Telamon wrote:
That depends on the core. Some are no good for 3 to 30 MHz.
TV operates at 50MHz.
--
Telamon
Ventura, California
-----------------------
So people keeep telling me.
I have built many, around 30,of these little gems and they all
worked down to below MW with minimal losses. Read Mr Doty's
comments. They have all worked great from .5MHz to above
20MHz with every scalvaged TV core I have used. Perhaps I
have just been lucky, or maybe, just maybe for cores so small
loss are not as high as many people think.
I will include an observation from Mr Doty himself:"
------------------
"John Doty Sep 6 2000, 1:48 pm show options
Newsgroups: rec.radio.shortwave
From: "John Doty"
Date: Wed, 06 Sep 2000 08:47:12 -0500
Subject: "Improved" matching transformer
In article (Who Did you Expect?) wrote:
I use the same core that I do for the matching transformer, but with 10
turns for both primary and secondary. And while we are on the subject of
cores, I still can not find a lot of difference in ANY [ferritte not
powdered iron!] core that I have tried. Anyone have any thoughts on
that?
The biggest difference in performance due to different ferrites will be
at
longwave, not shortwave. At the highest shortwave frequencies, size
will
matter: 10 turns on a big core will have too much capacitance and
leakage
inductance at 30 MHz. At 10 MHz, 10 turns at 50 ohms looks good with
almost any ferrite.
--
| John Doty "You can't confuse me, that's my job."
--------------------------
and
--------------------------
John Doty Sep 8 2000, 2:57 pm
Newsgroups: rec.radio.shortwave
From: "John Doty"
Date: Fri, 08 Sep 2000 09:56:55 -0500
Subject: "Improved" matching transformer
In article (Who Did you Expect?) wrote:
By "large", what do you mean? I am using the 1.2" OD type 43 Amidon
ferritte core.
That's about the size where the problem becomes interesting. The rule
of
thumb I use is that if the total amount of wire in the transformer is
less
than 5% of a wavelength, the parasitics are likely to be negligible no
matter how you wind it, but if you have more wire, the winding details
will matter in complex ways. 5% of the wavelength at 30 MHz (10 m) is
50
cm, or 20". One turn on your core is about an inch, so a 10:10 is at
the
edge, and a 30:10 violates the rule by a factor of 2 (but it's only a
rule
of thumb).
10 turns was a conservative estimate based on the minimum inductance
expected from a small TV core. With detailed specs, you can compute the
turns required more accurately. If you want a lower 3 dB cutoff of 500
kHz
(this is likely to suffice even for LW DX), you want the magnetizing
inductance on the 50 ohm side to have a reactance of 50 ohms at 500
kHz:
Z = 2*Pi*f*L, so L = Z/(2 Pi f ) = 50/( 2*3.14*500000) = ~16 uH
This calculation is only valid if the lower cutoff is below the knee in
the permeability versus frequency curve. For 43 material, this is at 5
Mhz, so we're fine.
An FT-114-43 has a core constant of about 0.6uH/turn^2, so 5 turns
yields
25*0.6 = 15 uH, which is close enough. A 5:5 is comfortably within the
rule of thumb, while a 15:5 is at the edge (but is probably better than
a 30:10).
I have also used a BLN-73-202 with 7 turns triflar wound, both worked
about the same.
This is a really good core for this kind of transformer. Just *two*
turns
on the 50 ohm side yields 34 uH, for a lower cutoff below 250 kHz.
Since a turn is less than 4 cm, a transformer like this should have
good performance well into the VHF region. This appears to be the
core ICE uses in their transformer (if it isn't this one, it's
something
with very similar specs).
--
| John Doty "You can't confuse me, that's my job."
-------------------------------
I have found his 5% of a wavelength "rule" to be very true.
And by using it Ihave never had a "TV" core, either the
binoculer or the small "O" , fail.
Before getting lost in theory, why not wind one and
see how it works.
Terry
|