[Kenwood] TS-870s and roofing filter

[Steve Harrison]

At 10:20 PM 2/17/2007 +0100, SkunkWeed wrote:
>Hello.
>After some emailing with Inrad they said, that they can run serie of
>roofing filters for TS870 and similar.
>Tell me, is that worth installing inside? I want to improve receiver
>performance without loosing "hifi" sound.
>They said that filter will be 5KHz at 3dB wide.
>Can anyone post comment on that?
>Mike

Someone else responded that a sharp-skirted roofing filter would be better
than one with slowly-rolling-off skirts. According to the Sherwood site,
the slowly-rolling filter, which is typically a 4-pole filter, appeared to
be a better choice because it has less insertion loss. Insertion loss of
the filter is important because if too much, it requires addition of
amplification in order to avoid reducing the noise floor of the receiver.
The addition of amplification to compensate for the increased loss of a
narrower filter can also result in reducing the ability of the receiver's
first stages to handle strong signals and thus increase 3rd order IMD
products, negating the reason for adding the filter in the first place. All
of this is explained in an article published on the Inrad site explaining
the use of roofing filters, specifically the 3rd page (look for the frame
"Performance Enhancements" then the link to "A few words about roofing
filters" in the description of roofing filters for the yaesuuey FT-1000 MP).

I've been working the ARRL CW DX contest on 3.5 over this weekend, first
attempting to use my otherwise-neat TS-680S and then having to switch to my
venerable TS-130S (which has the Kenwood 250 Hz filter). The difference is
that the 680S was not only generating scads of IMD products even with the
attenuator in and RF gain turned down, but was also still being desensed by
several very strong stations, especially W3LPL. I'm in southwestern
Connecticut; LPL is in the WDC area. At one time last night while using the
130, they were so strong up here that they started generating IMD in my 130
even with the attenuator in and RF gain reduced to S9 on the meter. I've
just set up the S-meter over the past week and S9 represents around 50
microvolts at the antler port. The attenuator is 20 dB. LPL was obviously
squirting a BODACIOUS signal up this way at that time
==============================8-O  In comparision, none of the New England
stations ever came anywhere close, including KC1XX. Even K8AZ never caused
problems. It was just propagation and the type of antennas that all those
big-gun stations were using; LPL apparently just happened to be
transmitting on some antenna that had a very high radiation angle that
caused them to dump a high amount of RF energy down on New England  ;o((((((

The TS-680S has "balanced mixers", both before and after its pitiful
attempt at a first IF filter. The TS-130S (and the similar but earlier
TS-120S radios) have a single dual-gate MOSFET front end and mixer and a
similarly-pitiful first-IF filter. But the 120/130 front ends and 1st IF
stages don't exhibit nearly the sensitivity to overload that my 680 does.
I'm going have to do some measurements to determine what's compressing
first in the 680: the front end or the 1st mixer stage. Mesuspects it's the
front end rather than the balanced mixer FET stage. In this case, a roofing
filter will NOT help at all. If the same situation exists in the TS-870S,
then similarly, a roofing filter won't help because the front end, itself,
is overloading and generating the problems.

Steve, K0XP