Optical components

JDS Goes Undersea

It appears TriQuint Semiconductor Inc. (Nasdaq: TQNT) has found one buyer, at least, in its quest to trim down its optical-components division.

A recent SEC filing from JDS Uniphase Corp. (Nasdaq: JDSU; Toronto: JDU) shows the company acquired TriQuint's undersea pump-laser packaging technology on May 6, for $6.6 million in cash. JDSU will account for the purchase in its fourth fiscal quarter, which ends in June.

Note that JDSU didn't acquire actual lasers, which would be a bit nonsensical, considering JDS and SDL Inc. had so much pump-laser power that they had to divest some of it during their merger (see Nortel Buys JDSU Plant for $2.5B and JDSU and SDL: The Saga Continues). Rather, JDSU wanted the packaging technology in order to develop entire pump modules, a JDSU spokesman says.

In addition to the technology, JDSU will pick up a couple of TriQuint's employees, the spokesman says.

A TriQuint spokeswoman wouldn't elaborate on the deal, but it's likely that the pump-laser packaging came from the acquisition of the optical components division of Agere Systems (NYSE: AGR.A) early this year. TriQuint was primarily interested in Agere's lasers and detectors, and the company has been seeking buyers for the other product lines (see TriQuint Closes Agere Acquisition and TriQuint Talking Deals at OFC).

Of course, the undersea business right now is, well, underwater (see Report: Undersea Cable Market Has Sunk and Elmic Shows IP Storage Stack). But the JDSU spokesman says the company wants to prepare for that legendary day when long-haul and undersea buildouts come back to life. In the undersea case, specifically, qualification takes much longer, and JDSU wants to be ready for any new contracts that surface.

— Craig Matsumoto, Senior Editor, Light Reading

Page 1 / 2   >   >>
redface 12/5/2012 | 12:02:11 AM
re: JDS Goes Undersea Could someone explain the requirements for undersea operation? Is the optical fiber cable immersed in sea water? Are optical components sealed in hermetic packages?

Dr.Q 12/5/2012 | 12:02:05 AM
re: JDS Goes Undersea "Could someone explain the requirements
for undersea operation? Is the optical
fiber cable immersed in sea water? Are
optical components sealed in hermetic

The principal requirements for undersea systems are reliability, reliability, and reliability. The cost and difficulty of repairing equipment that is under two miles of ocean is enormous in comparison to sending a service truck with a repair technician.
For example, the pump lasers used on the current EDFA systems are rated for 28 year lifetime.
The cable contains several optical fibers, a strength member, electrical supply, and very durable overcoat to protect it from sea water. The repeater station (about every 40 km) is roughly 30 inches long by 10 inch diameter. This houses the EDFAs and the control electronics. The active component in the repeater station are hermetically sealed, but some of the passive components do not need to be sealed. Each of the components and modules is tested at many points in the assembly process to make sure that it not only meets the requirements, but that there are no anomalies (which might lead to later failure). As one example of this regimen, before the lid was sealed on the pump laser module, it was examined under a microscope and a video was made of the inspection, which could later be viewed in case of anomalies in its performance. The repeaters station itself has heavy duty seals to keep out the water.
Once the repeater is down on the ocean floor, it is quite a benign environment. It is a constant temperature (~2 C for deep water, up to 38C for coastal areas), which avoids the stress of thermal cycling that terrestrial exterior plant equipment suffers.

Another strategy for undersea systems, called 'Coastal Stitching' puts the equipment on land, and only the optical cable under the water. The reach of these systems has increased substantially in the past 5 years, so going 120 or 200 km is quite feasible. This strategy can be used if the distance to be covered is relatively short (e.g. ATT's first EDFA undersea system was Israel-Cypres link in 1993). It can also be very economical in regions like coastal Africa, where there is a very high cost to running terrestrial cable (especially due to vandalism), but where the equipment ('Shore terminal') can reside in a secure facility near the shore, and a protected cable run a couple of kilometers into the ocean, then run up the coast for >100 km, finally coming ashore. The cost of laying the cable this way is very economical by comparison.

optical_rookie 12/5/2012 | 12:02:04 AM
re: JDS Goes Undersea Dr. Q,
Great summation on the requirements for submarine systems. Given your apparent knowledge on the topic, can you comment on the quality of the packaging technology being purchased by JDSU? I heard that one of the AT&T systems had some major failures due to pump lasers provided by Agere. Do you have any knowledge of this situation and, if so, please comment. Also, does other pump laser suppliers like Bookham, Corning or Furukawa have any product offerings in this space?

dwdm2 12/5/2012 | 12:02:03 AM
re: JDS Goes Undersea "...The repeater station (about every 40 km) is roughly 30 inches long by 10 inch diameter..."

I keep hearing that integrated planar optics with high power components can go more than 100 km without repeater. Can anybody please discuss which companies are working in this area and what products are available and is there any deployment with bigger distance?
Dr.Q 12/5/2012 | 12:01:58 AM
re: JDS Goes Undersea dwdm2 asks in post #4,
"I keep hearing that integrated planar
optics with high power components can
go more than 100 km without repeater.
Can anybody please discuss which companies
are working in this area and what products
are available and is there any deployment
with bigger distance?"

Part one of the answer is that good 10G transmitters that can go 120 km without repeater are available today.

Part two of the answer is that with high optical launched power (~+10 dBm) and with backward Raman pumping you can go >200 km without a repeater. If you wanted to build a trans-Atlantic system with amplifier spacing substantially larger than 40 km, you could do so. The reason for the (relatively) small amplifier spacing is system performance. The noise buildup is significantly better for a system with many low gain amplifiers closely spaced than for a system with a few high gain amplifiers widely spaced -- with the same total gain in each system. The choice of 40 km is a system tradeoff between cost, performance and reliability.

Part three of the answer is that conventional bulk optic devices can do optical powers well in excess of a watt. (Bell Labs researchers did hero experiments in the mid 90's where they got 100 Watts [that's right, Watts, not milliwatts] out of an 8 micron core fiber laser -- higher brightness than the surface of the sun.) Planar waveguide devices are somewhat limited in their optical power handling capability, especially devices like waveguide amplifiers (EDWAs)

For part four of the answer, there are a number of companies doing planar waveguide amplifiers, but there is no market yet. See for example,
- www.nktintegration.com/news/ EDWA_press_release_Feb_03.pdf
- www.inplanephotonics.com
- www.genoa.com/download/Technol...
- www.kigre.com/er117.doc
just to name a few players.

Dr.Q 12/5/2012 | 12:01:57 AM
re: JDS Goes Undersea optical_rookie asks
"...can you comment on the quality of the packaging technology being purchased by JDSU? "

It's a bit early to comment, as I haven't found out yet exactly what equipment/facilities/people was sold to JDSU. Making the reasonable assumption that it is the equipment, facilities, and process information that Lucent (later Agere, now TriQuint) had for subcable pump packaging in the late 1990's, then JDSU got a world class operation.

However, they would need to have at least the core of the engineering staff who know the what, how and why of the technology. The key thing with submarine components, as I mentioned earlier, is RELIABILITY, and being able to demonstrate and document the reliability. The hardware is almost an afterthought here. Without the people, it will take JDSU quite a lot of effort to get to submarine quality.

[Comment on the reliability levels.
Reliability level is typically expressed in terms of FIT's, which are Failure In Ten to the 9th hours. Submarine system customers require components that have reliability on the order of 1 to 10 FITs. That is 1 to 10 failures in 1,000,000,000 device hours (~114,000 device years). Imagine aging 114,000 widgets for a year to demonstrate that fewer than 10 of them fail. People don't actually DO that, but this illustrates the magnitude of work required to demonstrate that reliability level.)

To answer another part of the question, yes there have been field failures from the Lucent pump lasers in the mid 90's. I know it was resolved, but I don't know what the resolution was.

I'm not current on who else does submarine level lasers, but check the Nortel website for their 980 nm laser. SDL did a lot of reliability work on their 980 lasers in the late 90's, so they certainly have the laser chips to put into the packages.

dwdm2 12/5/2012 | 12:01:55 AM
re: JDS Goes Undersea "The choice of 40 km is a system tradeoff between cost, performance and reliability."

Thanks DrQ for the elaboration.

It seems like it is the usual inertia towards a newer technology that is the biggest barrier which is conveniently reasoned (by the present art) in the economics terms. I can see the convenience of resorting to the known and proven technology even if it costs more. However, the 40 km scheme is not all that reliable either!

Now, somewhere I saw that the repeaters cost in excess of $50k per pop (please correct my number!). So perhaps cost is not the biggest factor, because, reduction of the number of repeaters will result in proportional savings.

That leaves one with performance and reliability. I'd tend to pin things on reliability, because, performance is lot better in newer systems, like the 10G transmitters you mentioned. Now, even if we consider similar standard for both 40 km and >100 km systems, it is hard to see how 40 km scheme is more reliable than the other. I understand both will involve O-E conversion, BER checking, amplification and regen, etc., until true OOO becomes reality.

In fact reliability is one of the important issues (among many others) that drives the integrated approach. Not to mention smaller and cheaper, etc.

So if there is no significant market for the newer technology that seems to mean that folks who are already comfortable milking the 40 km cow, would discourage everybody else to even consider the next gen.

fiber_r_us 12/5/2012 | 12:01:53 AM
re: JDS Goes Undersea dwdm2, I think you missed DrQ's description. Current submarine systems are already all optical. Submarine systems are not "repeated" (which is an OEO process), they are optically amplified (via EDFAs and/or RAMAN pumps, which is completely optical). The reason the spacing is limited to around 40km is that the optical system performs better when you can use lower-power amplifiers with shorter spans. The OSNR is better this way. It is one of the fundamental optical things that lets you go trans-oceanic distances completely optically. In the grand scheme of things, having twice as many amplifiers in the system does not make it cost significantly more. The real cost is the overall system (cable, power plant, terminals, fiber, manufacturing, testing, and most of all, installation).

This is all mostly meaningless anyway as there is so much submarine capacity at this time that no one is looking to deploy anything new.
Milano 12/5/2012 | 12:01:53 AM
re: JDS Goes Undersea One more thing on JDS' shopping list: a vendor of passive components with the right packaging technology for sub applications. For that they are looking at ITF in Montreal for a couple of millions and bring the 5-6 people that really know something about packaging.

The government agency that invested in ITF last year is now the joke of the investment community for that move. The new government in Quebec is eager to move away from industrial management and selling to JDS may be the only way to get anything out of that deal.

Bongiorno 12/5/2012 | 12:01:28 AM
re: JDS Goes Undersea
The Zurich ex IBM ex JDSU ex Nortel now Bookham
chip fab was supplying the qualified chips for the Lucent/Agere packages and ultimatelyTyco cables. Interesting to know if they still are, but presumably JDSU will use the ex SDL chips which are already submarine qualified being used in the Alcatel soon to be Avanex submarine package and sunsequently Alcatel cable systems. Corning did have a submarine qualified pump in the ex Pirelli laser (package and chip) but that was shut down. KDD have I think with a link up with Mitsubishi for the chip (package?).
Then there is Furukawa but I'm afraid I don't know how they were doing.
Page 1 / 2   >   >>
Sign In