NimbleSig III 230 MHz Low Pass Filter Network Analyzer Test Results



20090215

A photo of the connectorization to the PC board  used for the completion of these tests is shown below.

The above photo array illustrates swept frequency insertion loss tests done on four 230 MHz NS3 low pass filters. Note the markers are at 50 MHz intervals, the center frequency of the top row is 250 MHz and the bottom row 500 MHz. The band pass regions are very consistent from filter to filter as is the cutoff region. The reject band remains >60 dB throughout 500 MHz. The bottoms of the 350 MHz and 500 MHz trap responses are very deep and could be moved around by physically moving the test cables suggesting the shielding effectiveness limitations of the cables was a factor in the tests at the dynamic range extremidies.

I am very pleased with these results which I think speak for themselves. It would be nicer if the 500 MHz nulls were not shifted by up to 50 MHz from 500 MHz however this is rather difficult to do much about.  Considering the resonating capacitors for the 500 MHz traps are only 2pF  a less than 0.2pF change would be needed to tune them exactly to 500 MHz. Possibly a 1.8 pF value could be used with a gimmick added to tweek this null. However the critical nature of such a tuned circuit could be affected by the enclosure cover. If the existing 60 dB rejection of the reference clock is found insufficient for some applications then an external supplementary filter should be considered to provide a more reliable solution.

I think these test results confirm the recommended 230 MHz LPF design conforms nicely with the simulation and provides a sufficiently steep cutoff to effectively reject the reference clock LSB image.

This photo shows one of the two test  jigs used. The other test jig used BNC connector pig tails which we found were a bit too heavy to work with. These SMB pigtails which use 1/10" Teflon insulated coax were much more compatible to the tiny NS3 PC board.


73 de VA7TA