CC1101: Unable to recevie packets with RSSI value less than -70 dBm at 904MHz

Hi,

Anaren has provided register settings for Data rate = 38.3835 KHz, RX filter BW = 464.285714 KHz, and  Deviation = 165.039062 KHZ. They have not provided the expected RX sensitivity for their settings.

At 250 kBaud data rate, and 540 KHz RX filter BW, CC1101 should be able to receive packets with RSSI value -95 dBm. In my case, the data rate and RX filter BW are lower therefore it should be able to receive packets with an RSSI value of -95 BdBm. But the maximum RSSI value of packets is -70 dBm. what should be the register values for optimum sensitivity with existing settings? The module is 15.247a2 and 15.249 certified with the above-mentioned configuration.

Regards

Ashok

I am a little confused as there are several different PHYs being mentioned - I will summarise what I understand to be the case so far:

Is this correct?

Please can you specify what your required PHY settings are (datarate, deviation, ect.)? I cannot advise on recommended register setttings without knowing more details.

The App Note SWRA122 that I linked to in my previous reply will help you with finding the needed settings. Additionally, you can also refer to SWRA682 (Finding Settings for new PHYs for the CC13xx Family): https://www.ti.com/lit/swra682

Regards,
Zack

ZC said:

Data rate = 38.4 kBaud, Deviation = 18 KHz, and RX BW = 70 KHz

Hi,

Yes, Anaren settings are not providing Rx sensitivity less than -70 dBm. I want to use the following configuration with optimization for sensitivity and expect to achieve an RX sensitivity of -95 dBm.

Carrier frequency = 904 MHz
Data rate = 38.3835 kBaud
Deviation = 165.039062 KHz
RX BW = 464.285714 KHz
Modulation format = 2-FSK
PA ramping = false
Channel spacing = 97.991943 KHz
Sync word qualifier mode = 30/32 sync word bits detected
Manchester enable = false
CRC enable = true

Xtal Frequency 26 MHz

Regards

Ashok

A couple of follow-on questions:

• What are you basing your expected RX Sensitivity value on?
• What parameters can you change for this, i.e. what is the most important factor here: the range or the TX PHY settings?

Regards,

Zack

Hi,

I am expecting RX sensitivity based on the assumption that CC1101 has RX sensitivity of -95 dBm at 250 kBaud and 540 KHz RX BW. It is capable of the same or better RX sensitivity for 464 KHz RX BW (which is the required RX BW).

PHY settings take precedence over range; I'm willing to sacrifice a few meters of coverage for optimal PHY settings.

Regards

Ashok

I am not able to test HW right now, but can you try increasing the RX BW setting to 540 kHz (I think the nearest value possible is 541.666667 kHz)?

You can use SmartRF Studio 7 to get the register settings for this.

Regards,

Zack

Additionally, please can you share details of how you are measuring this (your test setup)?

I'll adjust the RX BW to 541.6 KHz and report the results. As for the test setup, I have a node transmitting data to the gateway, and moving the node to monitor the RSSI value.

Hi,

Could you share the register settings I should use for testing, so I can accurately perform the test you're requesting?

Regards

Ashok

The easiest way (to speed up your development) would be to install SmartRF Studio 7 so that you can use it to generate the register settings for you in order to test different settings more rapidly; this is the recommended approach. I will share the register settings for the increased RX BW below.

For future reference, the CC1101 registers are detailed in CC1101 Datasheet: https://www.ti.com/lit/swrs061 with the RX BW setting being discussed in Section 13.

To increase the RX BW to 541.666667 kHz (i.e. 540 kHz), the register MDMCFG4 is changed to the following:

0x2A,  // MDMCFG4       Modem Configuration

Please share the impact on performance you observe.

Are you able to perform conducted testing initially? Details on how to do this are given in:

• Section 3 of SWRA603 (Debugging Communication Range): https://www.ti.com/lit/swra603  - it is also worth reading the rest of this document.
• Section 6 and Section 7 of SWRA370 (Basic RF Testing of CCxxxx Devices): https://www.ti.com/lit/swra370

The reason for suggesting conducted performance initially is that the datasheet numbers given are conducted measurements and it reduces the likelihood of additional factors negatively impact performance (such as a noisy environment) whilst tuning the PHY settings, i.e. simplifying the task to begin with. 

Regards,

Zack

Hi,

I have installed RF Studio 7 but I am not sure about the configuration to use. I changed the MDMCFG4 value to 0x2A and there was no improvement in the performance. I have attached the register settings used for the test. Please let me know the next steps or the register settings to be used for next testing.

:0003:47	FIFOTHR
:0004:D3	SYNC1
:0005:91	SYNC0
:0006:FF	PKTLEN
:0007:0C	PKTCTRL1
:0008:05	PKTCTRL0
:0009:00	ADDR
:000A:00	CHANNR
:000B:0C	FSCTRL1
:000C:00	FSCTRL0
:000D:22	FREQ2
:000E:C4	FREQ1
:000F:EC	FREQ0
:0010:2A	MDMCFG4
:0011:83	MDMCFG3
:0012:03	MDMCFG2
:0013:01	MDMCFG1
:0014:E5	MDMCFG0
:0015:65	DEVIATN
:0016:07	MCSM2
:0017:3c	MCSM1
:0018:18	MCSM0
:0019:15	FOCCFG
:001A:6C	BSCFG
:001B:47	AGCCTRL2
:001C:40	AGCCTRL1
:001D:91	AGCCTRL0
:001E:87	WOREVT1
:001F:6B	WOREVT0
:0020:F8	WORCTRL
:0021:57	FREND1
:0022:10	FREND0
:0023:E9	FSCAL3
:0024:2A	FSCAL2
:0025:00	FSCAL1
:0026:1F	FSCAL0
:0027:41	RCCTRL1
:0028:00	RCCTRL0
:0029:59	FSTEST
:002A:7F	PTEST
:002B:3F	AGCTEST
:002C:88	TEST2
:002D:31	TEST1
:002E:09	TEST0
:0030:00	PARTNUM
:0031:04	VERSION
:0032:00	FREQEST
:0033:00	LQI
:0034:80	RSSI
:0035:01	MARCSTATE
:0036:00	WORTIME1
:0037:00	WORTIME0
:0038:00	PKTSTATUS
:0039:94	VCO_VC_DAC
:003A:00	TXBYTES
:003B:00	RXBYTES
:003C:00	RCCTRL1_STATUS
:003D:00	RCCTRL0_STATUS

Regards

Ashok

Hi,

I attempted configurations at 250 kBaud with 127 kHz deviation and 540 kHz using register settings from Smart RF Studio. Despite this, the sensitivity only reached -79 dBm, which is still higher than the specified -95 dBm in the datasheet. For your reference, I have included the register settings.

Regards

Ashok

cc1101_register_settings.xlsx

Hi Ashok,

The settings you are proposing are not the same as the datasheet, so the expected performance will not be the same as that PHY setting; for example, the deviation for the PHY you wish to optimise is signficantly larger (20 kHz for the default TI PHY compared to 165 kHz for the Anaren PHY).

Without knowing the expected performance that Anaren have reported for this PHY (which we do not know unless I have missed something in the thread), I do not think it is possible to say that an RX Sensitivity of -95 dBm is expected for this and -95 dBm is not specified for these settings, the measurements presented in the datasheet are for the settings listed in the datasheet.

From what you indicated in your original post, other settings given by Anaren (that are closer to the default SmartRF Studio 7 settings) perform closer to the TI-given PHY. The thread I linked to in my first reply explains in more detail some of the other parameters affecting the expected RX Sensitivity.

Are you able to test the 902.5 MHz, 38.4 kBaud data rate, RX Sensitivity optimized, MDMCFG2.DEM_DCFILT_OFF=0
(GFSK, 1% packet error rate, 20 bytes packet length, 20 kHz deviation, 100 kHz digital channel filter bandwidth) PHY using SmartRF Studio (i.e. recommended settings)? It is important to establish that your measurement(s) of this known PHY agrees with the datasheet in your environment before continuing.

Another approach to finding settings that have performance closer to what you need are to use the process outlined in SWRA682 (Finding Settings for new PHYs for the CC13xx Family): https://www.ti.com/lit/swra682 :

• Begin with one of the recommended settings given in SmartRF Studio 7.
• Test the PHY to check your measured performance agrees with the datasheet numbers.
• Work through the App Note to find a combination of settings that fit your criteria.

As stated in SWRA682, all measurements should be conducted measurements for this: "All testing should be done conducted. Strong unwanted transmitters could impact the results and use of a shielded test setup should be considered if a base station or similar impacts the results."

Regards,

Zack

Hi,

I did test 902.5 MHz, 38.4 kBaud data rate, RX Sensitivity optimized, MDMCFG2.DEM_DCFILT_OFF=0 (GFSK, 1% packet error rate, 20 bytes packet length, 20 kHz deviation, 100 kHz digital channel filter bandwidth) and was able to receive packets with RSSI value -106 dBm. I can confirm that the test results match the results mentioned in the datasheet, the register settings is included below.

If the settings provided by Smart RF Studio differ from those that were used to achieve -95 dBm sensitivity mentioned in the datasheet for 250 kBaud, a deviation of 127 KHz, modulation set to 2GFSK, and RX bandwidth (RXBW) of 540 KHz. I would appreciate it if you could share the register settings that led to that level of sensitivity.

5355.cc1101_register_settings.xlsx

Regards

Ashok

Hi Ashok,

Thanks for clarifying - my point about the Data Rate = 38.4 kBaud, Deviation = 165 kHz, and RX BW = 465 kHz PHY was that as this is not a recommended PHY in SmartRF Studio (or the CC1101 datasheet) the same RX Sensitivity performance as another, recommended PHY cannot necessarily be expected, especially with no information provided by Anaren.

Please measure this PHY using a conducted measurement as it could be that the noise floor is higher in your location.

However, your measurement results for the Data Rate 250 kBaud, Deviation = 127 kHz, RX BW = 540 kHz SmartRF Studio PHY should be around what the datasheet figure is if they are conducted (the datasheet measurements are conducted), so figuring out the cause is the next step.

Taking a look at your attached register settings for the Data Rate 250 kBaud, Deviation = 127 kHz, RX BW = 540 kHz SmartRF Studio PHY the settings do not seem to match the settings those in the "Register Export" (setting the Base Frequency to 904 MHz).

I have highlighted the differences below:

The modulation format was set to 2-FSK instead of GSFK. So, please first try with the settings listed in the right hand column (for both TX and RX)

Regards,

Zack