Test on-resistance of pin-diode

[ThermalRunaway]

Hi all,

Does anyone have suggestions with regard to testing the on-resistance of a pin diode at a set frequency and bias current?

Brian

 

[RadioRon]

The obvious choice for me would be to prepare a jig for attachment to a vector network analyzer. The jig would place the diode across the end of a 50 ohm transmission line in such a way that lead impedance is identical to how the diode would be mounted in its final application circuit. The analyzer would be set up to measure the impedance of the diode at the desired frequency.

The DC bias could be coupled into the jig through a suitable pair of RF chokes directly next to the diode, and of course a pair of DC blocking capacitors is needed to prevent DC from flowing back to the analyzer. The DC bias current would be metered and fed from a DC power supply set up at specific current limits.

The details of the jig design and choice of VNA depend a great deal on the frequency that you want to test at. The higher the frequency, the more fussy you must be about the jig layout.

 

[ThermalRunaway]

Thanks for that Ron.

The test you have described is similar to that which I have already put in place so I'm quite pleased with that. Unfortunately the results I am getting do not compare favourably with the datasheet (when the same test parameters are used). I think this might be due to the fact that I did not use D.C. blocking capacitors after applying the 10mA bias. Thanks for that suggestion, it would seem quite obvious now that you've pointed it out.

Brian

 

[ThermalRunaway]

By the way, I don't think I need to worry about matching the impedance to that of the final application circuit. The story behind this is that a number of test failures are occuring in an attenuator circuit and it is my opinion that the pin-diodes might be the cause. Swapping pin-diodes results in the attenuator circuit performing differently.

The reason for testing the pin diodes was to see if they were in spec with respect to the datasheet. If a spread of results could be obtained, the pin-diodes could then be arranged in batches of similar performance and each type could be subsituted into the final application circuit to see what effect it has on the performance of the attenuator.

The datasheet quotes an on-resistance of 0.6Ohms @100Mhz with a bias current of 10mA, and hence my reason for wanting to test the components with these parameters.

Thanks again Ron,

Brian

 

[RadioRon]

Brian, that on-resistance is quite low compared to the 50 ohm system you are using to measure it with. This means that even minor errors in your jig impedance or calibration could result in significant error in your resulting measurement. If you are just doing relative checks against a known good part I wouldn't worry so much, but for absolute accuracy you must be careful. At 100MHz things are reasonably forgiving (well, compared to 2GHz for example!) but keep your RF path lengths between tx line and diode terminal as short as you can and as high a common mode Z as you can.