Hello,

I'm modifying a 28-50 MHz transverter to work the new (for us) 70 MHz
band, not having a spectrum analyzer I'm trying to get as close as
possible with substitutions. I have a doubt about the low pass filter
after the final amplifier:
on the arrl handbook I have found formulas for pi-filters:

----|----L------|--------
=C/2 =C/2 R
----|-----------|---------

(I hope the ascii art is readable)

L=R/(pi Fc) C=1/(pi Fc R)

Now the 50 MHz original filter is composed of three cascaded filters
as the one above, the capacitor are 47pF at the two ends and 100 pF
the two middle ones, which is consistent with the formula as the
middle ones can be actually seen as two parallel capacitor for the
cascaded pi-cells, the formula indicates an Fc of about 63 MHz with
these values, which seems correct.
The inductors are 6 turns, internal diameter 6 mm, enameled 0.8 mm
wire, no spaces between turns for the two "external" inductors and 7
turns same diameter for the internal one.
These inductors should have an inductance of 0.17uH the 6 turns ones
and 0.21 uH the 7 turns one and the formula don't agree with these
values, can anyone explain where I'm wrong?
Thanks in advance

Francesco IZ8DWF

Hi Allison,

On 13 Lug, 04:13, wrote:


Generally if it's two cascaded 1/4wave sections as a half wave filter
they are identical. However the designer may also have the filter
going from 50ohms to some other non 50 ohm impedence.
It's not uncommon to see that. Also the middle cap for the two
sections usually is the some of the two end caps if the sections
were viewed in isolation.

In most cases when I encounter that I scale the values by 50/70 (.714)
so that the 47 pf becomes around 33 and the 100pf around 68 and same
for the coils. That maintains the same relationships and impedences
overall.

Actually I found my mistake, the formula I was using for inductance
required the external coil diameter and I was using the internal
diameter, so everything is clear now.
For instance the values you suggest are just right even calculating
with the formulas for a scaled cutoff frequency of about 90 MHz.
By the way, this filter is clearly for 50 ohms both input and output,
one side is the antenna and the other side is the output impedance
adapter filter of the final transistor.

Now the question is if I should try to use the same final (and driver)
transistor of the original 50 MHz transverter (2sc2166 driver and
2sc1969 final) which are 27 MHz transistor (working fine at 50 MHz).
On the datasheets I have been able to find there's no indication of
their ft.
I have a better final transistor but it would require me to move the
amplifier part outside of the original box (different case).

Thanks and 73

Francesco IZ8DWF

On Jul 12, 1:17 pm, wrote:
Hello,

I'm modifying a 28-50 MHz transverter to work the new (for us) 70 MHz
band, not having a spectrum analyzer I'm trying to get as close as
possible with substitutions. I have a doubt about the low pass filter
after the final amplifier:
on the arrl handbook I have found formulas for pi-filters:

----|----L------|--------
=C/2 =C/2 R
----|-----------|---------

(I hope the ascii art is readable)

L=R/(pi Fc) C=1/(pi Fc R)

Now the 50 MHz original filter is composed of three cascaded filters
as the one above, the capacitor are 47pF at the two ends and 100 pF
the two middle ones, which is consistent with the formula as the
middle ones can be actually seen as two parallel capacitor for the
cascaded pi-cells, the formula indicates an Fc of about 63 MHz with
these values, which seems correct.
The inductors are 6 turns, internal diameter 6 mm, enameled 0.8 mm
wire, no spaces between turns for the two "external" inductors and 7
turns same diameter for the internal one.
These inductors should have an inductance of 0.17uH the 6 turns ones
and 0.21 uH the 7 turns one and the formula don't agree with these
values, can anyone explain where I'm wrong?
Thanks in advance

Francesco IZ8DWF

Hi Francesco,

I guess you have things figured out pretty well now, but for future
reference, I'll offer a few additional points:

-- You can get some pretty decent filter design software for free from
www.aade.com. It will tell you a lot about the expected performance
of the designed filter. It also has coil design help, though the
dimensions are in (ugh) English units.

-- You can also get the evaluation/student/ham version of Elsie from
www.tonnesoftware.com. I've been using the professional version for a
while now, and I think the value/price ratio is very good. I've
evaluated some much more expensive filter design programs that offer
only marginally better features.

-- For coils used at HF and above, you will get better Q if you use a
little space between turns. That is, if you need a coil with, say,
1mm turn pitch, it's better to use wire that's about 0.5 - 0.6mm
diameter, than to use wire that's 1mm diameter. For example,
according to a coil program I use and trust, your 6 turn coil should
be about 186nH, with a Qu of 150 at 50MHz. That's using 21AWG
(0.724mm) wire with 0.80mm turn pitch. Changing to 26AWG wire, same
turn pitch, now the pitch is about 1.9 wire diameters, so there's
almost as much space between turns as the wire diameter. Now the
inductance is predicted to be about half a nanohenry less, but the
predicted Q is 175. You might think at first that the larger wire
will have lower RF resistance, but the proximity effect of closely
spaced wires dominates as the turns are immediately next to each
other.

-- You can wind coils with nice even spacing by winding them on the
threads of a bolt. After you unscrew the bolt, if the wire diameter
is fairly large for the particular coil diameter, the coil may well be
self-supporting enough to be useful as-is. You can further stabilize
it by putting a couple rows of appropriate glue parallel to the axis
of the coil, making it look a little like the old "Air-Dux" coils.
Use glue with reasonably low dielectric constant and low RF loss.

Hope this is helpful.

Cheers,
Tom

In article . com,
K7ITM wrote:

-- You can wind coils with nice even spacing by winding them on the
threads of a bolt. After you unscrew the bolt, if the wire diameter
is fairly large for the particular coil diameter, the coil may well be
self-supporting enough to be useful as-is.

An article with some graphs showing the inductance and Q of coils
which were wound on commonly-available bolts is at

http://www.radagast.org/~dplatt/hamr...rted-coils.pdf

--
Dave Platt AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!