FEATURE : TEST & MEASUREMENT
Rapid passive chip alignment
g involves migrating Keysight’s years of test and measurement experience and expertise from
the optical transceiver market. Keysight’s current photonic test and
measurement products are at least 10 times faster and more accurate than its first designs, with ‘hundreds of patented new technologies’, Haisch added. ‘As complexity of PIC design and implementation varies, some only work on passive structures, while others bring in ring resonators and even lasers,’ he said. ‘Te test and measurement requirements also vary, and we have been supporting each individual requirement from each foundry and intellectual property designer.’
Self-testing on the roadmap? To clearly define the interface between device and test setup, Keysight has introduced calibration and de-embedding methods Haisch noted. Tis removes the test setup’s impact on a device’s characteristics. ‘Keysight’s photonic applications soſtware suite speeds and simplifies both the process of test setup and the optical measurements,’ Haisch said. Keysight also partners with probe station vendors like Formfactor. Such relationships help integrate probe stations with optical, electrical and radio-frequency probes, and cards with test instruments in an overall test solution, Haisch said. Finally, Keysight Pathwave Automation provides a platorm to define and realise test flows and communicate with test instruments, probe stations and databases. Tis overall solution was demonstrated live at the OFC conference in March. Approaches that test PICs through external
interfaces can be complemented by self- testing functions, Haisch added. ‘Designers work on incorporating self-test features onto PICs, to check functional operation, rather than
12 FiBRE SYSTEMS n Issue 25 n Autumn 2019
performance,’ he said. Approaches include optical loopback between transmiter and receiver, or optical monitors tapping off signals. ‘Since electronics are not, and potentially will never be, monolithically integrated with PICs – maybe with the exception of very simple electronic functions – self-testing is for testing at module level, rather than wafer-level testing,’ he said. Haisch added that this is due to greatly differing semiconductor technology for high speed electronics, which uses 7nm-scale lithography, and silicon photonics, which uses 90-130nm lithography.
In tune However, Latkowski is more optimistic about electronic-photonic integration, and outlined some self-testing options that might be possible. ‘Initially PICs could, for example, monitor basic properties such as temperature, output power or wavelength of a laser.’ He hoped that it would be possible to implement relatively simple feedback circuitry that could then take corrective action. For example, it could increase current injection into a laser, should it face a drop in output power. Alternatively, the circuitry could cut off the current injection should the chip overheat. Another option would be to tune a laser should its wavelength be driſting. ‘Next to that, if we could implement
advanced logic – for which even more support from digital electronics may be needed – PICs could monitor a wide range of parameters, and adequately react to those,’ said Latkowski. Options include recalibrating laser arrays or switching between high-performance or energy-saving modes, like systems-on-chips in smartphones, Latkowski highlighted. ‘Tis should be possible for PICs, when intimate electronic-photonic integration matures,’ and is
PIC CUSTOMERS HAVE BEEN REPORTING INCREASED VOLUME SHIPMENT OF PIC-RELATED TECHNOLOGY FOR THEIR DATACOM PRODUCTS IN THE LAST TWO YEARS
on the Integrated Photonic Systems Roadmap. In the meantime, Latkowski highlighted that
standardisation efforts must become widely recognised as successful. And if they are, they will first benefit prototyping, multi-project wafer (MPW) foundries and other low-volume production scenarios. ‘For large volumes, at a certain threshold
there will always be a need for customisation in order to optimise the cost,’ Latkowski said. But the two do not conflict, he observed. ‘Standards, even if used in small prototyping batches, should be defined in a way that supports scaling to volume.’ Tat’s because test and measurement equipment suppliers, PIC foundries, assembly and packaging service providers all want the same thing, Latkowski said. ‘We are all looking for a “killer application” that will drive the volumes.’ n
www.fibre-systems.com @fibresystemsmag
Google
ficonTEC
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44
