• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 17 subscribers
  • Views 1316 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.

CC85XXDK-HEADSET: Bidirectional VS unidirectional --- RF RANGE

Antoine
Prodigy 80 points

Part Number: CC85XXDK-HEADSET

I recently received a CC88XXDK-headset, witch is a dev-kit for digital wireless audio streaming. At first use, I tried the bi-directional headset demo on PurePath wireless configurator, works well, no drop out until 7-8 meters. As I kept reading and testing the devices, I tried to put the devices on unidirectional mode, with stereo input on the master, because I will not use  bidirectional mode for my future applications. The problem is : when I select : downstream -> stereo input base (2/0) for the master and : downstream -> stereo headphones (2/0) for the slave, I lose all range (less than 1 meter) and there are drop-outs every 2 seconds.

As soon as I put at least one back at bidirectional, performances return to normal and I have decent range.

The only thing I changed on the board is the supply, which is at 3.6V to maximise the range.

Master AND slave settings while in unidirectional mode :

TX power : 5 dBm

range extender : CC2590 (HGM)

used RF channel : all

RF data rate : 5 Mbps

protocol version : A

Master settings : 17 ms latency (832 samples), streaming format : SLAC.

If there is something I did wrong (or didn't do) or if you experienced this situation, please let me know.

Thank you.

  • 0
    TI__Guru** 106705 points
    Hi Antoine,

    The default example for the Headset kit in PPW Configurator (CC85XXDK-HS Preloaded Demo) is unidirectional, have you tried this without any changes?

    Are you using Stereo output headphone or satellite on the slave side?

    Cheers,
    Fredrik
  • 0 in reply to Fredrik K
    Prodigy 80 points

    Hi Fedrik,

    Since I posted my problem, I tried many, many configurations and I probably tried the preloaded demo at one time. I do now have equal performances between bi and unidirectional settings.Latency and data rate were the game changer in audio-range and number of drop-outs.

    For the slave setting, I use Stereo Headphone.

    For now, my concern has gone to audio drop-outs, witch are still occurring too often. As I now test with 2 pairs (2 masters and 2 slaves, master sending to slave) with both same settings in a spacy room, with at least 2 feet of space between each devices, I experience drop-outs even if I'm close (1-4 meters).

    I used 2 Mbps to improve salve sensibility ; deactivated fade-out ; automatic selection for timeslot period (I didn't notice difference between max and min, no matter the latency) ; Protocol A ; TX power 5dBm ; 25.333 ms latency (48kHz) and slac format. These were the settings where I got the best results for unidirectional stereo streaming. The range isn't a problem anymore, I can go up to 20 meters and won't lose signal, but there are still those drop-out that come from nowhere and occur at any range. I do not have other signals that might interfere in the bandwidth.

    Now to improve robustness, i'm looking hardware : use the CC2592 and try antenna diversity. What are your feelings about the CC2592? I know it will increase the range, but will it increased the robustness within range compared to CC2590?


    Thank you,

    Antoine

  • 0 in reply to Antoine
    TI__Mastermind 27145 points
    Hi,

    If the robustness problem is due to the Weak signal reception at the Receiver then using CC2592 would increase the Range and Robustness because it has higher transmit power ( 22dBm) than CC2590.

    If you observe the drop out at any range then this may not be the Tx power issue. You can check the RSSI at the Receiver. If the RSSI is lot higher than the Receive sensitivity then Tx power may not be issue.

    I recommend to check your power supply noise. If there is a switching noise on the Power supply rail then it may cause some burst errors.

    Thanks,
    PM
  • 0 in reply to PM
    Prodigy 80 points
    Hi PM,

    To check the RSSI, I used Purepath Commander with two devices, one as TX and the other as RX. As I could see, the RSSI at receiver was at -30 dBm (average) at close range. It dropped to -60 dBm while being close to max range, but I didn't noticed any drops below -80 (value where I think everything should crash) except when I really wanted to kill the signal. Positioning (found the sweet angle where the DN0007 inverted-F antenna has 0 gain) also influenced the RSSI but as I experienced drop-out while the devices were steady, I don't think the problem is the radiation pattern of the antenna.

    For my supply, I scoped at L1 ( I use the eval-kit) and it does have some drops, but nothing out of ordinary for a switching regulator. As I checked the zip file of the eval-kit, I noticed in the changes (excel file) that there was once a LDO to supply devices. What we have on eval-kit is a BUCK/Charger 2-in-1 (BQ25015). Would it be better to use LDO to supply?

    Thank you,
    Antoine
  • 0 in reply to Antoine
    TI__Mastermind 27145 points

    Hi,

    In your post, you mentioned that you didn't see any drop outs below -80dBm. That means the link is working at it's optimum range with CC85xx.

    As per the CC85xx data sheet, the sensitivity is at -83dBm ( 5MBPS, 0.1% BER). You will see some errors, when the input level is around -83dBm.

    Yes, definitely LDO has less noise than Switching Regulators. So, better to use LDO.

    But to make sure whether this is a Power Supply (Switching Regulator) issue or not, you can disconnect the on-board power supply and use an External Power supply (Lab Power supplies) to the boards and test the link. If you don't see any drop outs then the issue with the Power Supply on the board. If you still see the issue then the problem will be something (may be external RF interference like WiFi signals) else.

    Thanks,

    PM

  • 0 in reply to PM
    Prodigy 80 points

    Hi PM,

    I back checked myself with the supply noise and I don't know what I've done to miss it, but the supplies are a lot more noisy than it should. It doesn't come from the buck regulator as the glitches are not occurring at 1MHz. Those glitches have epic drops (300 mV) and I suppose they occur when there is a packet transmitted because glitches happened less frequently when I probed the receiver than with the transmitter.

    So I've add some capacitances (10 uF) on each power supply and it has greatly reduce the noise on each device. I'll test them all day long and let you know if it solved the problem. I will also use external power supply if the capacitances doesn't help.

    Again, thanks for your help,
    Antoine

  • 0 in reply to Antoine
    Prodigy 80 points
    Adding those 10 uF on each power supply really helped (AVDD_A,AVDD_R,AVDD_CC2590 and VDD) . Still not perfect but I do have solid performances above 18 ms latency.

    With an LDO and better decouping, I should be able to have even better performances on my design than the eval-kit.

    Thanks for the support.
    Antoine