Topics

new bandpass filters for 6, 620, and 2200 Meters? #bpf

freefuel@...
 

would anybody else other than myself be interested in bandpass filters for the 6 Meter band, and maybe the lower 630 and 2200 Meter bands? perhaps as an add on kit of parts to alter the existing kits offered by QRP LABS. 

Just a thought

-Justin N2TOH 

N1BUG
 

I would be interested in 630 and 2200 meter bandpass filters as
experimenter items, especially if the passband could be narrow and
slightly tunable like the others. I realize, of course, a 30pf
trimmer won't do much at 2200 meters. :-)

Paul N1BUG

On 04/11/2018 03:36 PM, freefuel@... wrote:
would anybody else other than myself be interested in bandpass
filters for the 6 Meter band, and maybe the lower 630 and 2200 Meter
bands? perhaps as an add on kit of parts to alter the existing kits
offered by QRP LABS. 

Just a thought

-Justin N2TOH

Thomas Yarish
 

Same here.

Tom KJ6MKI

On Apr 11, 2018, at 2:04 PM, N1BUG <paul@...> wrote:

I would be interested in 630 and 2200 meter bandpass filters as
experimenter items, especially if the passband could be narrow and
slightly tunable like the others. I realize, of course, a 30pf
trimmer won't do much at 2200 meters. :-)

Paul N1BUG


On 04/11/2018 03:36 PM, freefuel@... wrote:
would anybody else other than myself be interested in bandpass
filters for the 6 Meter band, and maybe the lower 630 and 2200 Meter
bands? perhaps as an add on kit of parts to alter the existing kits
offered by QRP LABS.

Just a thought

-Justin N2TOH

Chris Wilson
 

Hello Justin, a sharp BPF for 2200 meters would be of great interest
to me. As sharp as possible above 139kHz and below 135kHz. I have
built a few designed using ELSIE but for whatever reason they aren't, in
reality, centred where I wish them to be. Thanks.

Wednesday, April 11, 2018

would anybody else other than myself be interested in bandpass
filters for the 6 Meter band, and maybe the lower 630 and 2200 Meter
bands? perhaps as an add on kit of parts to alter the existing kits offered by QRP LABS.
Just a thought
-Justin N2TOH _._,_._,_


--

2E0ILY
Best regards,
Chris mailto:chris@...
--
Best regards, Chris Wilson (2E0ILY)

freefuel@...
 

Hi, Paul 

yeah I was thinking about the limited adjustment range of the standard trimmer caps Hans uses on the other kits, perhaps double up and use two of them for each of the stock adjustable trimmers?  

outside of the cap requirements I'm interested to see what will be required for the magnetics.  will it be bigger cores, more of them or perhaps a different material. 

no matter the path I see a use for them, it's just a matter of figuring out how to get there. 

Justin N2TOH 

John Backo <jabac@...>
 

Regarding low frequency filters:

Consider the #26 iron core material. It is designed for
AT switching power supplies, operates at about the range you want,
and they are almost found everywhere a power supply is.

They are yellow and white (warning: similar to the #6 material which is yellow and
neutral or grey), and can easily be found in 1" or better size. The Amidon and other sites
list the characteristics.

The other common toroid in power supplies is grey in color; that would probably work also,
but I don't know its specs.

They are at least worth a try. Capacitors are another topic. Probably the best solution
is a silver mica cap with some kind of air-core or similar micro-adjuster.

john
AD5YE

David Bowman <g0mrf@...>
 

Hi all.

As you say, the 30pF trimmer will not do much for tuning on MF / LF

One possibility is a variable inductor but these are getting difficult to find in high values these days.
However, Coilcraft have one which is 275uH and adjustable.  Would probably work very well at 630m

It's the 7M3-224.   Specs can be downloaded from https://www.coilcraft.com/slot7.cfm

5 caps and 2 inductors would make a nice bandpass.

73

David G0MRF

Steve in Okinawa
 

Mica compression trimmers are available up to 1400 pf from RFParts. They
are bigger than the little trimmer on the QCX of course, but not
outlandish.

Maybe garden-variety subminiature 455 kHz IF transformers could also be
used somehow.

Steve JS6TMW

Hans Summers
 

Hi Justin

The 6m BPF is easy and there is already a design on the BPF kit page see http://qrp-labs.com/bpfkit - starting with the 12m BPF and modifying it. 

I think the 10m BPF is easily modifiable to 6m also. 

600m and 2200m are problematic for this particular design (double tuned resonator) on this compact PCB, because the number of turns would be very large and as others have pointed out, the trimmer capacitors won't do anything (already the case on 160m version). 

If anyone can figure out the resolution to these issues I would happily make it available as a version of the kit. 

73 Hans G0UPL

On Wed, Apr 11, 2018 at 10:36 PM, <freefuel@...> wrote:
would anybody else other than myself be interested in bandpass filters for the 6 Meter band, and maybe the lower 630 and 2200 Meter bands? perhaps as an add on kit of parts to alter the existing kits offered by QRP LABS. 

Just a thought

-Justin N2TOH 


Kelly Jack
 

On Thu, Apr 12, 2018 at 04:26 am, Hans Summers wrote:
Hi Justin
 
The 6m BPF is easy and there is already a design on the BPF kit page see http://qrp-labs.com/bpfkit - starting with the 12m BPF and modifying it. 
 
I think the 10m BPF is easily modifiable to 6m also. 
 
600m and 2200m are problematic for this particular design (double tuned resonator) on this compact PCB, because the number of turns would be very large and as others have pointed out, the trimmer capacitors won't do anything (already the case on 160m version). 
 
If anyone can figure out the resolution to these issues I would happily make it available as a version of the kit. 
 
73 Hans G0UPL

On Wed, Apr 11, 2018 at 10:36 PM, <freefuel@...> wrote:
would anybody else other than myself be interested in bandpass filters for the 6 Meter band, and maybe the lower 630 and 2200 Meter bands? perhaps as an add on kit of parts to alter the existing kits offered by QRP LABS. 

Just a thought

-Justin N2TOH 

 

 

 


 Somewhere to start.

http://www.spirat.com.au/vk5zvs/pic55.htm

Regards


Simon VK3ELH

Glen Leinweber
 

For  0.5 MHz band, Steve has a promising idea: 10mm 455 kHz I.F. transformers.
These cup-cores are self-shielding, and a big bonus is tune-ability. Most resonate
with a 180 pf capacitor, which is sometimes included, hidden in the base. The top
cup has screw threads that adjusts a small air gap, allowing about 600uH -700uH adj.
Unloaded Q is in the 100 ballpark.

I have found that most I.F. transformers can be disassembled (with care). All that's needed
is for a tri-filar 50 ohm winding to be added to the inner core. The original winding
(made of very fine #40 wire) would be left intact. The inner-core is a dumbell-shaped
ferrite, and there is usually enough winding space that the extra turns can be
added over the top...even if the trifilar winding is done with heavier wire than #40.

Yes, it is a tricky job - these cans are small and delicate, but with dexterity it is do-able.

cct.oei@...
 

Having recently designed/built some receive filters for the 630 and 2200 meter bands for QSD-based receivers I didn't bother making them tunable as they were 10s of kHz wide and the bandwidth being a sizable percentage of the operating frequency made the task pretty easy.  In the U.S. where we never really had any high-power LWBC stations, LORAN-C is offline (for now) and few people live very close to NDBs - we can probably get away with a filter than pretty wide, but this may not be the case in Europe where LWBC and teleswitch signals can be quite strong.

Having a known-accurate L/C meter (a universal "component tester" capable of measuring inductance really doesn't cut it this task...) helped.  Using 5% capacitors and being able to achieve the target inductance, they came out pretty close to where they were expected:  Their being a few kHz off-center was no problem owing to the overall bandwidth as they still covered the bands just fine.

One of the problems that I've run across is that few manufacturers make large-value variable inductors these days, and the parts suppliers like Mouser and Digi-Key barely stock any at all - at least those that have inductance ranges useful for the LF/MF amateur bands.  While Mouser still stocks 455kHz IF "cans", I have to wonder if they are selling many these days and if, when they run out, they will get more.  To be sure, one can find these types of components here and there, but if I were to publish any sort of design, I'd want it to be replicable by others.

73,
Clint
KA7OEI

Arv Evans
 

Clint  KA7OEI, and others...

Playing with the calculator on my cell phone indicates that for a parallel resonance frequency
of 175 KHz one might use 2100 pf, and 382 uh as a starting place.  Then going to
http://toroids.info I find that the 382 uh can be obtained with 33 turns (18.5 inches or 47 cm)
on an FT37-43 core.  That shows an impedance of 420 ohms which is dependent on resistance,
so some adjustment will be necessary to make a filter. 

Refining this a bit further with the cell phone app...Android Electronic Toolbox Ver-3.78
by Faust Nijhuis
Android Design.nl

 Chebyshev LPF says this:
175 KHz Fc
220 Khz Fs
1.5 db ripple
3db insertion loss
50 ohms
2 nodes
50 ohms in
39.45 nf
29.42 uh  ==  9.2 turns on an FT37-43 core (6.6 inches or 16.8 cm of wire needed)
14.9 ohms out
Connected in reverse that seems to match the output impedance of an IRF510 RF PA
to a 50 ohm load.

The calculator built into http://toroids.info lets you play with many different cores to see which
one might be the best for your needs.

Hopefully this will help designers get beyond the coil-on-a-bucket size for high value inductors. 

Arv K7HKL
_._


On Fri, Apr 13, 2018 at 12:15 PM, <cct.oei@...> wrote:
Having recently designed/built some receive filters for the 630 and 2200 meter bands for QSD-based receivers I didn't bother making them tunable as they were 10s of kHz wide and the bandwidth being a sizable percentage of the operating frequency made the task pretty easy.  In the U.S. where we never really had any high-power LWBC stations, LORAN-C is offline (for now) and few people live very close to NDBs - we can probably get away with a filter than pretty wide, but this may not be the case in Europe where LWBC and teleswitch signals can be quite strong.

Having a known-accurate L/C meter (a universal "component tester" capable of measuring inductance really doesn't cut it this task...) helped.  Using 5% capacitors and being able to achieve the target inductance, they came out pretty close to where they were expected:  Their being a few kHz off-center was no problem owing to the overall bandwidth as they still covered the bands just fine.

One of the problems that I've run across is that few manufacturers make large-value variable inductors these days, and the parts suppliers like Mouser and Digi-Key barely stock any at all - at least those that have inductance ranges useful for the LF/MF amateur bands.  While Mouser still stocks 455kHz IF "cans", I have to wonder if they are selling many these days and if, when they run out, they will get more.  To be sure, one can find these types of components here and there, but if I were to publish any sort of design, I'd want it to be replicable by others.

73,
Clint
KA7OEI


Alan
 

There are a few different formats for 10/7/5 mm IFT’s, the larger Toko 10 series and 10E series are particularly easy to adapt. Often a 455KHz IFT will tune 475KHz but if it is stubbornly LF then the integral C is easy to remove, it is a very thin ceramic tube with contacts each end, a small driver can be used to carefully break up the capacitor and thus completely remove it from circuit. You can then use an external capacitor to resonate the primary for a 475KHz filter.  A 3.9pF top coupling capacitor would produce a bandpass pair at 475KHz with 180pF of internal tuning.  ( not all IFT’s are tuned with 180pF or are on 455KHz)

 

The 135KHz band can be reached with around 1nF across the primary of a 180PF tuned 455KHz IFT, it’s not going to have the Q of an ideal-for-the-task inductor but it is the bird-in-the-hand solution. Murata do still produce 5mm SM L’s but the Q is remarkably low!

 

The dumbbell ferrite bobbin enclosed in the ferrite cup adjuster is a nice part in many ways. After easing off the can and unscrewing the cup the bobbin is revealed. It is often held in place  with candlewax. It is possible to hold a soldering iron near and remove the bobbin after breaking the connecting wires.  If you are able it is possible to rewind the bobbin to much lower frequencies with fine wire as one of those memorable mountain climbing  type exploits  however it is a lot easier to mess with the tuning C !

 

Some very old ( early transistor radio )  IFT’s were very easy to modify, they had ferrite slugs in a conventional  core former with a slip on ferrite cup. Some were wound with litz and had good unloaded Q’s around 200-250. So search that Junk Box.

 

73

Alan

G8LCO

 

 

 

Sent from Mail for Windows 10

 

Chris Wilson
 

Hello Alan,

Talking of LF band pass filters I created this one linked to below, in
ELSIE and built it. I have taken a closer look, using a noise source I
treated myself to, but which I previously didn't understand in so far
as I hadn't realised what a low output level it had. I just tried
using two Chinese broad band amps in series to up the level and fed
the output through my single filter, into an SDR receiver and a
bandscope. It seems the reality of the filter, whilst quite sharp, is
NOT centered on the desired 136 to 137khz. Whilst I appreciate this is
maybe not an ideal filter design, it's what I have on a stripboard (and
you'll possibly condemn the use of miniature leaded inductors, I don't
know... ;)). But it is what is for now, how might I shift the passband
a bit lower, using readily available caps or standard leaded inductor
values please? ELSIE now has me seeing double and not getting very far
varying component values to shift it a bit lower in frequency
retaining low loss. I don't think a trimmer cap anywhere in the circuit is
practical, is it? Lots to learn, especially that a modelled circuit
may not be the same when built!

http://www.chriswilson.tv/lf-circuit.jpg

http://www.chriswilson.tv/lf-plot.jpg

http://www.chriswilson.tv/passband.jpg

Wednesday, April 18, 2018

There are a few different formats for 10/7/5 mm IFT’s, the larger
Toko 10 series and 10E series are particularly easy to adapt. Often
a 455KHz IFT will tune 475KHz but if it is stubbornly LF then the
integral C is easy to remove, it is a very thin ceramic tube with
contacts each end, a small driver can be used to carefully break up
the capacitor and thus completely remove it from circuit. You can
then use an external capacitor to resonate the primary for a 475KHz
filter.  A 3.9pF top coupling capacitor would produce a bandpass
pair at 475KHz with 180pF of internal tuning.  ( not all IFT’s are
tuned with 180pF or are on 455KHz)
The 135KHz band can be reached with around 1nF across the primary
of a 180PF tuned 455KHz IFT, it’s not going to have the Q of an
ideal-for-the-task inductor but it is the bird-in-the-hand solution.
Murata do still produce 5mm SM L’s but the Q is remarkably low!
The dumbbell ferrite bobbin enclosed in the ferrite cup adjuster is
a nice part in many ways. After easing off the can and unscrewing
the cup the bobbin is revealed. It is often held in place  with
candlewax. It is possible to hold a soldering iron near and remove
the bobbin after breaking the connecting wires.  If you are able it
is possible to rewind the bobbin to much lower frequencies with fine
wire as one of those memorable mountain climbing  type exploits 
however it is a lot easier to mess with the tuning C !
Some very old ( early transistor radio )  IFT’s were very easy to
modify, they had ferrite slugs in a conventional  core former with a
slip on ferrite cup. Some were wound with litz and had good unloaded
Q’s around 200-250. So search that Junk Box.
73
Alan
G8LCO




--

2E0ILY
Best regards,
Chris mailto:chris@...
--
Best regards, Chris Wilson (2E0ILY)

Nick Norman
 

On 19/04/18 12:31, Chris Wilson wrote:
Hello Alan,

Talking of LF band pass filters I created this one linked to below, in
ELSIE and built it. I have taken a closer look, using a noise source I
treated myself to, but which I previously didn't understand in so far
as I hadn't realised what a low output level it had. I just tried
using two Chinese broad band amps in series to up the level and fed
the output through my single filter, into an SDR receiver and a
bandscope. It seems the reality of the filter, whilst quite sharp, is
NOT centered on the desired 136 to 137khz. Whilst I appreciate this is
maybe not an ideal filter design, it's what I have on a stripboard (and
you'll possibly condemn the use of miniature leaded inductors, I don't
know... ;)). But it is what is for now, how might I shift the passband
a bit lower, using readily available caps or standard leaded inductor
values please? ELSIE now has me seeing double and not getting very far
varying component values to shift it a bit lower in frequency
retaining low loss. I don't think a trimmer cap anywhere in the circuit is
practical, is it? Lots to learn, especially that a modelled circuit
may not be the same when built!

http://www.chriswilson.tv/lf-circuit.jpg

http://www.chriswilson.tv/lf-plot.jpg

http://www.chriswilson.tv/passband.jpg
Chris,

How about this ...

<http://www.dctower.co.uk/images/137-bpf-sch.png>
<http://www.dctower.co.uk/images/137-bpf-plt.png>

I can't test it but ...

Regards

Nick
M0HGU

--
Instead, people would take pains to tell her that beauty was
only skin-deep, as if a man ever fell for an attractive pair of
kidneys.
(Maskerade)
12:50:01 up 20 days, 23:18, 16 users, load average: 0.24, 0.45, 0.57


--
73 Nick M0HGU

N1BUG
 

Hello Chris,

Looking at your ELSIE plot vs. your real world plot I see the latter
has a soft, rounded peak where the former does not. I suspect the Q
of you components is less than what you specified in ELSIE. This
will also cause the filter loss to be higher than predicted. Did the
inductors have a Q specification at or near the frequency you are
using them on? The capacitors may be poor for RF use also. Ceramics
can be hit or miss when you are looking for tight performance like
this. I'd try poly-something, silver mica, or mylar perhaps.

As for why it is off frequency, you got me there. I would suspect
part values are off, but I changed values in the model by as much as
10% and it didn't move this far. If the SDR you are using has an
input filter, the two filters may be interacting.

At the moment those are all the thoughts I have but more may sneak
up on me later.

73,
Paul

On 04/19/2018 07:31 AM, Chris Wilson wrote:
Hello Alan,

Talking of LF band pass filters I created this one linked to below, in
ELSIE and built it. I have taken a closer look, using a noise source I
treated myself to, but which I previously didn't understand in so far
as I hadn't realised what a low output level it had. I just tried
using two Chinese broad band amps in series to up the level and fed
the output through my single filter, into an SDR receiver and a
bandscope. It seems the reality of the filter, whilst quite sharp, is
NOT centered on the desired 136 to 137khz. Whilst I appreciate this is
maybe not an ideal filter design, it's what I have on a stripboard (and
you'll possibly condemn the use of miniature leaded inductors, I don't
know... ;)). But it is what is for now, how might I shift the passband
a bit lower, using readily available caps or standard leaded inductor
values please? ELSIE now has me seeing double and not getting very far
varying component values to shift it a bit lower in frequency
retaining low loss. I don't think a trimmer cap anywhere in the circuit is
practical, is it? Lots to learn, especially that a modelled circuit
may not be the same when built!

http://www.chriswilson.tv/lf-circuit.jpg

http://www.chriswilson.tv/lf-plot.jpg

http://www.chriswilson.tv/passband.jpg

Wednesday, April 18, 2018

There are a few different formats for 10/7/5 mm IFT’s, the larger
Toko 10 series and 10E series are particularly easy to adapt. Often
a 455KHz IFT will tune 475KHz but if it is stubbornly LF then the
integral C is easy to remove, it is a very thin ceramic tube with
contacts each end, a small driver can be used to carefully break up
the capacitor and thus completely remove it from circuit. You can
then use an external capacitor to resonate the primary for a 475KHz
filter.  A 3.9pF top coupling capacitor would produce a bandpass
pair at 475KHz with 180pF of internal tuning.  ( not all IFT’s are
tuned with 180pF or are on 455KHz)
The 135KHz band can be reached with around 1nF across the primary
of a 180PF tuned 455KHz IFT, it’s not going to have the Q of an
ideal-for-the-task inductor but it is the bird-in-the-hand solution.
Murata do still produce 5mm SM L’s but the Q is remarkably low!
The dumbbell ferrite bobbin enclosed in the ferrite cup adjuster is
a nice part in many ways. After easing off the can and unscrewing
the cup the bobbin is revealed. It is often held in place  with
candlewax. It is possible to hold a soldering iron near and remove
the bobbin after breaking the connecting wires.  If you are able it
is possible to rewind the bobbin to much lower frequencies with fine
wire as one of those memorable mountain climbing  type exploits 
however it is a lot easier to mess with the tuning C !
Some very old ( early transistor radio )  IFT’s were very easy to
modify, they had ferrite slugs in a conventional  core former with a
slip on ferrite cup. Some were wound with litz and had good unloaded
Q’s around 200-250. So search that Junk Box.
73
Alan
G8LCO



--
Paul
N1BUG 160m-2m DXCC Honor Roll
WI2XTC 2200m-630m Experimental license
FN55mf ME Piscataquis County
http://www.n1bug.com
http://www.aurorasentry.com

Alan
 

Hi Chris,

Welcome to the real world.

The original thread was about extending  bandpass pairs down in frequency.  If you took a 455KHz IFT, added 1nF of capacitance and  top coupled a pair with 18 or 22pF then you would have a bandpass pair around 135KHz that would be tuneable. You can use the existing coupling windings or make a capacitive tap by adding a largish C -say around 10nF- between the 1nF and ground. ( If you do this  it reduces the tuning capacitance so you might need to add 100pF across the 1nF’s. In practice the Q may well be somewhat low and the top coupling may need to be larger to avoid excessive loss.

 

However you have modelled a different filter with fixed values. So the actual values are too small and the filter is HF. For a filter of this type the best way is to open up the two series elements and the parallel element and tune them individually to the centre frequency.   Resonate  the series 680uH //2nF by adding C to the 2nF to the band centre, do the same for the other series tuned circuit then  couple the one end grounded parallel tuned circuit via  small value capacitors  - say 1nF to the source and detector and resonate by adding C to the 607nF capacitor. Then when you re-connect the elements all should be not too far off.

 

Using wire ended inductors that are wound onto a ferrite rod  core may result in inductive coupling if the bodies are close, if they are a body length apart the coupling will be small- hint -that is yet another way of making bandpass filters!

 

At LF many film capacitors have good Q, polystyrene is very good but very few manufacturers left. Mylar or better polycarb or PS are excellent. Many film C’s are 20% tolerance as bog standard but 5% or sometimes 2% or 1% tolerance are available for a higher cost. Do not waste time with Hi K ceramics.

 

Some wire ended inductors have poor Q, that will show up on a resonance test.

 

Yet another way would be to grab a few small ferrite rods from scrapped motherboards and wind some wire onto a cardboard former that can slide along the core- a presettable fixed inductor.

 

If you want to go further might I suggest that a RF  (or even an AF for LF) generator is an invaluable tool , the SDR and noise source is a viable alternative but even a simple LC oscillator can be very useful when working with filters of all kinds.

 

 

73,

Alan

G8LCO

 

 

 

Sent from Mail for Windows 10

 

 

 

N1BUG
 

I looked at this again. The first time I hadn't played with values
in the parallel resonant center section. Of course as I should hvae
realized that's the critical bit! You'll need to be within 1% of
correct values there. In a filter this narrow, Q of components will
play havoc too.

I'm going to have a look in my parts bins later. Id like to see if I
have the right things to build one of these up and experiment with it.

Paul

On 04/19/2018 08:31 AM, N1BUG wrote:
Hello Chris,

Looking at your ELSIE plot vs. your real world plot I see the latter
has a soft, rounded peak where the former does not. I suspect the Q
of you components is less than what you specified in ELSIE. This
will also cause the filter loss to be higher than predicted. Did the
inductors have a Q specification at or near the frequency you are
using them on? The capacitors may be poor for RF use also. Ceramics
can be hit or miss when you are looking for tight performance like
this. I'd try poly-something, silver mica, or mylar perhaps.

As for why it is off frequency, you got me there. I would suspect
part values are off, but I changed values in the model by as much as
10% and it didn't move this far. If the SDR you are using has an
input filter, the two filters may be interacting.

At the moment those are all the thoughts I have but more may sneak
up on me later.

73,
Paul


On 04/19/2018 07:31 AM, Chris Wilson wrote:
Hello Alan,

Talking of LF band pass filters I created this one linked to below, in
ELSIE and built it. I have taken a closer look, using a noise source I
treated myself to, but which I previously didn't understand in so far
as I hadn't realised what a low output level it had. I just tried
using two Chinese broad band amps in series to up the level and fed
the output through my single filter, into an SDR receiver and a
bandscope. It seems the reality of the filter, whilst quite sharp, is
NOT centered on the desired 136 to 137khz. Whilst I appreciate this is
maybe not an ideal filter design, it's what I have on a stripboard (and
you'll possibly condemn the use of miniature leaded inductors, I don't
know... ;)). But it is what is for now, how might I shift the passband
a bit lower, using readily available caps or standard leaded inductor
values please? ELSIE now has me seeing double and not getting very far
varying component values to shift it a bit lower in frequency
retaining low loss. I don't think a trimmer cap anywhere in the circuit is
practical, is it? Lots to learn, especially that a modelled circuit
may not be the same when built!

http://www.chriswilson.tv/lf-circuit.jpg

http://www.chriswilson.tv/lf-plot.jpg

http://www.chriswilson.tv/passband.jpg

Wednesday, April 18, 2018

There are a few different formats for 10/7/5 mm IFT’s, the larger
Toko 10 series and 10E series are particularly easy to adapt. Often
a 455KHz IFT will tune 475KHz but if it is stubbornly LF then the
integral C is easy to remove, it is a very thin ceramic tube with
contacts each end, a small driver can be used to carefully break up
the capacitor and thus completely remove it from circuit. You can
then use an external capacitor to resonate the primary for a 475KHz
filter.  A 3.9pF top coupling capacitor would produce a bandpass
pair at 475KHz with 180pF of internal tuning.  ( not all IFT’s are
tuned with 180pF or are on 455KHz)
The 135KHz band can be reached with around 1nF across the primary
of a 180PF tuned 455KHz IFT, it’s not going to have the Q of an
ideal-for-the-task inductor but it is the bird-in-the-hand solution.
Murata do still produce 5mm SM L’s but the Q is remarkably low!
The dumbbell ferrite bobbin enclosed in the ferrite cup adjuster is
a nice part in many ways. After easing off the can and unscrewing
the cup the bobbin is revealed. It is often held in place  with
candlewax. It is possible to hold a soldering iron near and remove
the bobbin after breaking the connecting wires.  If you are able it
is possible to rewind the bobbin to much lower frequencies with fine
wire as one of those memorable mountain climbing  type exploits 
however it is a lot easier to mess with the tuning C !
Some very old ( early transistor radio )  IFT’s were very easy to
modify, they had ferrite slugs in a conventional  core former with a
slip on ferrite cup. Some were wound with litz and had good unloaded
Q’s around 200-250. So search that Junk Box.
73
Alan
G8LCO



--
Paul
N1BUG 160m-2m DXCC Honor Roll
WI2XTC 2200m-630m Experimental license
FN55mf ME Piscataquis County
http://www.n1bug.com
http://www.aurorasentry.com

Alan
 

Filterers,

Look at the Centre Fx to Bandwidth ratio, for 136KHz and 2KHz that’s  68:1 so you need Q’s >>68  if the filter is not going to be very lossey.  Ideally there should be NO tuning error  but without continuously variable parts that’s tricky, the way that a filter tunes up when tweaked is a good diagnostic, top coupled with C bandpass filters tune in a different way to the equally terminated ladder that Chris modelled. (You normally end up on the HF edge of a bandpass curve when tuning for max signal with a single tone with top C coupling). Obviously an air spaced variable capacitor  can be a handy tool to add a tweakable amount of C which can then be replaced with a fixed part. If the C was too high then add a larger series C to bring the total C down then add the variable.

 

For an equally terminated ladder all of the series and parallel sections are tuned to the centre frequency normally, you can try for extra B/W by stagger tuning if needed but start with the centre frequency ( the geometric mean of the -3dB points I.e. sqr Fl*Fh).

 

There is an easy  way of calculating the product of L and C for tuned circuits.

 

FOR  F UNDER 1MHz

 

L*C=   25330/ F*F  or   F*F=25330/L*C          with F= KHz,  L =mH,  C=nF

 

I.e. at 455KHz  L*C=  25330/455*455=0.1223        

 

So with C=0.2nF ( 180pF and 20pF strays) the inductance is ) 0.1223/0.2  =0.611mH    for an 180pF tuned IFT

 

So at 136 KHz the L*C number is   25330/136*136= 1.3695,    so  2nF and 0.6847mH,  0.0022mH and 622.5nF     ( NB that’s for centre 136.0 which is not band centre! You can refine it yourself!

 

 

FOR F >1MHz

 

Exactly the same process but F is in MHz, C in pF, L in uH 

 

e.g. a tuned circuit at 10MHz

 

L*C=25330/10*10  = 253.3   , so that could be  100pF and 2.53uH

 

At 100MHz,  L*C=25330/ 10000. Could be  10pF and 0.253uH

 

 

There is no snake oil involved, it’s a re-arrangement of the normal resonance formulae.

For SMALL changes of L or C we differentiate the product to indicate that we need a 2% change in C or L if we are 1%   out in frequency. Big changes half the C or L, 0.707 /1.414 change in Fx,  four times the L or C , 2/0.5  change in Fx. Ten times C or L is sqr(10) Fx change.(0.316/3.16)

 

The optimum form of the  BP filter depends on the centre frequency /bandwidth ratio. For  > 20 the top coupled parallel resonator takes some beating. The equally terminated ladder BPF that Chris modelled is  excellent for octave bandwidths but becomes tricky with CF/BW >10. The shunt element, the parallel tuned circuit becomes very low Z, very large C and tiny L  while the series part becomes very large L and small C    both of which are not often a good mixture. (It is used with the series element as a quartz crystal with very high L and tiny C).

 

For most narrow bands  the top C coupled parallel tuned circuit BPF is a good choice, the coupling C’s add extra roll-off on the LF side greatly reducing MW break-through for HF bands . At 136KHz in Europe two or three inductors may be enough however if you are close to the  “telecommand” FSK  TX  you can add a notch filter to greatly reduce the unwanted signal.

 

If you are winding coils then remember L is proportional to the turns squared, twice the turns half the frequency for the same tuning C ( in practice slightly different because of winding capacitance but as a Rule-Of-Thumb  it works well.) You can also say that turns are proportional to wavelength!

 

The above is a big    simplification for non-specialist filter makers. Filter design is specialised, after you get beyond simple LPF’s, HPF’s and BPF’s there is a mass of material.  The choice of magnetic materials is a BIG topic with many pitfalls with  ferrites and dust iron materials. It is particularly sad that many cores are no longer available, a well stocked “Junk Box” is even more important now than ever before with so many desirable bits unobtainable. A good time to hone those Hunter Gatherer instincts and fill the Loft with treasure!

 

73,

Alan

G8LCO

 

 

 

 

 

Sent from Mail for Windows 10