• Resolved

AFE7071: output of AFE7071 has unexpected spurious at 0.5*LO and 1.5*LO frequency

Prodigy 110 points

Replies: 10

Views: 304

Part Number: AFE7071

hi,i have some problem about AFE7071, i find the output of AFE7071 have some unexpected spurious at the 0.5*LO and 1.5*LO frequency.

when i change the LO frequency, the frequency of spurious will change also.  i use LMX2581 as the LO source, and check the output power 

of LMX2581 between -5dbm to +5dbm. and different LO power didn't affect the spurious much.

attached is the test photo of spectrum analyzer.   the output frequency is 400MHz and 600MHz.

anyone can help me to resolve this problem?


  • User, 


    Are you sending a pattern to the AFE7071? If so please specify the frequency & the number of tones. 

    Thanks

    Yusuf

  • In reply to Yusuf%20Agoro:

    i send a QAM modulation baseband signal to AFE7071, sample rate is 64Mhz, baseband signal bandwidth is 8Mhz. DVB-C standard.

    i make sure the baseband signal is no problem , it's a clear spectrum. 

    the photo is the RF output spectrum. i want to know why i find unwanted spectrum at 0.5x and 1.5x of the LO/RF frequency? 

    how to avoid this ?

  • In reply to user6150974:

    User, 

    can you send me the pattern file you are using so i am try and replicate results on my end?

    Yusuf

  • In reply to Yusuf%20Agoro:

    dac_in_data.rar

    attached is the pattern file.  it 's  txt file with 2 columns, the first is 14bits signed I data, the second column is Q data. 

  • In reply to Yusuf%20Agoro:

    hi Yusuf ,  when can i get your help? i want to resolve this problem in urgent.

  • In reply to Yusuf%20Agoro:

    hi Yusuf ,  when can i get your help? today is Sep 24 now, and i want to resolve this problem in urgent.

  • In reply to user6150974:

    User, 

    Can you send me a shematic of your board? I am not able to see this issue on my setup. I suspect that there is noise coupling onto the LO sourced by the LMX.  To test this,  disconnect the LMX and use another signal source for LO on the AFE. Is this possible to do?

    Yusuf

  • In reply to Yusuf%20Agoro:

    i don't think the hardware design has some problem. yesterday the local TI FAE bring the AFE707XEVM Evaluation Module to my office, and 

    we did some test together. the result is same. 

    the LO frequency is 400Mhz, we use a signal generator as LO source, the LO is clear sine wave.  when we output to the AFE7070 chip a  fixed

    DC signal.  the wanted signal is 400Mhz single tone, and the unwanted 200Mhz signal (0.5xLO frequency) is -57dbc. 

    but when we changed the data source to AFE7070 to a QAM(8Mhz bandwidth) signal(like the data file i provided last time), the signal at 200Mhz

    is about -38dbc compare to the main output signal at 400Mhz, like the photo i attached before. the same result as my own board.

    so can you test use my data file as source data to DAC? and it's the design or performance defect of this chip? or some special/reserved  register setting will

    help to resolve this ?  can you show me your test photo? 

  • In reply to user6150974:

    User,

    I was able to replicate the issue and find the solution. Please power down the LVDS output buffer by asserting bit 7 in register address 0x06. 

    Reference the following picture. 

    Thanks

    Yusuf

  • In reply to Yusuf%20Agoro:

    thanks a lot.  i have already resolve this problem after set pd_lvds to 1, the spectrum at 0.5*LO and 1.5*LO is clear now.