Optical components

Finns Claim Fiber Advance

Not far from the famous home of Sinter Klaas, deep in heart of Lapland, Finnish startup Liekki Oy is busily sintering something of its own -- and may well put something nice in the stockings of makers of Erbium Doped-Fiber Amplifiers (EDFAs) this Christmas.

Founded in 1999 by glass and ceramic materials expert Dr. Markku Rajala, Liekki looks like it could make a name for itself with a fiber-manufacturing process that permits much higher dopant concentrations to be incorporated into optical fiber than do conventional methods. The upshot is a fiber that could significantly lower the cost of making EDFAs.

Liekki's product ties in well with the industry's drive towards low-cost, compact amplifiers, particularly for metro applications (see Vendors Unveil Amplifier Advances).

This could be one reason the startup claims to be making good progress, despite the present weakness in demand for optical amplifiers (see Report: Slow Ramp for Optical Amps). Liekki's CEO, Terry Donovan, claims that "every major vendor of EDFAs" is testing its fiber. The company is sampling two types of erbium-doped fiber: one for C-band amplification (1525nm to 1565nm) and one for the L-band (1570nm to 1615nm).

The innovation behind the product is a process called "direct nanoparticle deposition," or DND. Raw materials are fed into a burner which, as the name suggests, results in the deposition of nanoparticles of glass and erbium. These are sintered to create a solid fiber preform. The advantages of this technique are that distribution of erbium inside the glass is exceptionally uniform and controllable.

This stands in contrast to the conventional fiber manufacturing method of solution-doping, which relies on diffusion -- a process that is difficult to control, resulting in an uneven distribution of erbium inside the fiber core.

The diffusion process is also slow, with the result that erbium ions tend to clump together in pairs. This leads to what's known as "handshaking" -- where erbium ions exchange energy without generating light, thereby limiting the maximum amount of amplification from the fiber. "If you have an uneven distribution [of erbium], you need to keep the overall distribution low to avoid handshaking," Donovan explains.

Since Liekki's DND process creates a highly uniform erbium distribution, this effect is avoided, allowing significantly more erbium to be incorporated into the fiber. As a result, much less fiber is needed to achieve the same degree of amplification as conventional erbium-doped fiber. "In the specification race of C-band fiber, you need four meters of our fiber, versus twenty meters of someone else's," Donovan claims, adding that Liekki aims to reduce C-band fiber down to one meter.

Shorter fibers lower costs in a couple of ways. For one, the cost of the fiber material is lower. Second, the assembly cost of the amplifier is lower because it's easier to pack in a shorter length of fiber.

DND also improves performance, Donovan claims. Good control over dopant distribution leads to perfectly round cores, which have lower polarization mode dispersion (PMD). Optical losses and Nonlinear Effects are also reduced in shorter fibers.

Donovan believes that Liekki is the only fiber vendor with an alternative to solution doping for fiber manufacturing -- a carefully worded sentence that may well turn out to be true.

That, however, doesn't mean the rest of the industry has ignored the fiber-doping issue. At least one other startup also claims to have developed a way of making highly-doped erbium fiber: Chuck Chandler, CEO of NP Photonics Inc., told Light Reading that his company's technique can be used to reduce fiber lengths from tens of meters down to centimeters -- a reduction of 100 times, compared to the one-third to one-fifth claimed by Liekki. NP's method is based on a new kind of host glass that doesn't allow erbium ions to clump together as they do in silicon.

A key differentiator, however, is that NP Photonics is building its own EDFAs, while Liekki sells fiber to the amplifier vendors.

Liekki has raised €5 million in funding to date, lead by Helsinki-based VC Stratos Ventures.

— Pauline Rigby, Senior Editor, Light Reading
mmmmm...forbidden fiber 12/5/2012 | 12:52:14 AM
re: Finns Claim Fiber Advance I think Mr. Donovan needs to do a little homework. While solution-doping is the most common method as far as number of companies using it, there are two other techniques (aside from DND) by which you can produce EDF that account for the majority of EDF in amps deployed today.

While short lengths may be appealing for reduction of volume in packaging of fiber they prevent challenges for splicing. If you get a bad splice and have to cut back your EDF, you've now lost a considerable amount of gain and may have to replace the entire coil.

Also Pauline, be a little careful. It's clustering in silica (an oxide, glass in this case), not silicon (a semiconductor). Two entirely different materials.

Hybrid 12/5/2012 | 12:28:27 AM
re: Finns Claim Fiber Advance Precisely correct.
The two other methods are homework for Pauline, but why use a messy solution techique if you can avoid it (unless it offers superior uniformity, which it doesn;t.
Hybrid 12/4/2012 | 9:09:00 PM
re: Finns Claim Fiber Advance Hogwash. Absolute BALDERDASH. Bah Humbug!

Solution doping is not the industry standard method for producing large wolumes of uniform, low cost EDF.

Smaller is indeed much better - the article missed that!

The cost of EDF is a miniscule fraction of the overall materials cost.

What a waste of Euros. There is still evidence of VC's flushin money down the toilet - perhaps we should all go cap in hand to Europe?

Merry Christmas! Bah Humbug!
Pauline Rigby 12/4/2012 | 9:08:59 PM
re: Finns Claim Fiber Advance Hybrid said (amongst other things):

>Solution doping is not the industry standard method for producing large wolumes of uniform, low cost EDF.

OK, so what is?

[email protected]
LightBeating 12/4/2012 | 9:08:58 PM
re: Finns Claim Fiber Advance Hybrid,

I don't agree. First of all, you don't know how much money was spent on that project, second I find it a valuable technical advance. Everybody wants shorter lengths of EDF in their amplifiers. I don't think the cost of the EDF is "minuscule". People want EDFA's below $1K, so any cost reduction is welcome. Of course, it remains to be seen if they will sell that fiber for four times the price of standard EDF...

zettabit 12/4/2012 | 9:08:57 PM
re: Finns Claim Fiber Advance Part of the benefit claimed by Liekki is not valid - this being that reducing Er-doped fiber length from 20m to 4m will save space and manufacturing cost.

This is because as anybody knowledgable of EDFA design knows, you must loop the fiber to fit within a reasonably sized package (we can't have 20m long EDFA gain blocks now, can we?). The diameter of this fiber loop is limited to a minimum bend radius defined by the fiber geometry, so as to prevent light loss from the fiber core and to minimize reliability-affecting curvature induced stresses on the fiber.

Therefore the only difference between a 4m and a 20m long Er-doped fiber is that the latter will need 5x more loops - HOWEVER the loop size is the same! Therefore there is no size reduction possible of the EDFA gain block.

And the contention that this will save assembly cost is bogus. The only savings will be needing to do 1/5 less looping of the fiber, which should save all of 2 minutes at most. Given Chinese assembly of many of these modules, and a $5/day labor rate, that should save a whole.... $0.01!

WOW! That is fantastic. Imagine saving 1 cents off every EDFA gain block!

Granted, higher Er concentrations may have other benefits in terms of optimizing EDFA gain profile & performance, but these have yet to be high-lighted.
Sign In