• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 20 subscribers
  • Views 119 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.

LMX2572LP: Fpd Spurs at 960MHz output

Andrew Small
Prodigy 20 points
Part Number: LMX2572LP

Hi,

I'm using the LMX2572LP as a TX/RX Vco, 30MHz to 960MHz. It's working great for the lower range of frequncies but I'm getting spurs at the higher frequncies 800+, looks to be the Fpd (50MHz) freq getting into the output. I've played around with the MASH, Fpd, Charge pump gain, and the loop filters to impove it the best I can. 

On the LMX2572LP EVM these spurs are much lower with the exact same registers programmed. The PCB Layout is very simliar, and component values are the exact same. The Ocsillator input (25MHz) on Eval board i'm using an external signal generator on my board I'm using a VCTCXO.

Any suggestions to reduce these spurs?

EVAL Board plot:

My Board plot:

Registers:

R125 0x7D0820
R124 0x7C0000
R123 0x7B0000
R122 0x7A0000
R121 0x790000
R120 0x780000
R119 0x770000
R118 0x760000
R117 0x750064
R116 0x740000
R115 0x730050
R114 0x727C27
R113 0x710000
R112 0x700000
R111 0x6F0000
R110 0x6E0000
R109 0x6D0000
R108 0x6C0000
R107 0x6B0000
R106 0x6A0007
R105 0x694440
R104 0x680000
R103 0x670000
R102 0x660000
R101 0x650000
R100 0x640000
R99 0x630000
R98 0x620000
R97 0x610000
R96 0x600000
R95 0x5F0000
R94 0x5E0000
R93 0x5D0000
R92 0x5C0000
R91 0x5B0000
R90 0x5A0000
R89 0x590000
R88 0x580000
R87 0x570000
R86 0x560000
R85 0x550000
R84 0x540000
R83 0x530000
R82 0x520000
R81 0x510000
R80 0x500000
R79 0x4F0000
R78 0x4E00CD
R77 0x4D0000
R76 0x4C000C
R75 0x4B0840
R74 0x4A0000
R73 0x49003F
R72 0x480001
R71 0x470081
R70 0x46C350
R69 0x450000
R68 0x4403E8
R67 0x430000
R66 0x4201F4
R65 0x410000
R64 0x401388
R63 0x3F0000
R62 0x3E00AF
R61 0x3D00A8
R60 0x3C03E8
R59 0x3B0001
R58 0x3A9001
R57 0x390020
R56 0x380000
R55 0x370000
R54 0x360000
R53 0x350000
R52 0x340421
R51 0x330080
R50 0x320080
R49 0x314180
R48 0x3003E0
R47 0x2F0300
R46 0x2E07F0
R45 0x2DC608
R44 0x2C12A1
R43 0x2BCCCC
R42 0x2ACCCC
R41 0x290000
R40 0x280000
R39 0x27FFFF
R38 0x26FFFF
R37 0x250105
R36 0x24004C
R35 0x230004
R34 0x220010
R33 0x211E01
R32 0x2005BF
R31 0x1FC3E6
R30 0x1E0CA6
R29 0x1D0000
R28 0x1C0488
R27 0x1B0002
R26 0x1A0808
R25 0x190624
R24 0x18071A
R23 0x17007C
R22 0x160001
R21 0x150409
R20 0x145048
R19 0x1327B7
R18 0x120064
R17 0x11009B
R16 0x100080
R15 0x0F060E
R14 0x0E1808
R13 0x0D4000
R12 0x0C5001
R11 0x0BB018
R10 0x0A1178
R9 0x090004
R8 0x082000
R7 0x0700B2
R6 0x06C802
R5 0x0538C8
R4 0x040A43
R3 0x030782
R2 0x020500
R1 0x010808
R0 0x00211C

Schematic:

  • 0
    TI__Guru* 88130 points

    Hi Andrew,

    Looks like you have set MULT=2 to get fpd=50MHz. This is not a valid configuration, you should use OSC_2X to double fosc. 

    In addition, you have set IPBUF_TERM = 1, which will enable an internal 50Ω termination to OSCin pin. This setting works fine with a 50Ω source such as a RF signal generator. TXCO, especially clipped-sine wave output with 1Vpp output swing, is not able to drive 50Ω load. We should make this register bit = 0.

  • 0 in reply to Noel Fung
    Prodigy 20 points

    Hi Noel,

    Thanks, I changed both the suggestions made no difference to the spurs. We actually already were using OSC_2X, that multi=2 was a mistake.

    Originally we were using 100MHz Fpd, 100MHz works really well with most of our freq range. We also had the IPBUF_TERM=0 that must have got changed while we were exprimenting with different settings. We have played around with the fdp freq found that a lower freq improved the spurs at 960MHz on both EVM and my board. The spurs on the EVM seem to consistency about 20 to 30dB lower no matter the settings. So might be more of hardware or layout issue, just can't seem to find any glaring issues. 

    Thanks

    Andrew Small

  • 0 in reply to Andrew Small
    TI__Guru* 88130 points

    Hi Andrew,

    If possible, can you try not splitting the reference clock?

  • 0 in reply to Noel Fung
    Prodigy 20 points

    Hi Noel,

    I tried not spltting the reference, No change in the spurs.

    We are using this same reference for another PLL on the board so both are in sync.

    I also tried completely removing this reference and feeding in the same signal generator I'm using on the EVM, no change to the spurs.

    The output is connected to an IQ modulator and a voltage controlled attenuator I was measuring after these 2 parts. I disconnected all that and measuring directly at the output of the LMX2572LP now. Spurs did improve slightly but still a good 15 to 20dB higher then then EVM. 

  • 0 in reply to Andrew Small
    Prodigy 20 points

    Hi Noel,

    I have it working much better now, Worst spur is at 50MHz @ -75dBc everything else is more then -90dBc. This should be acceptable but I could add a switched in highpass filter to reduce it further. A filter wouldn't have helped before because the spurs were too close to the fundamental frequency.

    I was using the output as single ended, loading the unused side with a 51ohms resistor. I noticed on the EVM when I was measuring it, the unused side was not terminated. Removing the termination resistor on my board greatly reduced the spurs. The IQ modulator we are using has the option for a differential input. Having both the RFoutAP and RFoutAM connected to the modulator vastly improved the spurs.

    We still have a spur issue on the RX side RFoutB, That side goes to a single ended mixer so I don't really have the option to connect both M and P sides. I will experiment with it worst case I can switch to a mixer that has a differential inputs. Removing the termination on the unused side helps but probably not good to leave it floating I'm assuming.