Daft question time again.

I've got some ceramic filters in some radios that are too wide.

First:
I've got a repeater with a 455khz IF filter that's the 3 pin type, marked
"55F" and "CM". It's in a FX5000 reciever.

Second:
I've got some mobiles with the larger filters, marked LT455FW.


I want to put these all onto a nice and tight filter. The problem is that
I'm trying to use 145.625, but there's two repeaters on 145.6125 and
145.6375 locally, so a frequency shift won't do anything. Deviation is
2.5khz peak

1: What sort of bandwidth do people recommend? I have been told that 9khz
would work a treat, but I don't know if I should go narrower or wider?

2: I've found this site: http://www.euroquartz.co.uk/ceramic-filt.htm

What are all the filter types, what do they mean (like the LTW33, LTS, etc,
etc), and which models would be suitable substitutes?


Look forward to people's responses, because this is confusing me now!


Sam
M1FJB

Sam,



This is what Engineers must struggle through. Looking at the data
sheets I only see size differences in the different types. The electrical
specs appear to be the same.



When filtering FM you must consider the bandwidth of the transmitted signal
which includes the sidebands. Both of your received signals AND the
adjacent channel transmitter SIDEBANDS. There needs to be space between the
carriers to prevent interference and sidebands from one Tx shouldn't be
within the next receiver's bandwidth. Not having my text with me I hope I
analyze your numbers correctly. Just remember that FM is wider than just
the modulation frequency (according to the Bessel functions).

Narrow band FM (B 1) is narrow, but it does have some bandwidth and it is
twice the modulation frequency for sure and some at three time.



If I do the math correctly, your modulation index is about 2.5 / 3 or .833
(assuming 300-3,000 audio). I always have to think carefully how this ratio
goes.

Using:

http://www.rfcafe.com/references/electrical/bessel.htm



This puts us to the left of the B=1 (on the horizontal axis on the bottom)
and I see that the first and second sidebands (J1-black & J2-green dotted)
are significant, but the third (J3-red dotted) is pretty small, so it may be
attenuated in your filter with little resulting distortion. HOWEVER, this
sideband IS PRESENT in yours and the other transmitters. This says that the
second transmitter sidebands go out to +/- 6Kc AND with the third at +/- 9Kc
being pretty small.



Now, The sidebands from the other stations do the same (assuming they are
modulating correctly). This means that the one above you has sideband
energy at its Fc-6kc (second sideband) and some at Fc-9kc (third sideband).
Since its carrier is 12.5 kc. above you this puts them at 6.5kc. and 3.5kc.
ABOVE your carrier frequency!



So, with the filter, you're trying to get rid of a component that is 6.5 Kc.
above your carrier, that is the larger one, and one is 3.5kc. above your
carrier, that is smaller one. (The others are there, but much, much smaller
according to the Bessel curves).

The problem here is that you need to RECEIVE sidebands that are 3, 6, and to
some extent 9 Kc. above your carrier – from the desired stations. There is
a conflict here and it is due to carrier spacing that is too close for FM.

You have to look at the filter curves to see which one will provide some
attenuation at +/-6.5kc. The trouble is that it must pass +/- 6 Kc!
You'll also never get rid of the component at 3.5Kc and still receive your
desired second sideband at 6Kc.



The other unknown here is the location s of these systems and the stations
that will be using them. If someone is far from and using your system, but
close to one of the other systems you mention, those sidebands can be much
stronger and cause more "splatter" in the eyes of the receiver.



Receivers, at least the ones I have been connected with the design of, take
this into account and use filters that have all these considerations
included in their design and try to keep the bandwidth wide enough to
receive what it needs to receive and narrow enough – and with steep enough
skirts -- to attenuate adjacent and alternate channel sidebands as much as
feasible given economic constraints. There is, however a limit to how close
you can put carriers and not get interference *REGARDLESS* of the shape of
the filter at ANY expense.. I think you are past that limit.



I hope you understand "Kc." I 'm from a time long ago (:-)



I find that Hams don't understand the spacing of 2 meter systems and expect
to put them on adjacent channels. You can (should) only do this when the
systems are far enough apart so this sideband energy is low enough to keep
this kind of interference below audible levels.



73, Steve, K,9.D;C'I







"Samuel Hunt" wrote in message
...

Daft question time again.

I've got some ceramic filters in some radios that are too wide.

First:
I've got a repeater with a 455khz IF filter that's the 3 pin type, marked
"55F" and "CM". It's in a FX5000 reciever.

Second:
I've got some mobiles with the larger filters, marked LT455FW.


I want to put these all onto a nice and tight filter. The problem is that
I'm trying to use 145.625, but there's two repeaters on 145.6125 and
145.6375 locally, so a frequency shift won't do anything. Deviation is
2.5khz peak

1: What sort of bandwidth do people recommend? I have been told that 9khz
would work a treat, but I don't know if I should go narrower or wider?

2: I've found this site: http://www.euroquartz.co.uk/ceramic-filt.htm

What are all the filter types, what do they mean (like the LTW33, LTS,
etc,
etc), and which models would be suitable substitutes?
Look forward to people's responses, because this is confusing me now!
Sam, M1FJB

Carson's Rule:
The required filter bandwidth for accurate recovery of an FM signal is
(2*deltaF) + (2*Fmod).

Assuming from the requestor's previous that deltaF = 2.5 khz, and assuming
that the highest audio frequency is 2.5 hHz, then the required filter
bandwidth is 7.5 kHz. Whether this filter will reject the adjacent channel
is another matter.

Joe
W3JDR



"Steve Nosko" wrote in message
...
Sam,



This is what Engineers must struggle through. Looking at the data
sheets I only see size differences in the different types. The electrical
specs appear to be the same.



When filtering FM you must consider the bandwidth of the transmitted
signal
which includes the sidebands. Both of your received signals AND the
adjacent channel transmitter SIDEBANDS. There needs to be space between
the
carriers to prevent interference and sidebands from one Tx shouldn't be
within the next receiver's bandwidth. Not having my text with me I hope
I
analyze your numbers correctly. Just remember that FM is wider than just
the modulation frequency (according to the Bessel functions).

Narrow band FM (B 1) is narrow, but it does have some bandwidth and it is
twice the modulation frequency for sure and some at three time.



If I do the math correctly, your modulation index is about 2.5 / 3 or .833
(assuming 300-3,000 audio). I always have to think carefully how this
ratio
goes.

Using:

http://www.rfcafe.com/references/electrical/bessel.htm



This puts us to the left of the B=1 (on the horizontal axis on the bottom)
and I see that the first and second sidebands (J1-black & J2-green dotted)
are significant, but the third (J3-red dotted) is pretty small, so it may
be
attenuated in your filter with little resulting distortion. HOWEVER, this
sideband IS PRESENT in yours and the other transmitters. This says that
the
second transmitter sidebands go out to +/- 6Kc AND with the third at +/-
9Kc
being pretty small.



Now, The sidebands from the other stations do the same (assuming they are
modulating correctly). This means that the one above you has sideband
energy at its Fc-6kc (second sideband) and some at Fc-9kc (third
sideband).
Since its carrier is 12.5 kc. above you this puts them at 6.5kc. and
3.5kc.
ABOVE your carrier frequency!



So, with the filter, you're trying to get rid of a component that is 6.5
Kc.
above your carrier, that is the larger one, and one is 3.5kc. above your
carrier, that is smaller one. (The others are there, but much, much
smaller
according to the Bessel curves).

The problem here is that you need to RECEIVE sidebands that are 3, 6, and
to
some extent 9 Kc. above your carrier – from the desired stations. There
is
a conflict here and it is due to carrier spacing that is too close for FM.

You have to look at the filter curves to see which one will provide some
attenuation at +/-6.5kc. The trouble is that it must pass +/- 6 Kc!
You'll also never get rid of the component at 3.5Kc and still receive your
desired second sideband at 6Kc.



The other unknown here is the location s of these systems and the stations
that will be using them. If someone is far from and using your system,
but
close to one of the other systems you mention, those sidebands can be much
stronger and cause more "splatter" in the eyes of the receiver.



Receivers, at least the ones I have been connected with the design of,
take
this into account and use filters that have all these considerations
included in their design and try to keep the bandwidth wide enough to
receive what it needs to receive and narrow enough – and with steep enough
skirts -- to attenuate adjacent and alternate channel sidebands as much
as
feasible given economic constraints. There is, however a limit to how
close
you can put carriers and not get interference *REGARDLESS* of the shape of
the filter at ANY expense.. I think you are past that limit.



I hope you understand "Kc." I 'm from a time long ago (:-)



I find that Hams don't understand the spacing of 2 meter systems and
expect
to put them on adjacent channels. You can (should) only do this when the
systems are far enough apart so this sideband energy is low enough to keep
this kind of interference below audible levels.



73, Steve, K,9.D;C'I







"Samuel Hunt" wrote in message
...

Daft question time again.

I've got some ceramic filters in some radios that are too wide.

First:
I've got a repeater with a 455khz IF filter that's the 3 pin type, marked
"55F" and "CM". It's in a FX5000 reciever.

Second:
I've got some mobiles with the larger filters, marked LT455FW.


I want to put these all onto a nice and tight filter. The problem is that
I'm trying to use 145.625, but there's two repeaters on 145.6125 and
145.6375 locally, so a frequency shift won't do anything. Deviation is
2.5khz peak

1: What sort of bandwidth do people recommend? I have been told that 9khz
would work a treat, but I don't know if I should go narrower or wider?

2: I've found this site: http://www.euroquartz.co.uk/ceramic-filt.htm

What are all the filter types, what do they mean (like the LTW33, LTS,
etc,
etc), and which models would be suitable substitutes?
Look forward to people's responses, because this is confusing me now!
Sam, M1FJB