DS110DF410: Un-lock phenomenon depends on specific temperature

Takumi Suzuki1

Part Number: DS110DF410

Hi,
I would like to ask you some questions about DS110DF410SQ as below.

My customer has found the phenomenon that the device has "un-locked" when the temperature is changed.
The configuration of their circuit is following.

Optical Signal OUT : Framer ---> CH0(CDR) IN --- CH0(CDR) OUT ---> SFP+ ---> BER Tester
Optical Signal IN : Framer <--- CH1(CDR) OUT --- CH1(CDR) IN <--- SFP+ <--- BER Tester

The un-lock condition has happened at output signal path "Framer ---> CH0 IN" (0ch input).
And the un-lock has happened in case specific temperature change around "+20 deg - +17 deg".
(During the un-locked condition, BER tester has showed "LOF".)

And the "un-lock" condition is automatically recovered to "lock" condition under the following conditions.

1) Temp chaned under the specific temp range (under +17 deg)
2) Ch Reg 0x0A[3:2] = "11" ---> "00" (executed adaptation of Ch0)
3) After taking out and put in the receive cable of "SFP+ <--- BER Tester" path, the Ch0 "un-locked" is changed to "locked".
4) In case of setting Ch0 is "Bypass mode", the "un-locked" phenomenon has not happened.

Questions
Q1) Will you give me a comment or advice about this phenomenon?
Q2) About recover case of "3", in this case, Ch0 condition(signal out side) has been changed by Ch1 side (reciver side) condition.
Will you tell me your idea about the root cause of this phenomenon?
Q3) This phenomenon has happened at 2 devices (both are different lot number).
Do you know if this kind of phenomenon depends on "lot" number?
Q4) I attached CDR register setting file.
Will you check it if there are some mistakes?

Thank you for your support.
Best Regards,
Takumi

CDR setting.xls

over 2 years ago

0 Takumi Suzuki1 over 2 years ago

Mastermind 8480 points

Hi,

Let me tell you additional information about my question as below.

About the configuration, "Framer --- CDR(DS110DF410) --- SFP+" path is "one board" configuration.

About "SFP+ --- BER Tester", this path is connected by optical cable.

Optical Signal OUT : Framer ---> CH0(CDR) IN --- CH0(CDR) OUT ---> SFP+ -- optical cable --> STM64 BER Tester
Optical Signal IN : Framer <--- CH1(CDR) OUT --- CH1(CDR) IN <--- SFP+ <-- optical cable -- STM64 BER Tester

※ SFP+ : 10GBASE-LR

CDR has detected "un-lock" condition at "Framer ---> CH0(CDR) IN" path.

Thank you in advance.

Takumi

0 Rodrigo Natal over 2 years ago in reply to Takumi Suzuki1

TI__Mastermind 19165 points

Hi,

I have some questions to help debug,

Could you clarify the retimer operating temperature both at the start of the test as well as when the CDR unlocked behavior is observed?
Would you be able to provide the values for the retimer channel registers below both for the initial temp as well as the temp where issue is observed?
- 0x02, 0x03, 0x27, 0x28

Thanks,

Rodrigo

0 Takumi Suzuki1 over 2 years ago in reply to Rodrigo Natal

Mastermind 8480 points

Hi Rodrigo-san,

Thank you for your reply !

Let me answer your question No1 and about No2, I will ask the customer to get.

The start of the test temperature is 30℃ and then the temperature has been reduced linearly to 10℃.

The phenomenon happened around 20℃ - 17℃.

As I have some other doucuments to show you from the customer, will you tell me your e-mail address, if possible?

Thank you for your support.

Best Regards,

Takumi

0 Rodrigo Natal over 2 years ago in reply to Takumi Suzuki1

TI__Mastermind 19165 points

Thanks. This temperature delta is small relative to what the retimer can support. Thus it is very unusual that the customer is observing retimer CDR unlock for this test case. A retimer registers dump for the problem condition would be useful for debug.

You may provide your work email address and I can message you directly.

Regards,

Rodrigo