AnsweredAssumed Answered

Experienced w/ ADS @ 77 GHz

Question asked by STEINBUCH on Mar 7, 2004
Latest reply on Mar 8, 2004 by ALEPEREZ
I post this message to share my experiences with you:

I was faced with the task to realize a balanced mixer using a quadrature hybrid at 77 GHz. Simulation just the passive structure in Momentum and comparing that with my measurement results really puzzeled me ???.

The measueremts of the hybrid iteslf with the two diode ports open (w/o matching structure) and feeding the other two showed pretty good consistency:). Then I added the matching networks (just a short line, a stub and another, longer line to the diodes), and mesured that sturucture. The measurement result compared to the Momentum simulation showed significant differences! :'( The transmission (which should be a couple of dB since the ports were open) showed a dip at center frequency, which made the hybrid unusable. Couldn't explain why, did numerous tests, was stucked!

Then I got back to ADS....in my despair:P, I tried the thick metal expansion feature, since all my other ideas were gone. I had the simulation run over the weekend. Getting back to the results on Monday (today), I was amazed! The simulation results looked completely different compared to the simulations results w/o metal expansion. And intersting enough, the simulation was right on the measurement results! That made my week! Great!;D;D

The good thing is that I found means of optimizing the hybrid. The bad thing is that I don't know the root couse of that physical behaviour. Maybe somebody here has a clue....???

PS: When I cut off the stub, just having the 0.9 mm line to the diodes, the regular simulation was right on, just adding the stub (0.5 mm) caused that trouble in inconstistent results.

(Ro3003, 5 mil, t_line=40 µm)

Dirk Steinbuch, Robert Bosch GmbH  

Outcomes