[Premium-Rx] 6790 filters

[FRANCIS CARCIA]

Dan,
Cubic did FET filter switching in the R3030 that works well.
I have a 6830 with 8kHz Roofers in both positions with dynamic range ill effects. Frank WA1GFZ


________________________________
From: Dan Rae <danrae at verizon.net>
To: premium-rx at mailman.qth.net 
Sent: Wednesday, August 22, 2012 6:16 PM
Subject: Re: [Premium-Rx] 6790 filters

Assuming that it is the 6790/GM that is under discussion here, there are 
a few "features" of this radio that are seemingly a bit odd, and since 
the designers of Racal radios at the time certainly knew what they were 
doing, one has to assume that these features were requested by the main 
customer, i.e. various branches of the US government.

  In the GM the degradation of ultimate selectivity by the filter 
switching is one thing, the choice of roofing filter width, the odd 
selection of tuning rates, the lack of memories are others and finally 
the removal of any provision for muting and the allied removal of any 
input protection as found in just about every other Racal receiver.  One 
has to assume that these were /all/ done at the request of the customer, 
as must have been the use of plug in filters and the BITE routine to 
detect them.

I have an RA1772 which has a switchable narrow (1.5 kHz width) roofing 
filter installed, switched by DC wetted sealed relay, so in what, the 
late seventies?  Racal knew the advantages of sometimes using a narrow 
roofer.    DC wetting as a technique in low level signal circuits has 
been known since early telecomms days., maybe even pre wireless.

For switching in a 50 Ohm environment now, there is a lot to be said for 
FET switching.  Many have used FET bus switches like the FST 3125 series 
with great success, see the work of PA3AKE for example, but DC wetted 
relays are still the ultimate.  I'm not sure what would be the best for 
the High Z environment of the GM IF filters, but FET switches would 
certainly work here and are compact.  For what it's worth there is no 
point in trying to use PIN diodes here at 455 kHz.

From past discussions with an acquaintance who spent a large part of 
his working life in front of intercept receivers, including his 
favourite, the RA6217, on land and aboard ship in various exotic parts 
of the world, it seems that ultimate skirt selectivity was not liked 
since it made it possible to miss signals like morse that were on / off 
keyed, while tuning.  Now having a home brew receiver that has very 
good IF DSP with better than crystal filter performance, I can 
understand why this was his preference, particularly since the filters 
in the 1217 series are only four pole.  I have no evidence that this was 
a customer choice therefore, but it does seem possible.  If then some 
reduction of skirt selectivity was not unwanted, the choice of filter 
switching methods follows.

Similarly for the roofing filter widths.  Since Racal US did supply some 
civilian versions of the GM with filters down to 8 kHz in width, I 
assume the usual 20 kHz pair in the GM was a request.    I know there 
was a government requirement for use with multichannel data signals up 
to 16 kHz in width, which does kind of require the 20 kHz roofers.

In the RA1792 by the way, the standard roofer width is 16 kHz, and there 
is only the one.  In the 1792 second mixer there is no second roofing 
filter but they did include a wide 455 kHz ceramic filter which Michael 
mentioned, replaced in the GM by a simple LC filter. Again, I suspect 
this width was required if only for the provision of a wide bandwidth 
for FM reception since narrower filters were available.  In these radios 
if one does alter the roofer width to a 6 or 8 kHz one then it becomes 
necessary to edit the eprom firmware to get the correct widths 
displayed.  Thanks to the brilliant detective efforts of Guido Keppers 
in years past I have been able to do this.

One point about using a narrower roofer, and I'm sure that there is 
nothing that would be hard about producing them somewhere in the far 
east in lieu of Bosnia, is that the rapid phase changes at the edges of 
the passband start to affect the mixer dramatically, and your IP3 will 
suffer.  This is why in the GMs that I have seen with narrow roofers, 
they still use a filter at the first mixer of 16 kHz width and put the 8 
kHz one in the second mixer card well removed from the first mixer.

Inrad, I'm sure would gladly have a batch made if there were demand enough.

Personally once the Eprom firmware has been altered to give tuning steps 
of 1, 10 and 100 Hz, I am happy to accept all the other quirks of the GM 
and like it well enough the way it is :^)

Dan

ac6ao / g3ncr




______________________________________________________________
Premium-Rx mailing list
Home: http://mailman.qth.net/mailman/listinfo/premium-rx
Help Page: http://mailman.qth.net/mmfaq.htm
Post: mailto:Premium-Rx at mailman.qth.net
Help Contact eMail:  paul at 8zo.com
Home Page:  http://www.premium-rx.org/