• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 20 subscribers
  • Views 677 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Anyone get CC2520 CC2591 combo working consistently?

cleidigh
Prodigy 140 points
Other Parts Discussed in Thread: CC2591, CC2520, CC2530

We have been working for quite some time to get a MSP430 CC2520 + CC2591 design working.  The CC2591 needed a lot of work to improve it's stability through decoupling changes.  That effort seems to have worked.  We are not able to get consistency in either power or EVM accross even a small number of PC boards.  We have been trying several different match approaches including the eval module designs.  Bottom line is that we get EVM all over the place and no hope of a consistent design\manufacture picture.  The experiments have been done with quite a lot of analysis and even simulation, this has not been a haphazard approach.  Ultimately the design must be stable and have decent EVM\power across -40C to +70C, however, we cannot even get romm temp to behave enough at this point.  Most indictions are a poor EVM\feedback problem with the CC2520.

 

Has anyone got a board or module to work in a statistically consistent fashion??  If so would you be willing to share the match or other key info?

 

Much appreciated

Very frustrated CC2520 user.....

 

  • 0
    Prodigy 20 points

    Hi,

     

    We have the same problems :(

    Stability is a main issue for cc2591!

    We are using ref design but so far no luck :(

    Maybe you managed to work out what was wrong!

    Good Luck,

    Dominik

  • 0 in reply to Dominik Cirmirakis
    Prodigy 60 points

    Hi,

    I've also observed some instablility of the CC2591 PA. I saw this while playing with a 3-stub tuner on the output. I could find some settings of the stub tuner where spurious oscillations did appear. The setup was the following: Transmitter+CC2591 PA on the same PCB with SMA output, 10cm cable, 3-stub tuner, 20cm cable, spectrum analyzer.

    Then I took the official evaluation module from TI and measured its stability factors (K, u1 and u2) on a VNA. There is no information about stability in the datasheet, and this is a pitty. I found that around 1.8 GHz, all 3 stability factors can go below 1, which means that the PA is not unconditionnaly stable. I've done the meeasurement at 0dBm input and -30 dBm input. In both cases the problem appear. (see measurements below). So already on the official design the problem is visible, and the PA is not unconditionnaly stable.

    For the moment I could only see oscillations using the tripple stub tuner, probably under bad matching conditions, but I fear that when the PA is connected to an antenna, it might oscillated in certain circonstumances. And this because the antenna is a printed one for a portable device, and the environment is may change quite a lot, and so the matching conditions.

    I've also tried with a 3dB attenuator between the PA output and the 3-stub tuner, and this seems to have stabilized the PA. At least I couldn't find a position of the 3-stub tuner where an oscillation appeared. But it's a crude testing method, I cannot pretend that I've checked all possible loads. And the source matching didn'vary. I've also tested with an additional  cable of about 1 dB loss instead of the attenuator, but oscillation could still be found. So 1 dB seems not to be enough to stabilitze the device, and 3 do the job, but is far too much for our application.

    Any comment or idea to improve the stability factor of the device, without loosing too much power? As engineer we know that Murphy is never far away, and if you've seen something strange, even once on a prototype, and didn't understand and solve it, it will FOR SURE appear later on the market....

    Best Regards

    Yves OESCH

    PS: Here are my stability factor measurement of an official CC2591 Evaluation kit module (CC2591EM)

    HB9DTX

  • 0 in reply to Yves OESCH
    Prodigy 60 points

    I've added a 10 pF capacitor between the BIAS pinand ground. It helped massively. I've not been able anymore to find an oscillation, although I looked for by using a triple-stub tuner on the output of the PA, to mimic different loads conditions.

    Earlier this capacitor was 1nF and oscillations were very easy to obtain. To be noted: on the TI application board, this BIAS node is  not decoupled at all.

    Anybody has experience with decoupling this BIAS node?

  • 0 in reply to Yves OESCH
    Intellectual 260 points

    Hi all,

    I am currently testing CC2591 (custom design) and unfortunately I have encountered stability problems too. I do not have a VNA so I can't perform detailed stability analysis.

    I follow OESCH suggestion and I put  a capacitor in parallel to RBIAS.

    It helps so much but instead of 10 pF I have used 0.68 pF.

    I am been using a FR4 board with dielectric thickness 0.175 um.

    Anyway the output power is 4 dB below under the specification.

    Best regards

    Oscar

  • 0 in reply to SmartWater
    Intellectual 260 points

    Hi,

    we tested the LNA in HGM and it seems that we reach a gain of about 20 dB (Input power, -93 dBm @ 2.4 GHz; output power -73 dBm).

    Best regards

    Oscar

  • 0 in reply to SmartWater
    Expert 1510 points

    Hi all,

    I understand you are banging your head agains the wall here. The design around CC2591 must be done with great care!

    If you want a good way out and can use the CC2530, I would recommend RC24xx module from Radiocrafts. All FCC certified and with different antenna options.

    I have personally tested these at -40 degree and with VSWR 3:1 and different phases with a line stretcher.  And output power still is 20 dBm. REALLY!

    BR

    Ørjan