CC1120+cc1190 @ 915MHz cant sniff the traffic.

TER,
Can we communicate to you through Skype if we won't be able to solve the problem offline?

We will try to make a video tomorrow.

TER,

Since we got some result with SimpliciTI, we postponed the sniffing video yet. Will return to it later.

For now there is another question. When we start PER test we can see strong signal on the spectrum analyzer. However, could not get it during the SimpliciTI communication. Running the stack  on CC430 (Olimex board) showed RSSI 38, while our board only 14. The analyzer could not catch the short packets. After that we looped sending data and could discover the signal, it was very poor - about -70dBm on our board.

Also after setting correct value to the CC1190 control got signal about -50dBm.Found that we need to change PA_CFG2 which set to 0x2B by default. Playing with that value did not give any change. Then checked, what is set in SmartRF Studio - it is 0x6B, while the documentation says that the bit 6 in this register reserved. After setting the same value in the beginning we still did not have any changes, then suddenly found that the level is high - about the level we would expected. However, it does not go to this level immediately after start. It runs some time (about a second) with low level, than sends several frequencies for a shirt time and only after that goes to the high level. See the moment of multiple frequencies at the picture

906MHz is the working frequency and the others are shortly appeared signals.

We think it is not right thing, something is wrong. Could you comment, what it might be and how to fix?

At the meantime Pavel is going to write a terminal support to have ability to change parameters (starting from the Tx power) from a computer.

- PA_CFG2[6] is PA ramp enable. The time the PA uses to get to the final output power is the same if this bit is set to '0' or '1' hence the bit is set to reserved and equal to '1'.

- Just to summarize some, please let me know if I got something wrong: You have tested the CC1120 + CC1190 and got expected output power. If you use the PER test you get the expected power. With SimpliciTI you get lower than expected.

- When debugging try to measure the output power conducted. Reading the RSSI doesn't tell too much about what is going on.

- Have you checked that you have turned off frequency agility and that the control signal to the CC1190 is correctly set for all modes?

Hi Ter,

Thanks.
Using your advice, We can see better power level on Spectrum analyzer. And looks like, we are ready to make range test, with SIMPLICITI.

But there are another question: since CC1190 has low efficiency below 20dbm output power, what to do to transmit from CC1120 trough CC1190, just using

CC1120 power up to 14dbm? Is it possible? Or any kind of settings to increase transmit efficiency below  20dbm?

Thanks.

Pavel Kotelnikov

TER,

Here what we got powerwise:

1. When we originaly tested CC1120+CC1190 we did not have any tools to measure the power. Just assumed, that PER Test program says the correct number.
2. After we got the spectrum analyzer (actually a scope with SA module up to 3.5GHz) we got estimation of that signal. With 5dBi  antennas on modules and 2dBi antenna on input of the SA we got the signal level about 0dBm with distance between antennas  about 50cm. All air testing we provide with this set of antennas. 
3. When we started the SimpliciTI and looped sending data we got about -70dBm
4. After playing with the register settings we got the picture shown above: in the beginning the level is about -30dBm, then four frequencies appear and after that 1 frequency with power about 0dBm.
5. At least we know now that in our configuration and SimpliciTI we can get the similar level as we get with PER test. 

After reading your post with the frequency agility mention we figured out, that this is probably the case. Could you tell how to turn it off tor testing purposes?

With 15dB attenuator and RF cables we have around, we touched output of our module and got +8.5dBm. So, we actually close to the expected Tx power. Ordered more appropriate cable to connect directly to the module output connector and measure the power more accurate. Pavel is going to write a UART comminucation protocol that will allow us to change Tx Power level from a computer.

And, yes, we understand that RSSI gives just some estimation :-)

And let me rephrase Pavel's posting a little bit.

Our project is battery operated, so we are going to change Tx power - get the Tx power lower fot the closest to the central stations end devices and increase for those that go far away. Reading the datasheets we figured out, that we need HGM only if we need TX power higher than +23dBm. Using LGM we can save some power. And one of the questions is: what happens if we reset PA_EN? Will the ouptut signal go to the antenna passing the PA internally or it is not possible and we need to set PA_EN every time when we want to send thу data? Passing PA could be very usefull for Tx Power up to +8-9 dBm (absolute maximum input is only 10dBm, otherwise would be great to have +14dBm directly from CC1120 :-))

Any advice about this idea?

I'm not too familiar with the SImpliciTI code but I think you set it in smpl_nwk_config.dat, deactivate DxFREQUENCY_AGILITY.

If PA_EN is set low the signal will be attenuated 30 - 40 dB through CC1190, the chip does not have a bypass function. If bypass is needed you have to implement this with RF switches.

1) How to read real received signal strength of pair CC1120+CC1190?

2) How to adjust threshold of SIMPLICITI or any other parameters if  CC1120+CC1190 used vs CC1120?

3) in SENSOR NETWORK example, provided by TI when AP is restarted, previously  joined ED cannot rejoin again,

how to resolve it?