CC1312R: noise floor 433MHz

Hi René,

If you increase the rx bandwidth you will have a higher noise floor. In this case changing it from 78 kHz to 273 kHz the noise floor should rise by about 5-6 dB. 25 dB does seem a bit high.

Have you tried measuring the noise floor on a different frequency? For example, 434 MHz

Hi Diego,

in 43x MHz band it is the same. 868MHz is good, only 4-5dB higher noise. Could you test this with an eval board, LaunchXL-CC1312R1 e.g.?

br, René

I have tested with the LaunchXL-CC1312R1 and have no issues.

Have you tried measuring the noise floor on the whole 433 MHz band with an antenna connected to a spectrum analyzer?

Yes, we have testetd whole band. Terminated with 50Ω (spectrum analyzer) there is no problem, but with an external antenna direct connected to antenna port  (J7) we have the problem with the increase of the noise floor. Wtih an external antenna with 1m cable no problem, good noise floor. It seems that the problem comes from the LaunchXL-CC1312R1 .

br, René

Yes, we have testetd whole band. Terminated with 50Ω (spectrum analyzer) there is no problem, but with an external antenna direct connected to antenna port  (J7) we have the problem with the increase of the noise floor. Wtih an external antenna with 1m cable no problem, good noise floor. It seems that the problem comes from the LaunchXL-CC1312R1 .

br, René

By terminated with 50Ω (spectrum analyzer), do you mean that you connected the LaunchXL-CC1312R1 to a spectrum analyzer using a coaxial cable?

If that is correct, could you also test with the external antenna directly connected to the spectrum analyzer.

Conducted measurement is ok, but while connecting an antenna to port J7 the noise floor increases dramatically. C37 is mounted, C36 is removed. Connecting an antenna with 1m cable, RSSI / noise floor get better.

Please verify this setup.

We are aware that the conducted performance is OK.

It is useful to measure on the spectrum analyzer (SA) with an antenna connected directly to the SA to see what the noise floor you measure there. Please use max hold, set the span to the same as the Rx bandwidth and please share your results. 

There is no interferer (span / RX BW is 273kHz).

We also measured in a shielding box, disturbance comes directly from the board / CC1312.

Do you have another LaunchPad to test to see if you get the same results?

Yes, we tesetd with your EVBs and different ICs (CC1312R1, CC1312R7, CC1352P). The noise floor is to high in 433MHz band. With higher frequency noise floor gets better (8686MHz).

Hi Rene,

Could you share more information on how you did the measurement for "image 3: spectrum crystal" and "image 4: spectrum TCXO"?

We measured in mode "cont rx" and detect the spurs with a sniffer probe over the LaunchPad.

Do you have the same issue using a LaunchPad, cont rx, external rod antenna. How is your noise floor in 433MHz?

This is how the noise floor looks on my end. There is an increase, but not as much as in your setup.

Could you share images of the RSSI of both settings at 431 MHz and 435 MHz?

Here are the results at 431MHz with data rate 50k and 200k and also at 435MHz with 50k and 200k

We have the same problems with noise floor increasing when fixing (timeline off-on-off) an external antenna.

br, René

Hi René,

When you connect an antenna the noise floor will increase. That is normal behavior.

In the frequencies 431 and 435 MHz the noise floor increase when the RX bandwidth changed from 77.7 to 273.1 kHz, roughly less than 10 dB, is lower than when you did the same test at 433 MHz. This could indicate that around the 433 MHz frequency there is a external noise in your environment.

It is not really possible with the external noise. It comes directly from the LaunchPad, we test this in a shielding box with no external noise source.