DS90UB962-Q1: EYE DIAGRAM

Hi Sherry,

The CMLOUT gives the FPD-Link signal after it is recovered by the AEQ circuit.  Therefore it will still have spread spectrum and probably not be captured on the scope very well.  I would recommend that you remove the SSC for a proper eye diagram.

As for seeing the signal as it comes through the coax cable, recall the AEQ circuit I mentioned above.  The AEQ will amplify and recover the FPD-Link signal which the scope will not do.  You should always read on the CMLOUT pins to see what the deserializer is actually recovering.

Regards,

Nick

Hi Nick,

Thanks for your reply!

Can we test eye diagram on the 935 DOUT directly?

Thanks & Best Regards,

Sherry

Hi Sherry,

Are you asking if you can measure the FPD-Link signal directly from the cable or at the actual pin?  Either way I don't think it would be very meaningful and you probably wouldn't see an open eye.  The reason you need to measure it using the CMLOUT pins is because the DES has internal circuitry to add gain at the FPD-Link frequency and open the eye.  By measuring at the CMLOUT pins you are basically seeing what the DES sees which is most important.

Regards,

Nick

Hi Nick,

Yes, the customer wants to check the signal quality of 935's output, so they want to test the eye diagram of 935's DOUT pin. When they test the eye diagram of 960's CMLOUT pin, it also includes the coax cable's signal quality. If the result is bad, they don't know which one plays an important role, cable or 935.

Thanks & Best Regards,

Sherry

Hi Sherry,

You can measure at the DOUT pins of the serializer but again it would not be very meaningful.  Really the best way is to see the eye as it is seen by the deserializer.  You could compare the opening when using different lengths of cable to see the affect but you could also use the MAP tool that shows how many available AEQ settings.

Regards,

Nick

Hi Nick,

Thanks for your reply!

I can understand "The reason you need to measure it using the CMLOUT pins is because the DES has internal circuitry to add gain at the FPD-Link frequency and open the eye". Because the signal will be attenuated through a long cable, and the internal circuit of DES can compensates for cable losses. Therefore we measure eye diagram at CMLOUT pin instead of RIN pins. But I still can't understand why we can't measure it at DOUT pins because there is no cable losses at this pin, then the internal circuit is not necessary when tested at DOUT pins. Could you please help explain more about this? Is it related to the format of the signal?

Thanks & Best Regards,

Sherry

Hi Sherry,

You can certainly measure the signal on the DOUT pins of the serializer and you will see an eye.  I am just trying to say that I don't really think this would be meaningful.  Like you said, there is no loss so the only information you would get out of this measurement is that the signal is present.  Feel free to measure it but what are your going to prove with that measurement?

Regards,

Nick

Hi Nick,

We are summarizing the factors which will impact the eye diagram result. For example, if the eye diagram result at 960's CMLOUT is not good, we need to find which part impacts the result. We can break the link into three parts, 935, cable and 960. If the eye diagram of 935 DOUT is good, then we can locate the root cause should be cable or 960. If the eye diagram of 935 DOUT is not good, then the issue should come from MIPI source, sensor or 935.

Thanks & Best Regards,

Sherry

Hi Sherry,

If you are going to make several measurements along the cable then feel free to compare them to the eye at the DOUT pins.  I feel at any length of cable away from the device it is going to be hard to get an eye but it may work.  Let me know the results you get.

Regards,

Nick