Aiming to become the global leader in chip-scale photonic solutions by deploying Optical Interposer technology to enable the seamless integration of electronics and photonics for a broad range of vertical market applications

Free
Message: Patent: METHOD OF FORMING AN HERMETIC SEAL ON ELECTRONIC AND OPTOELECTRONIC PACKAGES

Wow! A very big question that I have been trying to figure out for a long time has just been answered. It is truly ingenious.

I have speculated on how it is that POET is able to achieve wafer level testing of the optical circuits without an access grating at the surface.

The following is from a recent post:

Silicon photonics uses grating couplers which can be placed at convenient locations within the SOI chip but they are very inefficient at capturing light. Silicon photonics requires light and electronics to coexist within the same die so it is important that fiber can connect at the surface of the wafer (in typical applications) within the real-estate that occupies the photonic waveguides.  

The article that I posted yesterday sums it up in a semi-understandable way:

7.4.1.2 Grating Coupler

A wafer scale testing is not possible with an edge coupler. Therefore, there is a need to develop vertical coupling mechanism… A grating coupler consists of periodic etch structures to diffract the light in a certain direction… the coupling efficiency is very poor for a grating coupler. 

There are three main contributors behind this high coupling loss: diffraction toward the substrate, mode mismatches between the fiber mode and diffracted mode, and second order Bragg diffraction.

 POET uses highly efficient spot size converters to couple light at the edge of the photonic chips so it would appear to me that in order to perform wafer scale testing grating couplers or similar access points are required to be in place for wafer level testing. No doubt associated with one of many trade secrets.

https://agoracom.com/ir/POETTechnologies/forums/discussion/topics/733707-autonomous-silicon-photonics-measurement-assistant/messages/2249464#message

So basically I thought they probably produced a surface grating to access the optical circuits which would just be for testing and would be sacrificed when the wafer was diced.

The ANSWER:

They have come up with an ingenious process whereby optical testing is achieved by energizing the laser using electrical probe points.

They do the complete end to end testing of the optical engine by using a looped back waveguide which is removed during dicing of the wafer.

So basically the whole optical engine is tested in this very unique approach that exercises all components of the engine in one step through just electrical injection in a way that emulates a complete end to end circuit. Both the transmit side and the receive side are tested without the need to connect fiber. The loopback temporarily acts as the fiber in both the transmit and receive circuits.

Figure 25 and 26

https://patentscope.wipo.int/search/en/detail.jsf?docId=US277541311&tab=DRAWINGS

 

 

Share
New Message
Please login to post a reply