• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 53 subscribers
  • Views 243 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CC1190: Lower Output Power than expected with CC1190 from one revision to another.

citizen
Expert 2780 points
Part Number: CC1190
Other Parts Discussed in Thread: CC1120

Hi,

We are going through testing the final revision of a custom prototype that includes a CC1190 PA/LNA and CC112X radio amongst other things on the board.

We are noticing a reduction in conducted output power (around 24,7dBm on average), and also in received post LNA power, about 2dB less on some of the most recent boards. This is in comparison to some prior prototypes that had an output power closer to 26dBm, albeit the sample size was smaller as we had not produced as many boards as more recently ...

Nevertheless also going by some prior experience with this Radio/PA/LNA combination that we have employed in other products without this inconsistency, it has made us a touch concerned about the reduction in Output/Input reception. This we have now observed on 8 different prototypes, and in some cases the output can vary from 24,2dBm to 25,4dBm. Our threshold commonly used to evaluate a well functioning RF output stage is to check an output carrier to be at about 26dBm or so, but less than 25dBm would be a "show-stopper" as far as production testing goes ... so we generally would not pass such radios and generally have not seen many ...

Between a PCB that passes, and one that does not pass, we have in both cases checked the CC190 input power, and in both cases its at about 7dBm, so it looks like the CC1190 may have some unforeseen attenuation somewhere that we cannot pin down as such as yet..

We are wondering why, and would like to know if this is simply within the expected device tolerances as this Max/Min values are not specified in the data sheet (only typical values are given for the absolute maximum output possible) ?? What are some of the expected minimum and maximum values expected for output power ? is a ± 2dB variation from device to device acceptable and within expectations or not ?

Furthermore, in one case where the power was 2dB lower, we removed the CC1190 and replaced it, and then noticed a 2,5dB to 3dB improvement in receiver LNA output gain in Rx mode, but no equivalent improvement in the output power in Tx mode as we were expecting to also see ... Generally the link budget would be severely affected if we loose, say 2dB in Tx mode, then also 2dB in Rx mode, all adds up to possibly to over 4dB loss in link-budget, hence our concern with what otherwise would be considered "a few dBs" ...

So we have a few possibilities that could be further looked at and perhaps some suggestions from TI could assist :

  1. Replacing the CC1190 with a new one on a PCB and observing some improvement on the LNA gain and post-LNA output, would seem to suggest there may have been something wrong with the CC1190 from the start, or with potentially the assembly process ?!? Any thoughts on this ?

  2. Generally speaking, how sensitive are these CC1190s PA/LNA devices (and may be also the CC112X radios) to humidity during PCB assembly or should this not be a major concern ? Otherwise a bad RF ground could present this type of observation ?

  3. Is a 2dB - 3dB variation in PA output, and/or LNA gain, a normal variation from device to device, or is this way too much, assuming the input power (from a CC112x) is pretty much constant ?

Looking forward to your reply.

Regards,

citizen

  • 0
    TI__Guru**** 317180 points

    I have seen something similar, see chapter 6 in https://www.ti.com/lit/an/swra542/swra542.pdf

    I will see if I'm able to pull some production test data. CC1120 + CC1190 has only been tested with 6 boards meaning that we don't have a lot of statistical data for this combination. 

    On the highest power setting I would expect a variation of +/- 1 dB for output power based on other PAs. The gain in the device will have some variation and this will cause a variation in output power from device to device. The variation will increase for lower output powers. 

    Since you see a delta if you swap the CC1190 I don't think the PCB variation impact much. 

  • 0 in reply to TER
    TI__Guru**** 317180 points

    I pulled some production test data and max-avg = 1.1 dB and avg-min = 2.8 dB for a population of 600 k. This means that what you measure is close to the min/ max data. I doubt that you manage to find two samples that are at min/ max of a full population, I would assume also some other factors contribute.  

    Note that the numbers are takes from production test where the variation in load impedance will be larger than on a PCB meaning that I would expect the variation on a PCB will be less since the impedance will be better controlled.