full duplex

[Radio_Man]

Hi 

I want to know if its recommended to insert RF filter  between Antenna and

 

 Rx input in full duplex mode to avoid LNA2 saturation when Tx is transmitting (Tx frequency =2.55Ghz , Rx frequency =2.3Ghz).

any will be help appreciated 

 

[Ricardas Vadoklis]

Hello,

Yes it is recommended, not only to filter the TX output but also to filter rest of the spectrum to avoid receiver saturation.

Best Regards,
Ricardas Vadoklis,
Senior System and Test Engineer,
Lime microsystems, 
r.vad...@limemicro.com
+44 0 1483 685 065

--
You received this message because you are subscribed to the Google Groups "Lime Microsystems open-source support channel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to limemicro-openso...@googlegroups.com.
Visit this group at http://groups.google.com/group/limemicro-opensource.

[Radio_Man]

hi

Thank you for your quick reply . please  can you clarify the following  

1-what's the maximum input power of LNA,

2-it's possible to  saturate TXVGA  if I set them to maximum gain , what's the recommended gain that I can use to minimize intermodulation distortion 

3- did you have any suggestion that i should apply  to improve the lms6002d  linearity 

sincerely 

[Ricardas Vadoklis]

Hello,

See my comments below.

Best Regards,
Ricardas Vadoklis,

On 9/3/2015 11:43 PM, Radio_Man wrote:

hi

Thank you for your quick reply . please  can you clarify the following  

1-what's the maximum input power of LNA,

+12 dBm with no damage.

2-it's possible to  saturate TXVGA  if I set them to maximum gain , what's the recommended gain that I can use to minimize intermodulation distortion

For best TX performance, set TXVGA1 to -10 dB and vary TxVGA2 as you require.

3- did you have any suggestion that i should apply  to improve the lms6002d  linearity 

sincerely 

For Tx path, mainly TxVAG1 gain settings defines Tx linearity, set it to -10 dB. Receiver linearity depends on the input signal level.  Receiver has three variable gain stages: LNA, VGA1 and VGA2. To maintain sensitivity and improve linearity with high input signal levels, reduce VGA2 first, followed by VGA1 and LNA. 

[Radio_Man]

HI Vadoklis

you said that maximum input to the LNA is 12dbm but in FAQ documents 5.10 I found that the max input is -60 dbm what's the difference  

[r.vad...@limemicro.com]

Hi, 

+12dBm maximum input signal to LNA input wit no damage, assuming that other gain stages are at min gain. If you were to set receiver path to max gain (VGA2, VGA1 and LNA) then -60 dBm is a maximum signal level not to saturate receiver path.

Best Regards,
Ricardas Vadoklis

From: Radio_Man Sent: Friday, 4 September 2015 21:11 To: Lime Microsystems open-source support channel Reply To: limemicro-...@googlegroups.com Cc: r.vad...@limemicro.com Subject: Re: full duplex

HI Vadoklis

you said that maximum input to the LNA is 12dbm but in FAQ documents 5.10 I found that the max input is -60 dbm what's the difference  

--

[Radio_Man]

Hi 
thank you for the clarification, I want to know if the following method is correct or not 

 I receive a signal with -40 dbm  , RXVGA are set to maximum gain,  after detection the AGC module will decrease them  to get  a suitable  values   

[r.vad...@limemicro.com]

Hi, 

That's correct. You would need to define your threshold to satisfy your SNR. AGC is automated function, reduces the gain based on your threshold.

Best Regards,
Ricardas Vadoklis

From: Radio_Man Sent: Saturday, 5 September 2015 00:42

To: Lime Microsystems open-source support channel

Cc: r.vad...@limemicro.com Subject: Re: full duplex

[Radio_Man]

Hi Vadoklis

thank you for clarifying this.  

please can you tell me how to use the rf loop back and peak detector

 to calibrate LO leakage and IQ imbalance automatically (Rx and Tx side)  , because I applied many algorithms which work with MATLAB models but fails when I apply  them to  received signal from the LMS6002D ,with this issue our project will fail because we want to use 64 QAM  modulation  and without a good SNR we can't achieve this.
best regards 

 

[Ricardas Vadoklis]

Hello,

Please find attached file that describes the calibration concept using RF loopback. It also shows the calibration blocks that have to be implement in BB in order to calibrate the chip.

To set up chip for RF loopback use the following procedure:

0. (optionally) Disable PAs. This is not strictly necessary, but reduce noise.
1. Register 0x75 LNASEL_RXFE[1:0] = lna_num: Selects the active LNA.
2. Register 0x25 SELOUT[1:0] = lna_num: Select output buffer in RX PLL
3. Register 0x7D[0] = 1: Power down LNAs with a test mode register
4. Register 0x70[1] = 1: Enable test mode for RX FE for the previous setting to take effect
5. Register 0x08 LBRFEN[3:0] = lna_num: Select RXMIX input for loopback

Best Regards,
Ricardas Vadoklis,

[Radio_Man]

Hi  Vadoklis 

Thank you for replying

I set the registers as you said to enable the loopback but In Rx I get very weak signal as you can see in the attached picture  (TXVGA1 set to -10,RXVGA to max )

,what's wrong ?? there are any other setup should be applied to get suitable values 
what about  AuxPA  should be enabled ???

Best Regards,

[Ricardas Vadoklis]

Hello,

Thank you for your graph.

AuxPA is enabled by default. Are you sending some signal through DAC's? FFT plot would me informative in this case.

You can try to set TxVGA2 gain to 22, this should increase the AuxPA gain to max as both are controlled through the same register.

I have attached the GUI project file and register setup list that set LMS6002D for RF loopback mode, for your reference.

Best Regards,
Ricardas Vadoklis,

[Radio_Man]

Hi 

Yeah I'm sending a sine wave and TXVGA2 set to 25 db

[Radio_Man]

Hi Vadoklis 

 

please find attached the FFT plot of the looped  sine wave  (figure FFT_intloopBack),

 and  register setup  (where Tx frequency =2500Mhz and Rxfrequency=2501Mhz) . 

can you tell me what's wrong with this setup .

[David]

Hi Radio_man,

Try the RF loopback at a lower frequency, say 400Mhz, and see if the base band signal is cleaner.

I had similar problems with RF loop back at 2.5GHz and didn't really solve it. It seems there is either more noise at the higher frequency in the loop path or less RF signal level into the RX at 2.5GHz. Either way, moving to a lower frequency gave better results.

Also, power down RXVGA2 and RXLPF DC calibration comparators after calibration as these add 6dB of noise I found. (See FAQ document, around 5.30 or 5.26 I recall, depending on which version you are reading.)

I hope this helps.

Cheers

David

[Radio_Man]

hi

thank you David, i will try it , but  in previous discussion I found  this

 

LMS6002D calibration is stable for entire 3GPP band, for e.g if your were to operate Tx path at Band1 ( frequency 2110 – 2170MHz), we would recommend to calibrate TX in the middle of band (at 2140 MHz). This calibration should work at any frequency with in the band. 

which means that we can't apply calibration done on  400Mhz  to 2.5 GHZ (target frequency )