I have been trying to bias the photo diode(PD) with bias Tee (combination of inductor and capacitor) but the cut-off is coming out to be only up to 2 GHz. My PD works upto 4.5 GHz but because of my biasing, its performance is limiting up to 2 GHz. What is the best way to get the maximum performance out of my PD.
Each PD has a specified frequency range where the response is flat.Outside it tends to decrease as like the response of an amplifier.By making structural changes or by adding more layers,the bandwidth of PD can be enhanced i suppose....
Hi, Surya,
Please don't take offense at any of this: often the least likely possibility at extreme frequencies is guaranteed to be the one that is both the cause of problems, and overlooked.
Much of this is from personal experience.
a) make sure your test sources are reliable, calibrated, in-cycle, appropriate to the measurement, and actually producing the outputs they claim. Even in-cycle calibrated signal generators can produce misleading results if loaded improperly, recently moved (bumped, jostled, dropped) by someone else and not mentioned, or simply presumed that impact, ESD, etc haven't caused changes.
b) Be very suspicious of components which should be passive, and make jigs to test them before trusting them. As an example (admittedly decades ago) is a bag of AVX .01uF multi-layer ceramic capacitors which oscillated at 940Mhz when biased to 5v DC. Only half the bag did, and I didn't have the time to figure out how, but the spurious output was repeatable with all the power supplies in the lab, consistent frequency, unaffected by series resistance (even more spooky) and could not be overcome. AVX, of course, is a highly-respected manufacturer, from whom we should be able to expect components which operate within published specifications, and it is possible that the capacitors were merely responding to a local signal source with contact inductance forming a tuned tank...but then, others in the same lot didn't, and no other caps of the same format from other manufacturers did. For that project, I had to jig and test all components just to be sure I could trust them.
c) As always, layout and other equipment nearby can be having an effect. Additionally, connectors are crucial over 1Ghz, and should be both undamanged and clean. Solder joints are worth reflowing or completely removing and rebonding, whether by high-quality braid or sucker, and remake them with the appropriate solder and flux to ensure a joint which is both mechanically and electrically sound without gaps, bubbles, "cold solder" effects, etc.
d) You haven't mentioned which PD, what technology, materials, geometries, etc. If you have a datasheet with the schematic of their test setup, try to duplicate it and their claimed response. If it is lacking in detail, don't be afraid to call and ask for their technical support, and consult with them. Their purpose is to enable you to do research, not to just sell stuff, so it is worth their time to be sure that you are satisfied with their product, or at least understand why it isn't doing what it should.
e) if it is a device you have produced yourself, however, your mileage may very.
Thank you very much to all of you for responding. William Brohinsky thank you very much sir for sharing your experience. As you said the PDs we are testing are not working as they specified in the data sheet. Thanks again for remanding to cross verify the components and saving lot of time..
- Badges
- Science topic