CC8520: + CC5292 - Adequate radiated power but poor practical range

Hi,

What type of Antenna you are using?
Did you check the Antenna Matching?
Did you orient the board in different directions and check the RSSI?
Did you check on different Frequencies ( Low Band, Mid Band and High Band)?
If you have strong interferers (Like WiFi), it may affect the range. Please check at different Freq and different board orientations.
If you are using Whip Antenna then make sure the connector has good contact with the board connector. Some times the Pin on the SMA Connector on the Antenna is not contacting properly.

Thanks,
PM

Hi,

The protocol master uses a 50 ohm FXP70.07.0053A from Taoglas. The protocol slave uses a custom sheet metal antenna that has been tuned for 2.4GHz.
The antenna matching has been checked and VSWR was acceptable. We do measure the expected +20dBm peak radiated power on both master and slave.
We tried different orientations and used the strongest one for the line of sight test. RSSI is approximately -30dBm at 25ft for line of sight in the direction of peak signal strength.
We did confirm the signal power for radiated and conducted on all 18 channels.
There is no wifi or bluetooth on either device and an emissions scan showed no jammers.
This has been a consistent issue across all our units but yes we are verifying the connectors make contact.

Thanks,
Arthur

Hi,

Did you test in an Open site environment?
Just to see whether this is a Multipath effect or not?
Did you use our Range Estimator spread sheet which is on E2E Forum?

Thanks,
PM

Hi,

Our over the air tests were in an open site environment thus far.

Can you link me to the range estimator spreadsheet?

Thanks,
Arthur

Hi,

The following is the link to the Range Estimator.

The latest version is available at: http://www.ti.com/tool/rf-range-estimator

Thanks,

PM

Hi, how do I fill in the drop-down menu for my chipset?
CC8520 + CC2592

Neither is listed.

Thanks,
Arthur

Hi,

You can use "Custom". It is on the Top of the List. Then you can enter the parameters.

Thanks,
PM

I am not quite certain what values to enter...

Sensitivity: -90dBm from CC8520 datasheet in 2mbps mode. OK this is straightforward.
Blocking level: I see 43 dB for "Alternate channel, ±8MHz, wanted 3dB above sensitivity. 5 Mbps " in CC8520 datasheet. Not sure this is correct and I'm definitely using 2mbps mode not 5mbps.

The CC2592 has a noise figure of 4.7dB and adds 11dB of gain. How would I input these into the custom entry?

How are you accounting for the LNA gain of the range extender?

Hi,

The antenna efficiency we measured was about +2dB for the protocol master (more directional) and -2dB for the protocol slave (more omnidirectional). So I think 0dBi for both the Mean Effective Gains in the spreadsheet would be more appropriate. In that case the predicted range is 213m (using 2mbps sensitivity level of -90dB) which is much less than we're getting.

Thanks,
Arthur

Hi,

Can you point me to a spec for PSRR of the RF chipsets? I understand that these chips were designed to work directly off a battery input voltage, but my systems power the chips via a buck-boost regulator with some switching noise. Is there a measurement I can take to verify the power supply is clean enough to meet the performance in the datasheet?

Thanks,
Arthur

Hi,

I need some more info to understand and find the root cause for this issue.

Did you measure the Sensitivity (Conductive) on your Boards?
What is the Sensitivity? How did you measure it?
Did you measure the Gain of the Antenna on your Boards or you took the Spec from the Data sheet?
Can you share your Antenna test plots (Gain)?
Could you please also share all the Conductive test data on your boards (Tx power, Sensitivity etc..)?

Yes, for Audio applications we don't recommend to use DC-DC Converter because they are very noisy.
If you observe the Sensitivity is not meeting the spec then you may disconnect DC-DC converter and Supply the power from an external LAB Power supply and retest the Sensitivity.

Thanks,
PM

Hi,

We did testing in a GTEM cell, alternating the Master and Slave inside the chamber, with the other device outside the chamber with a conducted connection to the chamber's antenna.  This allowed the devices to connect to each other and we were able to simulate range by adding RF attenuators to the conducted connection.  We found that when the Protocol Slave was in the chamber we couldn't connect lower than an RSSI of -50dBm or so.  With the Protocol Master in the chamber we could receive down to RSSIs around -80dBm.  

We implemented some improved power filtering on the power supplies to the CC8520 and CC2592 on the Slave and fixed the poor sensitivity issue we found in the chamber.

It would be really helpful to get a specification for the power rail ripple or something like that.  We have a few different options for power filtering and would like to compare them and get assurance they are adequate.

Thanks,

Arthur

Hi,

OK, it seems the Range was affected by the Poor Sensitivity (-80dBm) on your boards. If I put -80dBm as Sensitivity in the Range Estimator, it shows the Range is around 80Metrs with -2dB Antenna Gain. This is in-line with your Range measurement.

Still I am not sure about your Antenna Gains on your board.

Yes, I will try to get you the Noise Spec. It is not readily available.

Attached is an Appnote on Range Estimation on CC8520.

Thanks,

PM

swra320.pdf

Hi,

Is the 5mVpp spec actually mentioned in the user guide? www.ti.com/.../swru250m.pdf

Just couldn't find it in there and didn't see it in the diagram you pasted. It's a very tight spec.

Thanks,
Arthur

OK, and would you expect that I could take a voltage ripple measurement on the evaluation modules and pass the 5mVpp requirement?