Dear Cory,

        In my experience, on-wafer probes can be made to work quite nicely with TRL standards printed on PCB. I would consider establishing your first-tier electrical reference plane right in a microstrip line on the PCB. This will allow you to connect your TO can directly to the PCB in the same way you do in production, and determine its parasitics in the environment you intend to use it in. The best way to do this generally is to use the same transmission line in your TRL calibration as you use to connect to the TO can in your circuit, if that is possible.

Good luck,

Dylan

At 07:42 PM 2/14/2007, you wrote:

Hi,

I am trying to measure and model a comunication laser diode (TOSA) in
a TO can. In this case it is a 1550nm wavelength DFB laser intended
for 2.5Gbps. It will be used with a dual-loop laser driver, and we
are primarily interested in the monitor photodiode response rather
than the optical response.

Since I don't have an optical VNA, I am just doing electrical
measurements from the laser diode input to the photodiode output with
a standard 2-port VNA (8510C or 8753E). The TOSA has three
terminals: laser cathode, photodiode anode, common laser anode and
photodiode cathode. I am currently using a homemade fixture composed
of two SMA PCB connectors soldered end to end, with the center pins
shortened to allow the insertion of the TOSA. I am doing a coaxial
adapter removal calibration on the VNA, and I have created additional
fixtures in open, short, and thru configurations in order to develop
a model for the fixture.

I am having a couple of issues with this approach. First, the TOSA
is normally mounted on the edge of a PCB, with the laser and common
nodes connected on top of the PCB and the photodiode node connected
to the bottom of the PCB. This is obviously a different environment
than my fixture, so there will be some discrepancy. Second, in order
to connect the common lead I either have to bend the lead to reach
the fixture or add a short piece of wire. The inductance in the
ground lead significantly affects the results of the coupling between
input and output.

I am interested in obtaining a better fixture, but I don't know if
there is something commercially availible that would work well in
this situation. My other option is to design a PCB fixture myself
(with help from our PCB group). Please let me know if you have any
suggestions or comments.

Thanks,
Cory

-----------------------------------------------------
Dylan Williams               <dylan@...>
NIST 818.01                  <dylan@...>
325 Broadway                 [+1] (303)497-3138 (TEL)
Boulder CO 80305 USA         [+1] (303)497-3970 (FAX)
http://www.boulder.nist.gov/dylan/
-----------------------------------------------------