Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

[802.3_OMEGA] Ichiro Ogura: Reliability criteria: OMEGA original mission profile? Re: [802.3_OMEGA] Baseline test for IEEE 802.3cz PMD/MDI



All,

Allow me to ask additional questions about reliability in automobile.

As I and Kurata presented in Geneva mooting (Jan,2020), we take reliability issues how to address the random failures ( A study for highly-reliable optical transceiver based on Si Photonics technology).

The approach we take is redundancy and the Si-photonics chip can have built-in dual redundancy for laser light source with minimum additional cost.

Our philosophy for mission-critical applications is based on the idea that random failures should occur even if we are very careful in production and redundancy is mandatory in such systems. And of course we think automobile is the one.

What I continue asking is the goal we should have for reliable automobile PMD.
To be honest, I am in trouble to prepare my material for PMD comparison especially on cost and reliability.

I would very much appreciate any suggestions you have.

Thank you,

Ichiro



On 2021/03/15 12:07, Ichiro Ogura wrote:

Dear Ruben and all,

Thank you for your explanations.

Can I understand that the mission profile is based on the worst-case described in existing criteria authorized as LV124 and AC-Q100 and we don't have intention to create new standard on reliability test?

I will be fine if TF agree to set this as a goal for comparison of different technologies not only the case for VCSEL and get approval from OEMs.

We will check the model with Si-photonics.  Also, I ask my engineering team to follow the discussions and provide data to the TF.

Thank you,

Ichiro


On 2021/03/14 0:57, Rubén Pérez-Aranda wrote:
Dear Ichiro,

Let me explain the process.

During the study group I did a contribution (https://www.ieee802.org/3/OMEGA/public/nov_2019/perezaranda_OMEGA_05a_1119_VCSEL_Reliability.pdf
explaining the main concepts about reliability in general, the VCSEL reliability models and I used a mission profile in order to make a reliability assessment of the 
VCSEL device, necessary for the technical feasibility criteria. I used  some public papers and reports, but not a reliability model provided by a VCSEL manufacturer.
I took the mission profile from a use case that is not expected for multi-gigabit ports. The use case was battery management where chips are used not only 
during driving but also during battery charge.

Then, Roger King, affiliated with a VCSEL manufacturer company presented the contribution https://www.ieee802.org/3/cz/public/nov_2020/king_3cz_01_1120.pdf 
where he used the same original mission profile to examine how parameters like current density, lognormal scale and thermal resistance affects the wear-out  
reliability of a specific VCSEL device.

Then,  David Ortiz, presented in https://www.ieee802.org/3/cz/public/15_dec_2020/ortiz_3cz_01_151220_reliability_assesment.pdf a proposal more aligned with 
the expected use cases of OMEGA ports, LV124 and AC-Q100 grade 2. The new mission profile introduced two changes with respect to the original one:
  • It reduced the number of total operation hours from 32000 to 12000 h
  • It increased the ambient temperature of T2 from 50ºC to 70ºC 

Then, I was working with Roger to use the same reliability model in https://www.ieee802.org/3/cz/public/22_dec_2020/perezaranda_3cz_02a_221220_reliability_linkbdget.pdf
in order to make a link budget assessment with reliability constraints. In the slide 11, you can see that the max current density that can be used is 17.1 kA/cm^2 when
we consider the original mission profile, equivalent to 5.7 mA for a VCSEL with 6.5 um oxide aperture diameter. In the slide 12, you can see that the max current density that 
can be used is 19.7 kA/cm^2 when we consider the new mission profile, equivalent to 6.5 mA for a VCSEL with 6.5 um oxide aperture diameter. However, in slide 13 I proposed 
to use a more conservative current density of 15.2 kA/cm^2, equivalent to 5 mA in 6.5 um aperture diameter devices, in the link budget analysis presented from slides 15 to 20. 
As conclusion, the link reliability constrained link budget presented in this contribution supports the two mission profiles, the original and the new one, with margin. 

With respect to the temperature spectrum, in our experience the mission profile of a component depends on the specific location of the ECU in the car (it is different for roof-mounted components, for components in the passenger compartment, for components mounted in the engine compartment, …).
The mission profile that David Ortiz proposed in his contribution is a worst case for a grade 2 component (using LV124 and AEC-Q100, AEC-Q102 as a basis), so components that comply with that mission profile could be used in as many locations of the vehicle as possible.
I do not think that two different mission profiles might be proposed based on link length, since I think that both link lengths might interconnect ECUs on same class of locations in the vehicles. But as indicated in the presentation, there might be other mission profiles that could be analyzed, some one proposed by an OEM for example. But no other proposal have been received yet.

Best regards,


Rubén Pérez-Aranda 
CTO at KDPOF
_____________________________________________________________

Knowledge Development for POF, S.L.
A: Ronda de Poniente 14 2º CD, 28760, Tres Cantos (Madrid), Spain 
P: +34 91 804 33 87 Ext:110 
M: +34 689 319 866

El 13 mar 2021, a las 4:27, Ichiro Ogura <i-ogura@xxxxxxxxxxxx> escribió:

Dear Ruben and all,

Thank you for your information.

I just thought we have no consensus about reliability as we continue the proposal about it.

My question is about the change in mission profiles I have been informed to work for my reliability study in TF shown below;


Service time =15 years

Operation total time = 32000 hours mission Profile presented in Juneva meeting Jan.2020
(Ruben)
New mission profile in Dec. 2020 ad hoc
(Ortiz)

Ambient Distribution(%) Time (h) Time (h)
T0 -40 6 1920 720
T1 23 20 6400 2400
T2 50 65 20800 7800
T3 100 8 2560 960
T4 105 1 320 120


100 32000

I think the goal is not given even if the reliability is the most important issue in this TF.

From my technological background, the TF had better to ask OEMs to help the decision making about reliability that is not based on proposal-basis from suppliers.

If you allow additional question, 40m specification will need different reliability discussion for different application like bus and trucks that operate longer than passenger vehicle.

My apology again for asking primitive questions as  I have no background here.

Thank you,

Ichiro


On 2021/03/12 23:34, Rubén Pérez-Aranda wrote:
Dear Ichiro and all,

The contribution cited below does a mission profile proposal in slide 15 based on LV124 and AEC-Q100
attending to most reasonable use case expected for the OMEGA ports. According to my experience it is
a very good starting point for comparison. The number of hours correlates very well (even by excess) 
with the ones typically used by OEMs for applications that are active during driving hours. 

The reliability assessment example of slide 11 corresponds to an IC used in BMS. In this case the number 
of hours was higher, because the ICs were used also during battery charge, not only driving.

As I commented in our last meeting, although well argued and supported, the slide 15 of cited contribution is just 
a proposal of mission profile . If someone in the task force with more visibility, e.g. affiliated with OEMs and Tier 1s, 
considers that this proposal is relaxed for the expected use cases, please, I am really open to change and  make it harder.
Argumentation behind, of course, is needed.

From your comment "The new mission profile was somewhat relaxed from conventional one to meet VCSEL 
reliability estimation.” I can understand that you have a different proposal for a more stringent reliability mission profile
that is more conventional and that in your opinion should be used in the OMEGA (and in other automotive PHY project). 
Please, do a contribution with such proposal and let me know why the current proposal is relaxed with respect the most 
conventional ones used in automotive.

And no, we cannot decide whatever we like. Please, read AEC-Q100, AEC-Q200 (they are public) and LV124. Perhaps
you may also get access to mission profiles of Japanese OEM’s. After reading them, your conclusion will be that 
reasonable.
 
Best regards,


Rubén Pérez-Aranda 
CTO at KDPOF
_____________________________________________________________

Knowledge Development for POF, S.L.
A: Ronda de Poniente 14 2º CD, 28760, Tres Cantos (Madrid), Spain 
P: +34 91 804 33 87 Ext:110 
M: +34 689 319 866





El 2021 mar12, a las 12:57, Ichiro Ogura <i-ogura@xxxxxxxxxxxx> escribió:

Dear Ruben and all,

Yes my question had been raised after reading the presentation, that it is OMEGA original and is not authorized.

The new mission profile was somewhat relaxed from conventional one to meet VCSEL reliability estimation.

My question is the target is not given and we can decide whatever we like.

I had Telecordia as given target for comparison in my datacenter and telecom devices development.

Anyway we will decide the basis of reliability before we make decision.

Thank you,

Ichiro

On 2021/03/12 18:49, Rubén Pérez-Aranda wrote:
Dear Ichiro and all,

Mission profiles for reliability are not standardized in 802.3. In general terms, in the automotive industry
each use case in an OEM requires different mission profile (e.g. it is not the same the mission profile of
an IC embedded within electronic injection unit, an IC in BMS, an IC in a cockpit display, …) because 
number of hours and ambient temperatures are in general different.

Qualification and reliability assessment in automotive is a complex topic.

I recommend you to read the contribution 

In this contribution you can find:
  • Introduction to basic basic reliability approach identifying its limitations
  • Introduction to advanced reliability assessment including why it is necessary
    • Very important to identify most of the failure modes of the devices and base technology
  • Introduction to mission profile and how it is defined
  • A mission profile that David (author) proposed as reasonably applicable to OMEGA components.

Best regards,

Rubén Pérez-Aranda 
CTO at KDPOF
_____________________________________________________________

Knowledge Development for POF, S.L.
A: Ronda de Poniente 14 2º CD, 28760, Tres Cantos (Madrid), Spain 
P: +34 91 804 33 87 Ext:110 
M: +34 689 319 866





El 2021 mar12, a las 3:25, Ichiro Ogura <i-ogura@xxxxxxxxxxxx> escribió:

Ruben and all,

Thank you for your attention and comment for my Si-photonics presentation on 10th March.

I have a general question about reliability for my next step.

Will we have a authorized mission profile to make decision or will IEEE802.3 create new test standard for automobile phy?

Let us fix the basis of discussion for the target. My apology that I am quite stranger in this field and my data seem an overkill, though.

Thank you,

Ichiro  

On 2021/03/06 0:11, Ichiro Ogura wrote:

Hi Steve,

I would like to contribute a PMD matrix for discussion as attached.

As the wavelength is important choice of PMD,  I  would like to propose to leave it open and let wavelength options for more discussion.

Particularly, for the table 300-15 in your proposal, I would like to add wavelength options as 850/980/1310nm in page 1 of my file. I add some discussion points in page 2.

I think we'd better to start the D1.1 with wavelength options and keep discussion among TF members to leach consensus.

Let us discuss next week.

Let me know your suggestions.

Thank you,

Ichiro

Ichiro Ogura, PETRA

   


On 2021/03/02 11:42, Ichiro Ogura wrote:

Steve and all,
Thank you for your material for baseline proposal.

I would like to propose the baseline proposal to include all possible PMDs that leads open discussion to build consensus.

As I presented on 26th Jan., PETRA and AIO Core are working on a baseline proposal based on Si-photonics.
Si-photonics at 1310nm has advantages in high temperature operation and reliability typically needed in automobile applications.

I will prepare our text before 5th March that will be based on;
wavelength: 1310nm
Media: OM2(500MHz-km)/ OM3,OM4(effective BW 500MHz-km)/1310nm optimized MMF(2000MHz-km)
Power budget : TBD

I appreciate your consideration.

Thank you,
Ichiro

Ichiro Ogura, PETRA
    

On 2021/02/24 0:49, Swanson, Steven E wrote:
All,
 
On the IEEE 802.3cz call today, I provided a contribution on baseline text for the PMD/MDI (attached). I would like to make a motion to adopt this text as part of the baseline text at our next interim meeting as we move to D1.0.
 
The purpose of this note is twofold:
 
  1. Can I ask you to review the contribution and let me know if you have any comments to improve the text?
  2. Can I ask you to review the contribution and let me know if I can add you as a supporter for the contribution?
 
I would like to get your input prior to Friday, March 5 so I can make any necessary changes and submit for consideration by IEEE 802.3cz at our March 16 interim meeting.
 
Thanks,
 
Steve
 
Steven E. Swanson
Senior Standards Manager
Distinguished Associate
 
Global Technology & Industry Standards
MT&E
Corning Optical Communications
4200 Corning Place
Charlotte, NC 28216
 
m 607-725-1129
 
 
Standards are a bridge between markets and technologies; whoever controls the bridge controls the future…

To unsubscribe from the STDS-802-3-OMEGA list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-OMEGA&A=1


ウイルス フリー。 www.avast.com

To unsubscribe from the STDS-802-3-OMEGA list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-OMEGA&A=1



To unsubscribe from the STDS-802-3-OMEGA list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-OMEGA&A=1


To unsubscribe from the STDS-802-3-OMEGA list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-OMEGA&A=1




To unsubscribe from the STDS-802-3-OMEGA list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-OMEGA&A=1


To unsubscribe from the STDS-802-3-OMEGA list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-OMEGA&A=1