[802.3_OMEGA] Discussion about wavelength range

[Rubén Pérez-Aranda <rubenpda@xxxxxxxxx>]

Dear Colleagues,

In this mail, I would like to go in the detail about the question that Piers Dawe suggested to analyze: If we accept 840nm to 990nm wavelength range for TX and RX, 

which is the impact in the cost of the components and ecosystem?

As I stated today, I disagree with Ramana’s presentation in two points: the de-rating of failure rate and the extrapolation based on small sigma single populations.

But …. what if we assume 850nm is reliable enough, or if not, we decide to wait until we see what happen in the field with real devices … like we could do that.

First, let's start with comments:

• Comment I-107 intends to expand the TX wavelength range from 970 ~ 990 nm to 840 ~ 990 nm. 
• Comment I-108 intends to expand the RX wavelength range from 970 ~ 990 nm to 840 ~ 990 nm.

I will focus in a practical test. Let’s think how to run the SRS test, the one specified in 166.7.10, with the setup depicted in Figure 166–45. 

• Which set of wavelengths we should use?

With the current specification (970 ~ 990) I can argue that just using ONE reference transmitter is OK for testing the device (in P (process) V (voltage) T (temperature) 

characterization phase to get the design indexes for each parameter, in qualification tests, in final test, etc). Just one, like with  850nm receivers for 10BASE-SR, 25GBASE-SR, etc. 

But when I have to support  840 ~ 990 nm:

• How many different reference TXs do you think do we need in the lab? Just 850 and 980? 940nm too? Why not 880, 910, 950, 970 too? 
• As far as I know there are no tunable multimode  reference TX in the market in this range. I cannot talk about price/cost of these components for the lab. But I invite
  you to get quotation in the main T&M equipment suppliers to get the feeling of unitary price.
• How can I argue to the Tier1 and the OEM that just one wavelength is enough in the SRS test to validate my PHY implementation? 
• Should we write in our 802.3cz standard that just one wavelength is enough in the SRS test? 
• Which one and why? 
• Is it up to the implementor? How can we guarantee the interoperability then? 
• Is the Tier 1 and OEM going to accept single wavelength test report?  Consider that 20 nm width justification is not the same of 150 nm width justification.
• Why are we adding these complications to the standard?  
• Backward compatibility? — No, because automotive is a green field.
• Limitations of the fiber (CD, EMB, Att)? — No, because we are addressing 40 meters max length.
• More suppliers? Really? The yearly produced VCSEL in >= 940 nm is 2x orders of magnitude 850nm devices.

As you can see, a lot of questions not related with complex math of reliability. I appreciate that people participates in solving these questions, which are for the specific topic

of test. And I also invite people to make other questions. I think that if we find solutions to these questions and others, we can make better decisions. 

As I explained it is not just selecting all the wavelengths that are reliable or we can argue in one way or other that are going to be reliable. It is a more complex topic, specially reinforced 

in the scope of automotive application. 

Of course I recognized reliability as a very important topic from the beginning of the project. It is wrong to think in my opinion that optical multi gigabit PHYs will have more than 

ONE opportunity to success in automotive market. We cannot fail in the field. 

I appreciate participation of the group in the discussion.

Thanks

---

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