Post Go back to editing

ADG5404 cross talk?

hello,

ive some problem with my designed multiplexer card, i attach the schematic, see adg5404.jpg.

the address is set to 0:

if i switch +24V on out_1 and measure the results on out_mplex_1 and out_mplex_2,

now i get some heavy spark on the out_mplex_2, see osci_1.jpg.

the address is set to 1:

if i switch +24V on out_1 and measure the results on out_mplex_2 and out_mplex_3,

now i get some little spark on the out_mplex_3, see osci_2.jpg.

 

 

is this an design error?

why is the spark on out_mplex_2 such high?

some hints for me?

 

thanks a lot

attachments.zip
  • Hi Nico,

    Initially looking at your plots what comes to mind is switch charge injection effects.  I have attached an FAQ that has some Q&A on glitches caused by charge injection.  An experiment to do would be to increase the shunt capacitance on the output and see if it helps.

    Regards,

    Eric.

    Ask26-SwitchesandMultiplexers.pdf
  • Hi Nico,

    That's great that helped.  I think this would point towards the speed of the logic transition when using the 100k compared to the 10k affecting the charge injection mechanism.  The logic input is into a high impedance buffer with associated capacitance, see the equivalent circuit for the charge injection mechanism. If using pull up/down resistors I would typically use 10kohm resistors on our switch evaluation boards.

    Regards,

    Eric.


  • Hello,

    thanks for the information.

    i change the pulldown from the address line from 100k to 10k and it seems that the issue is gone.

    how could this?

    thx