DS250DF230: How to measure HEO

Yuta Nishie

Part Number: DS250DF230

Hi,

1) When measuring HEO, do you measure from the left end of the eye pattern? Or do you measure from the center of the eye pattern?

2) Please tell me the background of setting the default value of register address "0x6A" and "0x76" to "0x21."

Is it to maintain LOCK as much as possible even for poor quality input?

Best Regards,

Nishie

3 months ago

0 Drew Miller1 3 months ago

TI__Mastermind 20192 points

Hi Nishie-san,

1) My understanding is that measurement is started at the left side of the eye. Can you share your concern?

2) These are set to allow the retimer to lock to a signal even if the signal quality is not great.

Thanks,

Drew

0 Yuta Nishie 3 months ago in reply to Drew Miller1

Genius 3200 points

Hi Drew-san,

Thank you for your reply.

As shown in the following thread, a LOCK with an unintended data rate has occurred. To avoid this, we are considering countermeasures by changing the threshold of HEO/VEO.

So I wanted to check where the HEO measurement starts.

https://e2e.ti.com/support/interface-group/interface/f/interface-forum/1306638/ds250df230-method-for-discriminating-two-input-data/4974644

https://e2e.ti.com/support/interface-group/interface/f/interface-forum/1300746/ds250df230-behavior-when-signals-of-different-data-rates-are-input/4974639

Best Regards,

Nishie

+1 Drew Miller1 3 months ago in reply to Yuta Nishie

TI__Mastermind 20192 points

Hi Nishie-san,

Thanks for the details on this. I hope you're able to use the HEO/VEO lock criteria to filter out the false-lock condition. Let us know if you have additional questions.

Thanks,

Drew

0 Yuta Nishie 3 months ago in reply to Drew Miller1

Genius 3200 points

Hi Drew-san,

I have additional question.

I don’t think customer need to know the exact period but want to know approximate period. I suspect that, in order to measure HEO VEO, it needs to sample hundred thousand points. It may take several ten ms or more. What do you think of it?

Best Regards,

Nishie

+1 Drew Miller1 3 months ago in reply to Yuta Nishie

TI__Mastermind 20192 points

Hi Nishie-san,

Apologies for the delay. I'm looking to get more details on this, but I believe you are correct that it takes 10s of ms or more.

Thanks,

Drew

+1 Drew Miller1 3 months ago in reply to Drew Miller1

TI__Mastermind 20192 points

Hi Nishie-san,

Following up on this, the HEO/VEO measurement involves sweeping across the axes in the center of the eye and identifying the region without any hits. This is faster than a full 64x64 eye scan, but still takes some time. We'd estimate this takes in the 10s of ms. The amount of time spent at each voltage/phase position is shown in the table below.

Thanks,

Drew

Interface

Interface forum

DS250DF230: How to measure HEO