please confirm if I my approach towards changing the saw filter and the need to design the matching network for it as well as notch filter for my frequency band is correct?

I assume the given reference design would not be fit for my application since I am using 865-867 Mhz band. Is that right? Or am I missing on something?

Please confirm 

Also would you let me know if there is any reference available to design the matching networks or any online calculator or application notes related to it which could be good to start with.

Regards,

Nidhi

If you look at one of my previous replies I'm commenting on specific components in your schematic so yes, I have looked at it. 

The match between the SAW filter and CC1190 is partially due to the SAW filter and partially due to the impedance on the CC1190 pins. See page 6 and 7 in the  on the impedance of CC1190. If your SAW filter has the same impedance as the one we use you can use the same matching. Also try to see how the performance change if you replace the SAW filter with a 0 ohm resistor.

But first you have yo be able to do conducted measurements. It's very difficult to debug and validate a RF design with an external PA without a spectrum since you have to also check for spurious etc. 

Thanks, can you also post the CC1101 section, why do you have 1K pullup on the digital control lines. Those values are way too low and not required. If you want pullups, then use 100K.

Regards,
/TA

Hello,

Hello,

Sure i'll check it with 100k pullups too.

You may find the CC1110 section.

We have replaced the SMD 26Mhz crystal with its through hole equivalent in assembly.

Regards,

Nidhi Lodha