CC1352P: Is it possible to activate PA permanently without sending any data

What are you going to use this number for? A small signal impedance has no meaning for a PA with a high output swing. Also see "OptimumLoadImpedance" in https://www.ti.com/lit/an/swra640e/swra640e.pdf

Not sure why you are using 1720BL15B0200E since the optimal impedance seen from the RF pins is not 200 ohm.

For the 20 dBm PA the reference design should be followed. Note that we are probably going to change the BOM for the LAUNCHXL-CC1352P-2 since we have lately found that this BOM is not fully optimal for output powers lower than 20 dBm.

Hi,

We already had knowledge about the app note swra640e. We are using the chip with the same principle as on the launchpad LAUNCHXL-CC1352P. On our first version, we were using discrete components as on the launchpad. But during our tests, we had problems of board untimely reset. After investigations, we found that the reset was happening during 20dBm Tx emissions, so we tried another matching on the Tx lines to fix this problem and we had some good results but unfortunately, that was not that stable on all the devices, especially at low temperatures. The only answered we found to this problem was "follow the reference design". Not that easy cause our board form factor is not really the same, but for the RF part, we copied the design as close as possible (layout and stackup).

Then we found this app note: SWRZ076D. This is how comes the 1720BL15B0200E balun on the Tx 20dBm output. We manufactured some 12 new prototypes using this balun and no one is resetting at low temperature. And we have some good radio link range either which is quite positive to us.

Now you say:

->  "the optimal impedance seen from the RF pins is not 200 ohm." : which RF pins are you talking about ? Because on the app note SWRA640E that you mentioned on your answer, it is said that "863-928 MHz high power PA target load impedance: 200 Ω" (page 30)

-> "A small signal impedance has no meaning for a PA with a high output swing": We tried to measure the impedance of the Tx circuit, seen from the antenna, but for that, we need the PA is powered ON and not generating any signal (because high power signal is disturbing the VNA).

-> "we have lately found that this BOM is not fully optimal for output powers lower than 20 dBm.". As for us, we are using the board at 20dBm power during Tx phases,  not lower.  So it is ok.

-> "What are you going to use this number for". Outdoor application. Countryside area.

Regards.

Laurent.

Which frequency are you operatingon?  This comment "we have lately found that this BOM is not fully optimal for output powers lower than 20 dBm." is valid for 2.4 GHz.

The numbers we state is optimal load impedance. Meaning that this is the impedance the PA wants to see. Measuring this could be a bit difficult since you need to measure a differential impedance (remoiving the CC1352P and measure the impedance seen into the network). An optiona is to simulate it. This is something I normally recommend when it's not possible to follow the reference design due to PCB shape etc. If you use ADS to simulate S11, S21 for the reference design (use 50 ohm termination on all ports) and compare with your layout and adjust components, layout if different.

When we investigated on clock loss we found that the following override gave significant improvement:

// Set TXRX pin to 0 in RX/TX and high impedance in idle.

HW_REG_OVERRIDE(0x60A8,0x0001),

Please verify that you have this override.

Hi,

Thanks for your reply but this is not answering my previous request that is, by the way, mentioned in the title:

is it possible to set de power amplifier ON while not sending any signal (neither modulated nor CW), just amplified noise in its output ? Is it possible to set it ON via smart RF studio for instance or by using a dedicate software that turns ON the PA via internal registers ? Sorry to be insistent (and maybe arrogant) regarding this matter but the only thing I need is a "yes" or a "no". And if it is a "yes", I'd like to know how.

As far as I understand from smart RF studio (and I tried so many times to activate PA on different ways but I couldn't reach it), the PA is activated only on Tx demand, that is to say when a data is ready to be transmitted.

I also tried, in the past to measure the output impedance (we have a VNA) on our prototype but the chip is so small that this measure is complicated to me to have "clean" and reliable measurements. So I had to rely on the information mentioned in the application note if I don't want to waste too much time or money to built a dedicated custom prototype for RF measurements. Anyway I can't activate PA without sending any data so for now I can measure nothing (just the Rx circuits that are easy to measure, since you are on the receiver part and then there is no data out). The only thing I wanted to check is to see if the RF switch output match with our own antenna. Those measurements are easier for me since I can connect directly on the prototype output (supposed to be an unbalanced 50 ohms output). Moreover I am sure to have the real impedance since the product is encapsulated in its casing. On the other hand I have the impedance of the antenna with all its environment (boards and casing). This thing is also easy to measure for me.

Anyway, to answer your questions:

* I don't have ADS so I can't simulate anything. This software is quite expensive and we don't have enough volumes for now on radio products to buy such a tool. Anyway to you have the S-parameters of the TX 20dBm output ?

* We operate on the 902MHz - 928MHz band. We also use the 2.4GHz function but for closer radio link range.

Best regards and wishes for you and your family.

Laurent.

It's not possible to turn the PA on and not send something with available FW. We have never measured the impedance of the PA since we haven't seen any use of this number. It's not clear from the above how you would use the number.

The RF switch should be 50 ohms but you should match the antenna. How to tune the antenna, see http://www.ti.com/lit/swra161

Ok thanks, that's ok.

Regard.