CC1120: Max power output for the EVB of CC1120-CC1190 is low

This could be related to the powering of the board. Please test with an external supply as well.

We have seen previously that an external supply is sometimes needed to achieve the highest output power.

The output power is still around 21dBm when the customer apply external power supply, the current drawn is about 140mA, it seems that it is not the DC power issue, is it related to the register setting?

Send a continuous Tx unmodulated carrier in studio. Otherwise the settings look correct in SmartRF studio

Are you using the US or EU version of the CC1120 CC1190 ?

Could you check from below PCB and firmware revision to confirm whether it is US version.

I try to send a unmodulated continuous TX and the tx power is -30dBm, the setting and result is attached below.

We try to follow the document below to set the jumper to transmit continous TX, but we can not find IOCG2 and IOCG0 register, please advice.

The IOCFG registers are on the top in the register overview (The right hand pan in the SmartRF Studio window) 

Hi,

The default BOM should be for the 915 MHz band if the customer got the board from https://www.ti.com/tool/BOOSTXL-CC1120-90 and has not modified the BOM. We cannot tell from a photograph.

Without setting the IOCG2 and IOCG0 registers the device will not configure the CC1190 properly:

I have not tested the difference the HW revision settings make for this in SmartRF Studio, but the 868 MHz hardware revision is selected in your above SmartRF Studio screenshot.

Regards,

Zack

If the CC1190 PA is not enabled the signal will be 30 dB attenuated so I believe that the board just give poor output power. I have seen the same on CC1310-CC1190 LPs also. 

All my test is done at US band 902 - 928MHz,
The document mention about IOCG0 and IOCG2, but not IOCFG0 and IOCFG2, are they equal?
According to the swra492 document, we don't need to modify the register setting on the software when transmit packet mode as below, I use transmit packet mode and the TX power is also about 22dBm, is the register value has to be changed in order to get 30dBm at 915MHz?

Is there any feedback?

Thanks

With the jumpers on P6 connector you can set the PA_EN high and LNA_EN low; this should then set the CC1190 PA correctly for Tx mode.

HGM can also be set high to achieve a higher output power. 

Measure the CC1190 control signals just to make sure you have set these correctly.

Then you just need to transmit a constant Tx signal, and you should see a high output power.

What output power are you measuring now ?

Sending a constant Tx signal or packet Tx signal will not change the output power level.

Since there is some confusion here on setting the CC1190 control signals. Just start with the constant Tx signal first since to confirm you achieve a high output power. 

We set high to P6_1 and P6_11 and set low to P6_7, We still get 21dBm at both continuous TX and packet TX mode, we expected the output power should be about 27dBm. 

OK. Then the control of CC1190 is correct despite the output power being too low.

Is the antenna completely disconnected when you are measuring to the spectrum analyzer ?

Can you re-solder a new semi-rigid cable to double-check that the soldering is OK or the cable is not damaged ?

The soldering looks suspect but difficult to say from your picture.

Can you also measure the output power with a static unmodulated carrier as well ?

Adding another question:

The customer find the outdoor range result is not good. They think it relate to the PER counting method. They would like to modify the payload format. Could they get the EV board source code?

Please concentrate on achieving the conducted performance first.

Once the conducted performance is as expected; then we can move on to the radiated performance.

If they are missing 6 dB on the Tx side, then the range will be 4x less than expected.

Clear photo as below, I unsolder the 0ohm (C40) and resolder the semi-rigid cable again and the output power is still 21dBm.

Thanks for the update. 

For debugging purposes, can try with a higher output power from CC1120 to see if you see a higher output power with CC1190.

When choosing the CC1190 range extender, with 27 dBm power setting. PA_CFG is set to 0x77 which equates to 11 dBm from CC1120.

Can you change PA_CFG to the following and measure the output power:

PA_CFG: 0x79 (12 dBm from CC1120)

PA_CFG: 0x7B (13 dBm from CC1120)

PA_CFG: 0x7D (14 dBm from CC1120) 

Required power as below:

Could you pls help to comment on this?

Hi Fabio,

There is a possibility that their board has been damaged since they should be seeing a greater output power.

Do they have several boards to test ?

I would repeat the previous test and measure at the CC112x_RF position with the SAW filter removed. This will determine if the CC1120 is performing as expected or if the issue is related to the CC1190. But I would use another test boards since something is incorrect here.

They got 2 boards and both output power is 21dBm when PA_CFG is set 77, please review the below data, I can hardly believe both EVB are damaged, is measurement with SAW filter removed still meaningful?

Is their RF equipment calibrated ?

If they have poor ESD handling it is not impossible that two boards can be damaged if they have been handled in the same manner. 

I'll see if we have a board in the lab available so we can measure the output power out of the box. 

The feedback from the customer:

Our equipment is calibrated,
We handle the EVB in the same way as our own product and we do power measurement every day for different electronic product, please let me know what kind of special ESD handling should be done on the EVB.
Please let me know your measurement result if it is available.