• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 22 subscribers
  • Views 797 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.

LMK03806 IBIS simulation problems

Joseph Crop
Prodigy 20 points
Other Parts Discussed in Thread: LMK03806, CDCLVP1208

Hello, I am doing an IBIS simulation in hspice between the CDCLVP1208 and LMK03806 I/O buffer models using differential LVPECL signaling. The CDCLVP1208 simulates beautifully but the output buffer of the LMK03806 looks "horrible" and only outputs a 1000mVpp signal with the 2000mVpp setting which is not recoverable by the CDCLVP1208. I am trying to drive the CDCLVP1208 with the output of the LMK03806. Right now I have the output of the LMK03806 2000mVpp driver terminated with 150-ohms and ac-coupled to the input of the CDCLVP1208 termination component which has it's own 50-ohm termination. As you can see by the plot below there the LMK03806 output doesn't appear to have a proper pull down. I'm not sure what's going on and would really appreciate some help on this issue. Below is the hspice code I am using to generate these plots. Thanks for any help anyone can give!

*CDCLVP111RHB Simulation

* input source
v1 in0_p 0 pulse ( 0V 1V 1p 1p 1p 499p 1n )
v2 in0_n 0 pulse ( 1V 0V 1p 1p 1p 499p 1n )
v1a in2_p 0 pulse ( 0V 1V 1p 1p 1p 499p 1n )
v2a in2_n 0 pulse ( 1V 0V 1p 1p 1p 499p 1n )

vvdd vdd 0 3.3V

B_OUT_LVPECL2P0_P pu0_p pd0_p out0_p in0_p pc0_p gc0_p
+ file = 'lmk03806.ibs'
+ model = 'LVPECL2P0'
+ typ=typ
+ power=on
+ buffer=2
+ interpol=1

B_OUT_LVPECL2P0_N pu0_n pd0_n out0_n in0_n pc0_n gc0_n
+ file = 'lmk03806.ibs'
+ model = 'LVPECL2P0'
+ typ=typ
+ power=on
+ buffer=2
+ interpol=1


RC1 out0_n 0 150
RC2 out0_p 0 150
C1 out0_n out0_n_a 0.1u
C2 out0_p out0_p_a 0.1u

RT1 out0_n_t out0_n_a 50
RT2 out0_p_t out0_p_a 50

R1 out0_n_t term_out 50
R2 out0_p_t term_out 50

B_TERM pc_term gc_term term_out
+ file = 'cdclvpx.ibs'
+ model = 'VAC_REF'
+ typ=typ
+ power=on
+ buffer=17
+ interpol=1


B_IN_ECL_P v2p51 01 out0_p_t out1_p
+ file = 'cdclvpx.ibs'
+ model = 'INPUT'
+ typ=typ
+ power=on
+ buffer=1
+ interpol=1

B_IN_ECL_N v2p52 02 out0_n_t out1_n
+ file = 'cdclvpx.ibs'
+ model = 'INPUT'
+ typ=typ
+ power=on
+ buffer=1
+ interpol=1


B_OUT_ECL_N pu2_n out2_n in2_n pc2_n gc2_n
+ file = 'cdclvpx.ibs'
+ model = 'PECL_OUT'
+ typ=min
+ power=on
+ buffer=12
+ interpol=1

B_OUT_ECL_P pu2_p out2_p in2_p pc2_p gc2_p
+ file = 'cdclvpx.ibs'
+ model = 'PECL_OUT'
+ typ=min
+ power=on
+ buffer=12
+ interpol=1


r5t out2_p out2_p_t 50
r6t out2_n out2_n_t 50
r5 out2_p_t 0 50
r6 out2_n_t 0 50

vv2p5 v2p5 0 2.5V
vv1p3 v1p3 0 1.3V
vv1 v1 0 1V
vv2 v2 0 2V

.tran 1p 4n
.option post

.end

  • 0
    TI__Mastermind 32575 points

    Hello,

    Per the LMK03806 IBIS model notes:

                    -LVPECL simulation recommendation:
                      It is recommended that a small (1nF or less) AC coupling cap
                      be used. Larger values will result in long settling times 

    Does using smaller cap fix the simulation result? 

    Regards,
    Alan

  • 0 in reply to Alan O
    Prodigy 20 points

    Thanks for your reply Alan,

    Cap sizes doesn't seem to change the result. 

    The problem appears to be resulting from the "VAC_REF" termination model from the CDCVLP1208 chip.

    My understanding is that the termination is supposed to provide a common-mode reference that can be terminated by a 50-ohm resistor but the output node for the termination is following the signal on the other end of the resistor. In any case, when I provide my own reference everything starts looking good.

    Thanks for your help.

  • 0
    TI__Mastermind 28560 points

    Joseph Crop said:
    but the output buffer of the LMK03806 looks "horrible" and only outputs a 1000mVpp signal with the 2000mVpp setting

    Keep in mind, that 2000 mVpp is when considered differentially, so two 1000 mVpp single ended signals = 2000 mVpp differentially.

    73,

    Timothy