Hi --
I called APA a few months ago when I first saw their WDM receiver on the cover of FiberOptic Product News (a trade rag), so I am familiar with their offering. They have a 4 channel WDM receiver. I don't know how familiar you are with WDM, so if you already know about it, then skip the rest of this paragraph. In regular optical transmission over fiber, you have *one* laser operating on a single wavelength transmitting into the fiber, and at the output of the fiber you have a receiver (a photodiode) which absorbs the transmitter laser light and outputs an electrical signal proportional to the light intensity. In WDM, you have *many* lasers operating on *many* different wavelengths. They are all multiplexed together onto a single fiber with an optical multiplexer. After propagating down the fiber, the many different wavelengths are demultiplexed with an optical demultiplexer (i.e. the different wavelengths are spatially seperated), and then they are detected on different photodiodes, one for each transmitted wavelength.
The technology used for optical multiplexing or demultiplexing is often (but not always) a diffraction grating. The grating can be either bulk -- i.e. a standard diffraction grating like those found in spectrometers --, or it can be a photonic integrated circuit (PIC). In the case of APA, looking at the package and talking to the people on the phone leads me to suspect that they have a standard bulk diffraction grating in their box. After the grating, they have four photodiodes spatially offset from each other to receive the light. (They also have a lens to couple the light out of the fiber and collimate it for the grating, and probably have lenses for the photodiodes also.)
I imagine that you posted your question because you are interested in whether the APA product is useful for fiber communication. I shouldn't offer any *opinion* for a variety of reasons, but I will make a couple of technical observations about their product and the telco market for WDM devices:
1. State-of-the-art WDM systems vendors are busy readying systems having up to 40 +- 8 (i.e. from 32 to 48) wavelengths. Several vendors make PICs which demultiplex 32 channels, and the Japanese have published a device for 64 channels. The APA box is 4 channels.
2. The APA box's channel spacing is micrometer tunable from 1 nm to 4 nm. The last year, the ITU (the international telco standards body) defined the wavelengths to be used for WDM systems. The ITU wavelengths to which everybody must be compliant are exactly 0.8 nm apart. Some vendors make systems having 1.6nm channel spacing, but as the demand for bandwidth increases and more channels are added, people are migrating to 0.8 nm systems.
3. The micrometer tuning suggests that the device is fragile. To pass Bellcore qualifications, any device being considered for use in the telephone network must pass severe shock, vibration, thermal cycling, flammability, etc. tests. Can a mechanical assembly with micrometer screw tuning pass these tests?
4. Finally, if my suspicion that the APA box is a bulk diffration grating is correct, then the passband shape of the optical filter is not optimal. Like in any optical or RF system, you want your channel filter to have a flat top over the bandpass range, and steep walls outside. Bulk diffraction gratings tend to have rounded bandpass characteristics, which lead to more stringent requirements on the transmitter laser's wavelength stability.
Please note that these are my own technical judgements, and are not necessarily shared by any company which which I have ever been associated. Flames to /dev/null!!
Stuart |
