DS90UB948-Q1: MAP with DS90UB947 all locations are green

Hi Selva, 

Can you run at default dwell time of 1000ms and see if results are more inline with expectations? 

Regards, 

Logan

I run it for 1000ms too. The result is the same. 

Thanks,

selva s

Hi Selva, 

Can you share more about the set-up as well? What link rate/PCLK condition are these tests performed under? 

Can you paste a picture of the Information tab as well?

Regards, 

Logan

Hi Logan,

Please find the information tab. Pixel clock 112MHz. The link rate is around 1.35Gbps per port we are using dual-port data transfer. 

Let me know what else you are looking for to debug this ?

Thanks,

Selva S

Hi Selva, 

The all green result is a bit suspicious/non-realistic. 

Can you verify that the 948 is not be pre-initialized or tied up by a host-controller on the SER side? 

I've seen this behavior before when the SER side I2C host is interfering with MAP tool. 

One other thing to verify is confirm you are using the latest ALP profile for 948. 

Regards, 

Logan

Hi Logan,

Thank you for your comments.

The all-green result is a bit suspicious/non-realistic.  - Yes agree with you 

Can you verify that the 948 is not be pre-initialized - Is this before of serializer? 

or tied up by a host-controller on the SER side? Are you talking about the USB2ANY connection? It is connected at the de-serializer side only (948).

I've seen this behavior before when the SER side I2C host is interfering with the MAP tool. (The host interface is in the reset condition, I monitored the I2C bus activity at the serializer side during MAP analysis there is no communication )

One other thing to verify is to confirm you are using the latest ALP profile for 948. (Yes, I am. I installed both software too.)

Thanks,

selva

Hi Selva, 

It is not unheard of to get all green, but generally that is when the link is operating at lower than maximum/upper-end of PCLKs. 

If possible, can you up the PCLK up to the max possible and see if you still get all green results? Since this is a margin analysis of link, its good to exercise worst case PCLK anyways. 

From what you are reporting, I don't see any obvious issue with the set-up. 

Regards, 

Logan

Hi Logan,

 when the link is operating at lower than maximum/upper-end of PCLKs. 

 What is the recommended value: Currently I used 113MHz and 140MHz. 

Thanks,

Selva S

Hi Selva, 

170MHz for 947 and 949 SERs. 

Regards, 

Logan

Hi Logan,

we disconnected the serializer i2c from the host ic , by removing 0 ohms and performed margin test.

still we are looking complete green.

is there any way to debug this apart from increasing the pixel clock.

thanks,

selva

Hi Selva, 

According to your reports/feedback on set-up, I can't think of anything else to check at the moment. Maybe its the case that your cable/system is low-loss enough at these PCLKs that results are all green. 

Regards, 

Logan