Inside an RF mixer and RF sampler

[EggertEnjoyer123]

I took apart a dead HP sampler module, as well as a 1 - 26.5 GHz mixer (MMDC-00126). The HP sampler has a weird issue where both sampling diodes are fine, but a capacitor seems to be blown, and is now showing 55 ohms resistance. The return loss and everything looks normal since the diodes are fine, but it doesn't sample well.

(Tip for anyone who wants to take apart the sampler modules: just heat them on a hot plate to around 150 degrees C and the epoxy holding everything together will become brittle. Then you can just use a screwdriver to pry apart the top and bottom covers.)

For the mixer, there are four diodes, and the mystery black component in the mixer appears to be a capacitor, according to my multimeter. My guess is that the value is small to allow the LO signal to easily go through, but not the lower frequency IF signal. The mixer converts the single ended signals to differential through the taper in the ground plane, which transitions the microstrip into (not exactly sure what it's called).

For the sampler, there are two diodes and probably two capacitors inside the package. I believe one of the capacitors is bad, since there is 55 ohms from one of the bias inputs to ground. (The diode is definitely fine because the resistance across it is 105 ohms). You can see that a microstrip to slotline transition is being used to act as the balun. There also should be a shorted stub to shorten the pulse from the step recovery diode, which is probably the weird piece of copper on the LO input.

[EggertEnjoyer123]

Here's the equivalent circuit for both devices.

[EggertEnjoyer123]

Someone wanted better pictures of the microstrip to slotline transition so here it is:

I verified with a multimeter and the microstrip input for the LO is shorted to ground by the strip of copper. Same for the two sides of the slotline.

[arlo_g]

Thank  you for posting those additional pictures. 

It's much easier to see what's going on at the transition now, and to see that your sampler is similar to a 5086-7699 that I had seen photos of elsewhere.  It looks like they wrapped copper mesh from the microstrip line on the bottom side of the substrate over to half of the slotline opposite the incoming coax strobe.  Some of HP's samplers terminate the strobe microstrip with a resistor to the (half) slotline, but evidently not this one.  Can you tell if the gray object at the top of your photo 00055 is ferrite or something like it? 

The sampler chip looks much like one described in the HP Journal in February 1986.

[EggertEnjoyer123]

The gray material seems to be microwave absorptive material. Seems rubbery.