40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  40m Log  Not logged in ELOG logo
Entry  Thu Jan 17 18:22:14 2008, Alberto, Update, Electronics, RF Monitor Band-pass Filter RF166Mhz.pngChebyshevb.pngChebyshev2b.png
    Reply  Fri Jan 18 11:53:50 2008, Alberto, Update, Electronics, RF Monitor Band-pass Filter Chebyshevb.png
Message ID: 248     Entry time: Fri Jan 18 11:53:50 2008     In reply to: 246
Author: Alberto 
Type: Update 
Category: Electronics 
Subject: RF Monitor Band-pass Filter 
The response is asymmetric and on the left side of the peak, we have at least 33dB within 33Mhz, which is enough for all the frequencies. We probably don't need an higher order filter but just low pass filters in series.

The spike at 320MHz doesn't depend on the circuit board. It's either the cables, their connection, or the splitters.

Note that the frequency of this test circuit has still to be tuned exactly at 166MHz (now it's 149).


Alberto



Quote:
After we finalized the schematic for the RF monitor board based on buffered LC resonators, on Richard Abbott's suggestion to avoid the complication brought in by the fast op-amps, we gave another chance to the a passive configuration of the band-pass filter based on a Chebyshev topology. Rich and Ben gave me an old but very powerful software tool to design that kind of filters and showed me the way to circumvent many hassles in making RF test boards.

I made a test circuit for the 166MHz line (see attached schematic), using tunable inductors. The TF are also attached.
We get more than 20 dB of isolation after 33MHz (with a loss of only few dB at the resonance - it could be less), which is enough for all the other frequencies (33,133,199 MHz) but we would like more for the 166. We are going to add one or two extra orders to the filter.

We also have to understand the spike at about 320Mhz and eventually somehow get rid of it.


Alberto
Attachment 1: Chebyshevb.png  11 kB  | Show | Show all
ELOG V3.1.3-