[Collins] 75S3B info

[Gerald]

On Sat, 2005-06-25 at 20:43 +0700, Martin Sole wrote:
> I think Glen raises an interesting point. Do they last forever? Disclaimer:
> I have no hard laboratory spec measurements to back this up but...
> I have tried a number of S line SSB filters in my 75S3B. For the most part
> they all 'sound' the same but some do 'sound' different. Unmatched (see I
> said it wasn't very scientific) tests with a tracking generator and analyzer
> show different pictures of the pass band for different filters. Not to say
> they are not all in spec but just wondering if age might play a part.
> 
> Signed
> Curious
> 
> 
> 73
> Martin, HS0ZED

The filter element is made of resonant disks with coupling wires spot
welded. Then at each end there is a magnetic transducer to convert the
current in the coil to motion of a coupling wire or resonator. Due to
microscopic lack of cleanliness or low line voltage at the instant of
welding a spot weld can be weak, then normal operation added to shocks
of shipping and the environment the filter is used in can lead to breaks
at welds and changes in the response curves.

That said, its very easy to achieve odd looking response curve
measurements. The filter designs of the S-line mechanical filters were
very prone to ringing when hit by pulses. Long ago I modified my 75S-3B
to have an IF output jack which I often connected to a 5 MHz bandwidth
scope. When power line noise became an issue with a 2m converter, I
looked at the IF output. With the AM selectivity supplied by
transformers the noise was narrow time spikes 360 times a second (three
phase 60 Hz power lines), though I could shift the phase (while
triggering the scope with the local line) by turning the 2m array to see
a different power line. Typical substation transformers shift the phase
30 degrees and I could see that. Digression aside, when I switched to
the 2.1 KHz SSB filter there was no time modulation of the noise, the
envelope was constant. The ringing was longer than the time between
noise pulses. This means that sweep rate is very significant. The filter
both rings and has delay. The sweep speed of the analysis circuit has to
be reduced until the detected response curve stops changing, and then
I'd reduce the sweep speed another factor of 10 before taking data for
record. Without slowing the sweep so drastically, one will get a
different response curve for the same filter just from the speed of the
sweep.

And filters may have different responses depending on exactly where the
coupling wire is welded to each resonator.
>  
> 
> -----Original Message-----
> From: collins-bounces at mailman.qth.net
> [mailto:collins-bounces at mailman.qth.net] On Behalf Of Glen Zook
> Sent: 25 June 2005 08:47
> To: David Thompson; collins at mailman.qth.net
> Subject: Re: [Collins] 75S3B info
> 
> Collins filters use air and not any type of foam. 
> Therefore you don't have to worry about the foam disintegrating.  Unless
> they are dropped, Collins mechanical filters normally last "forever".
> 
> Glen, K9STH

When cared for some Collins mechanical filters may last forever, but
some will fail from poor workmanship and from shock and vibration or
electrical abuse such as the leaky coupling capacitor in the R390A. One
early (prototype) KWS-1 transmitter filter I have failed and I found,
when I opened it, that the filter was supported on rubber grommets that
didn't take the heat of the filter opening well, rubber parts might not
have an infinite life, even in a closed and protected environment such
as the sealed filter case. I have less confidence that the plastic
filters used in the later S-line are that well sealed.

> 
> 
> --- David  Thompson <thompson at mindspring.com> wrote:
> 
> Don't know about collins but I had a KW 2000E with a Kokusai 455 filter. The
> KW 2000 series was the RAF/RCAF version of the KWM2.
> 
> Glen, K9STH
> 
> Web sites
> 
> http://home.comcast.net/~k9sth
> http://home.comcast.net/~zcomco
> 

-- 
73, Jerry, K0CQ, Technical Advisor to the CRA
All content copyright Dr. Gerald N. Johnson, electrical engineer