1 Reply Latest reply on Aug 17, 2011 11:07 AM by Steve_McKinney

    Measuring waveform at device pins

    david.royle

      Greetings,

      I would like to observe the signal at the device pins (in order to verify eye opening spec at receiver pins). Would it make sense to use a pass through receiver AMI model without termination as shown?

      Ideally I would like to be able to insert jitter at the receiver as well.

      If so, does anyone have such a model they could share with me?

      I'd really appreciate it.

      Dave

       

      ami.jpg

        • 1. Re: Measuring waveform at device pins
          Steve_McKinney

          Hi Dave,

           

          This is a tough one.  The problem you'll find is that by doing a mid-signal probe, you're going to have a completely closed eye because of reflections. 

           

          What spec are you designing to?  Is it one of the standards (such as PCIe or something like that)?  Most of the time they have 'compliance points' where you need to validate the eye.  This typically means replacing whatever receiver you've got with a testload (often 100 ohms differential).  In this case, what I would suggest is to remove the package and the AMI model and just have the termination resistor at the end of the channel.  If your eye mets the spec here, then you should be OK. 

           

          However, if you have an equalizer in the receiver, you're really selling yourself short trying to meet a pin compliance point for a measurement.  You may be overly constraining the design when you have the ability to recover some additional losses within the receiver because of that equalizer.  You're better off knowing how that signal looks coming out of the equalizer which is the beauty of having an AMI model - you get that visbility you can't get with a pin measurement.

           

          -Steve