I am experimenting with some Murata Ceramic IF filters.
I have several that I salvaged from some old CBs.
They are all 455Is. I have been trying to stack them
in series. Given the loss figures I have used a simple
JEFT buffer between filter. While this kludge does work,
it doesn't work very well.

Since you design radios you have likely already invented
this wheel. Is it practical to even think about placing
ceramic IF filters in series to improve the skirt?

I have an R2000 with an Irnad 2.1KHz filter that is great.
Since I can't spring $185 at the moment for my other R2000,
I am trying to get the best performance that I can for the
least cash outlay.

Feel free to reply to
I normally only check that account enough to keep it alvie.
What with all the ads for enhancements etc it is too much to
routinly wade through.

Terry

Murata filters are not known for good shape factor(6db//60db) . When
stacked in series losses mount rapidly. And they seem to have a lot
more noise when compared to crystal or mechanical IF filters. Have you
tried to build a crystal ladder filter? Very effective and MUCH cheaper
than any other IF badpass filter ,and Inrad is not the costliest-
KVG,Collins etc.may set one back many hundreds of dollars...

Craig at Kiwa Electronics seems to have figured out how to make quality
filter modules from series-cascaded Murata ceramics. He uses low noise
buffer amps as part of the package in Kiwa's Standard and Premium Filter
Modules (PFMs). The latter are particularly good, with ultimate rejection
figures exceeding Collins mechanical filters and better shape factors than
quality crystal filters... and at a very good price for the performance.

I've used a number of PFMs over the years in various receivers, and they've
performed excellently for DXing and program listening.

http://www.kiwa.com/kiwa455.html

Guy Atkins
Puyallup, WA USA


"Arthur Pozner" wrote in message
...
Murata filters are not known for good shape factor(6db//60db) . When
stacked in series losses mount rapidly. And they seem to have a lot
more noise when compared to crystal or mechanical IF filters. Have you
tried to build a crystal ladder filter? Very effective and MUCH cheaper
than any other IF badpass filter ,and Inrad is not the costliest-
KVG,Collins etc.may set one back many hundreds of dollars...

The Rockwell-Collins Torsional Mode mechanical filters are right around 90
dollars, shipped. This is the low-cost
526-8695-XXX series that many of the manufacturers are selling as optional
filters for 120 dollars and up.
If you do look for one of the old coffin style mechanical filters, people
are selling those items for very silly prices so I would avoid those. A
disadvantage of the older style vs the Torsional Mode units is that the
older style filters require terminating capacitors at the inputs and
outputs. The Torsional Mode units don't require any external terminations,
as long as the stray capacitance on the PC board and its switching is below
30pF. These filters also have an Input/Output Z of 2k, which means that you
can replace one of the Murata ceramic ladder filters without any
modification to the existing circuit.
Buy them directly from Rockwell Filter Products Division, Costa Mesa,
California.
Lee Cornett is the sales person that you want to speak with.
As far as the buffer amplifier, what kind are you using? Source follower?
Common source with or without bypass capacitor?
If you use a JFET as a common source, you can use it is self-biased mode.
Use a 2k gate resistor to ground. The drain resistor should be 6.8k and the
source resistor should be 2.4k. This will give a Thevenin resistance of
around 1.7k...........close enough for matching the I/O characteristics of
the filter. Remember, the input capacitance of the JFET will be around 8pF,
but this will have minimal effect on the filter response.
Arthur is correct about that shape factor; crystal and mechanical filters
will have a much better characteristic.

Pete

"Arthur Pozner" wrote in message
...
Murata filters are not known for good shape factor(6db//60db) . When
stacked in series losses mount rapidly. And they seem to have a lot
more noise when compared to crystal or mechanical IF filters. Have you
tried to build a crystal ladder filter? Very effective and MUCH cheaper
than any other IF badpass filter ,and Inrad is not the costliest-
KVG,Collins etc.may set one back many hundreds of dollars...

Craig's filters are very good, with good ultimate rejection, but Mechanical
Filters have an ultimate rejection of 120dB. True, this figure is better
than the dynamic range of most receivers, and also true, you have to have
very good shielding/ground in your PC board layout scheme in order to obtain
this 120dB figure.
100dB ultimate rejection is a good figure to shoot for. I have spoken to
Craig a few times.....he is a very good designer.

Pete

"Guy Atkins" wrote in message
...
Craig at Kiwa Electronics seems to have figured out how to make quality
filter modules from series-cascaded Murata ceramics. He uses low noise
buffer amps as part of the package in Kiwa's Standard and Premium Filter
Modules (PFMs). The latter are particularly good, with ultimate rejection
figures exceeding Collins mechanical filters and better shape factors than
quality crystal filters... and at a very good price for the performance.

I've used a number of PFMs over the years in various receivers, and
they've performed excellently for DXing and program listening.

http://www.kiwa.com/kiwa455.html

Guy Atkins
Puyallup, WA USA


"Arthur Pozner" wrote in message
...
Murata filters are not known for good shape factor(6db//60db) . When
stacked in series losses mount rapidly. And they seem to have a lot
more noise when compared to crystal or mechanical IF filters. Have you
tried to build a crystal ladder filter? Very effective and MUCH cheaper
than any other IF badpass filter ,and Inrad is not the costliest-
KVG,Collins etc.may set one back many hundreds of dollars...

455KHz resonators are commonly used in remote controls. Have not tried
building a lattice or a ladder filter with them yet. But, what if
it just might be succesful-- that's why experimenting is fun.
Collins mechanical filters do have very good rejection just as Pete has
stated, in fact, at 455KHz it might as well have better
selectivity/rejection/low distortion/phase linearity than any other
filter type. I believe R-2000 is a triple conversion receiver, has
anybody ever try changing out the first (roofing) filter.Just for the
sake of experimenting.

Hi Pete,

Craig Siegenthaler and I have been friends since 1988, so we go back a ways,
and I'm very partial towards his design abilities and products. Apart from
my bias, there's a great value in Kiwa's Premium Filter Modules-- $75.00 for
a filter with typically greater than 100 dB USB ultimate rejection and shape
factor typically 1 to 1.65 is money well spent. The final, resulting
performance depends on the entire receiver system and PCB layout/shielding,
as you mentioned.

The very best filters I've encountered (before I went IF-DSP with the
756Pro) were matched-pair INRAD crystal filters, and the crystal filters in
my RA6790GM.

BTW, before Craig offered filter modules using cascaded ceramics, he tried
the same idea in the Kiwa Multiband AM Pickup (MAP) accessory in the
1989-1991 timeframe. This device provides outboard synchronous detection
plus two IF bandwidths, notch, tone tilt, IF output, low distortion speaker,
and other features. The closest competitor is (was) the Sherwood SE3.

The filter characteristics in my particular MAP are excellent-- the actual
measurements from the Certificate of Performance show an ultimate rejection
at 3.5 kHz of -108 dB and shape factor of 1 to 1.53 in the narrow bandwidth
(3.0 kHz @ -6db nominal) for my particular serial number. The passbands are
also kept flat within +/- 2.0 dB for both the wide and narrow IF bandwidths,
something you don't always find in discrete crystal & mechanical filters.

73,

Guy


"Pete KE9OA" wrote in message
...
Craig's filters are very good, with good ultimate rejection, but
Mechanical Filters have an ultimate rejection of 120dB. True, this figure
is better than the dynamic range of most receivers, and also true, you
have to have very good shielding/ground in your PC board layout scheme in
order to obtain this 120dB figure.
100dB ultimate rejection is a good figure to shoot for. I have spoken to
Craig a few times.....he is a very good designer.

Pete

"Guy Atkins" wrote in message
...
Craig at Kiwa Electronics seems to have figured out how to make quality
filter modules from series-cascaded Murata ceramics. He uses low noise
buffer amps as part of the package in Kiwa's Standard and Premium Filter
Modules (PFMs). The latter are particularly good, with ultimate rejection
figures exceeding Collins mechanical filters and better shape factors
than quality crystal filters... and at a very good price for the
performance.

I've used a number of PFMs over the years in various receivers, and
they've performed excellently for DXing and program listening.

http://www.kiwa.com/kiwa455.html

Guy Atkins
Puyallup, WA USA


"Arthur Pozner" wrote in message
...
Murata filters are not known for good shape factor(6db//60db) . When
stacked in series losses mount rapidly. And they seem to have a lot
more noise when compared to crystal or mechanical IF filters. Have you
tried to build a crystal ladder filter? Very effective and MUCH cheaper
than any other IF badpass filter ,and Inrad is not the costliest-
KVG,Collins etc.may set one back many hundreds of dollars...

Sorry for the botched who wrote what, but I haven't
had time to hunt up a better news reader.
-------------------------------

Arthur Pozner Apr 9, 9:31 am show options
Newsgroups: rec.radio.shortwave
From: (Arthur Pozner) - Find messages by this
author
Date: Sat, 9 Apr 2005 12:31:37 -0400
Local: Sat,Apr 9 2005 9:31 am
Subject: Question for Pete KE9OA
Reply | Reply to Author | Forward | Print | Individual Message | Show
original | Report Abuse

455KHz resonators are commonly used in remote controls. Have not tried
building a lattice or a ladder filter with them yet. But, what if
it just might be succesful-- that's why experimenting is fun.
Collins mechanical filters do have very good rejection just as Pete has
stated, in fact, at 455KHz it might as well have better
selectivity/rejection/low distortion/phase linearity than any other
filter type. I believe R-2000 is a triple conversion receiver, has
anybody ever try changing out the first (roofing) filter.Just for the
sake of experimenting.
------------------------------------------------
Sadly the R2000 uses a variable first IF. Or at least that is what
I tink after a long study of the service manual.
My large diagram is at work, I was in a hurry when I left Friday.

But looking at the service manual that I have at home:
Thefirst ceramic fitler is centered at 9.875, but is rated at +/-
130KHz for the 3db points, and +/- 2MHz! for -30dB points.
The manual says the first If 45.85~45.90MHz then that is
converted to 9.850~9.90 MHz.

I doubt if there is enough market numbers for anyone to make
and market a 9.865MHz +/-15KHz filter. Ihave looked, but have
found one. Building a latice filter is in my dream book.

I tried Pete's suggestions and a set of three stacked 555Is in
series may do the trick. I put the widest one first, followed the
narrower, ending with the narrowist. They are all within spec,
but after measuring 11, I went with the narrowist, that had the
proper center freq. 1 had the center freq at 453.8.

I am listining as we speak and the results are well worth the effort.
It pays to have frineds with well stocked junk rooms. I was given
12 older maxon MCB25 CBs. One was still in the plastic, so I
decided to add that one to the RF stable. That is just what I need,
another radio!

I tried to use the 455 3 pinresonators and found they make great
CW fitlers. The bnadwidth is too naroow, and by stacking them you
get pretty decent skirts.

Terry

You are right.........Craig is pretty sharp. He was telling me about how he
got started, in a conversation we had a couple of weeks ago. Still, I don't
know what kind of setup you were using with the Mechanical Filters, but when
they are properly terminated, there shouldn't be more than 1dB of passband
ripple. I used several versions of the disc-wire filters for years, before
switching to the Torsional Mode units.
About your filter..........................108dB is a very respectable spec
for that unit! Very good shape factor.
I do like those filters in the 6790..........I have one of them myself.
"Guy Atkins" wrote in message
...
Hi Pete,

Craig Siegenthaler and I have been friends since 1988, so we go back a
ways, and I'm very partial towards his design abilities and products.
Apart from my bias, there's a great value in Kiwa's Premium Filter
Modules-- $75.00 for a filter with typically greater than 100 dB USB
ultimate rejection and shape factor typically 1 to 1.65 is money well
spent. The final, resulting performance depends on the entire receiver
system and PCB layout/shielding, as you mentioned.

The very best filters I've encountered (before I went IF-DSP with the
756Pro) were matched-pair INRAD crystal filters, and the crystal filters
in my RA6790GM.

BTW, before Craig offered filter modules using cascaded ceramics, he tried
the same idea in the Kiwa Multiband AM Pickup (MAP) accessory in the
1989-1991 timeframe. This device provides outboard synchronous detection
plus two IF bandwidths, notch, tone tilt, IF output, low distortion
speaker, and other features. The closest competitor is (was) the Sherwood
SE3.

The filter characteristics in my particular MAP are excellent-- the actual
measurements from the Certificate of Performance show an ultimate
rejection at 3.5 kHz of -108 dB and shape factor of 1 to 1.53 in the
narrow bandwidth (3.0 kHz @ -6db nominal) for my particular serial number.
The passbands are also kept flat within +/- 2.0 dB for both the wide and
narrow IF bandwidths, something you don't always find in discrete crystal
& mechanical filters.

73,

Guy

Hi Terry,
I did try to e-mail you directly from work, but you e-mail
bounced, so doing it through the NG sounds good.
Anyway, the Torsional Mode filters are just as good as the disc-wire
filters..........I have been told that they are also more rugged. Exceptions
might be the avionics models that are used in the 651S-1, the PRC-515, etc.

Pete

wrote in message
ups.com...
I have been using a JFET source follower with 2K
gate resistor, and 2K source resistor with a Vdd of
~12V.

I will rebuild using your parts values. I am not looking for
crystal or mechanical filter performance, just as much
better then stock as I can get for the time being.

I plan on buying a Inrad 2.1KHZ crystal filter because I
know how good they work.

Ohter then the external capacitors, are the new torsional
mechanical filters better then the older "bathtub" units?

Terry