A VCSEL Test Wish List
Alexander E. Braun, Senior Editor -- Semiconductor International, 8/1/2001
This means that these devices must be tested and characterized more quickly and accurately in the production environment. Testing for lasers, VCSELs, photodiodes and modulators has not received much attention, although there are many difficult technical hurdles to be overcome — and many are making themselves known each day. In the case of VCSELs, the devices have the attractive property of coming out on the top with the light (Figure). This means that standard stations work somewhat better with VCSELs because they are designed to come in with probes from the top to procure the necessary signal.
The overall measurement approach with VCSELs depends on the testing that manufacturers require. Many want simple power measurements that just require bringing two probes down to get a light intensity reading. There are several methods to couple the light out of the device. One is to directly couple with a fiber probe. This is adequate for spectral measurements and also for directly modulated laser light because it also provides modulation information. However, it is not satisfactory for power measurements. Bad coupling efficiencies with the fiber enable relative power measurements but not absolute power measurements, due to the beam's broadness and the fact that it is not always emitted in a perfectly vertical fashion. An additional concern is the amount of back-reflected light and the effect this may have on the laser's performance.
|
|
A platform is needed to get absolute power measurements and modulation information at the same time, without having to switch detectors. Obviously, the solution is some kind of hybrid system. One problem is that there must be a precise vertical placement of the fiber and extremely accurate control of its distance to the wafer to within 10 µm from its surface —±1 µm — to ensure that the cone angle of light emission falls within the fiber's core diameter. This is difficult to do with a vertical fiber. Also, probe stations generally have straight-down microscopes, making accurate positioning nightmarish.
Thermal control is very important to VCSEL characterization, especially for spectral measurements, because the laser's wavelength can shift with temperature as the cavity length expands or shrinks — many want to do measurements between 0 and 100°C. The need to do back-side measurements is also necessary. VCSELs have emission through the back facet of the laser cavity, and it is used as a feedback mechanism at the packaged part level to set the forward laser power. Many manufacturers need to measure the front-to-back emission ratio. The capability to make this measurement, which requires a transparent wafer holder while maintaining thermal consistency, is not yet available.
Presently, what is urgently needed is a system capable of performing spectral measurements and rf characterization measurements for directly modulated lasers, as well as absolute power measurements. Other capabilities would derive from all this, such as jitter and LIV tests. There are some VCSEL test systems on the market — Karl Suss has one, and Cascade Microtech's is under development; however, a tool capable of doing everything is probably a year away. It will take a modicum of R&D and plenty of good engineering to prove to the optical community that the solution works — and the solution will probably not fit well with the traditional semiconductor fab probe station because it will likely require a unique solution for probe alignment that will work both on and off the light emission axis.
For additional information on inspection, measurement and test, go to www.semiconductor.net/imt