[CQ-Contest] Collins Filters in Contest Rigs

[K3BU at aol.com]

In a message dated 98-08-11 11:27:17 EDT, K2AV writes:

<<  The Collins filters vary quite a bit. In the ft1000mp, the Collins 500
 hz filter (455 if) and (the CW soldered in) Yaesu 500 (8m if) cascaded
 are significantly less selective and less deeply skirted than the INRAD
 400 hz filters for both IF's cascaded. If you want to hear the
 difference, let me know and we can arrange a telephone call, and I can
 let you *listen* to the difference. 
    In the 160 meter cw test I got between stations up & down 500 hz that
 are 20 over s9, and the INRAD cascade renders them *inaudible* unless
 they have key clicks. The difference after I switched was stunning. Hard
 to believe, but I am afraid that some of these new Collins filters seem
 quite inferior to the Collins of old.
  >>

Same experience here, I tried Collins and Inrad filters in TS870 and Inrad
filters seem to work better. One difference is 400 vs. 500 bandwidth and it
also appears that design and packaging has changed. I was surprised by the
small size of late Collins filters. Maybe technology advanced and shrank the
components. The only way to test or compare is to do some plotting of the
bandpass curve for various filters in your radio and do some comparison. In
TS870 the DSP seems to skew the results, it contributes to the shape of filter
curve - DSP seems to "improve" it and mask the true bandpass. The best thing
is to measure the signal after the filter's IF buffer stage, before detectors
and DSP stages. (Or use the network analyzer for those who have access.)
Sometimes even the "same" filters have different characteristics,
manufacturing tollerances can play games here.
   The best CW filter I have seen so far is the Collins 300 Hz 455 kHz IF
crystal filter. Amazingly it has less attenuation than wider Kenwood and Inrad
filters and very sharp skirts.
    When considering replacement/addition of filters one has to look at
impedances too. In some cases buffer amplifiers or matching devices should be
implememnted and in case of cascading the same filters in particular IF,
buffer amp might be needed to boost the insertion losses from additional
filters. The most benefit is obtained from cascading sharp filters at the
closest possible IF stage to the front end. 
   For the real rock crush proof front end there 50 ohm custom impedance
filters available (INRAD) for custom design frequency. Great for those with
their "own" CQing frequency. I haven't used them yet, but heard that some JA
stations in metropolitan areas were using them.
    Generally, if one wants to improve the receiver, the first thing should be
to consider replacing the "stock" filters with better ones - gives the most
bang for the money. Looks like some manufacturers are leaving this door open
for upgrades by leaving enough room in the IF PC board stages area for extra
and larger filters (Yaesu, Icom).
   73 Yuri Blanarovich, K3BU, VE3BMV


--
CQ-Contest on WWW:        http://www.contesting.com/_cq-contest/
Administrative requests:  cq-contest-REQUEST at contesting.com