Oddly, someone at work yesterday (who does have the Sevick book) was asking
just this question. Here you go:
FairRite, Micrometal, and Amidon both sell the cores in a bewildering
variety of sizes and materials. A google search will find their websites,
and once you've got that, you can find their local dealers, if needed. They
also do mailorder.
http://www.fair-rite.com
http://www.amidoncorp.com/
http://www.micrometals.com
Amidon sells kits for Ununs (Try:
http://www.amidon-inductive.com/associates_prod_chokes.htm) $12.50 for all
the parts.. (more to the point, it also tells you how much wire and what
kind of core you need)
A few weeks back, there was a post on this list giving a reference to an
article describing construction. Also, I think Amidon's website has an
application note.
Basic principle is this:
It's a transformer, so you need to decide what turns ratio you need.
Impedance ratio is turns ratio squared (i.e. a 2:1 turns ratio is a 4:1
impedance ratio), 3:2 is 9:4 (or pretty close to 2:1), etc.
The core material is determined by the frequency you want to work at. Every
mfr has their own coding for this. You DON'T want cores designed for EMI
filtering that are very lossy (like the ones clamped on the keyboard cable
of your computer).
The wire you select should be big enough to carry the current without too
much loss. Depending on how hot you expect it to get, you choose what kind
of insulation you want on the wire (you could wind it with AWG14 THHN house
wire for instance, but that's not good to as high temperature as say,
PolyThermaleze type motor winding wire.) "enamel" insulated wire in a wide
variety is available from places that rewind motors, and a bit of calling
around should find a place willing to sell/give/trade for a beer you a few
or ten feet of AWG whatever. Otherwise you can just order the stuff from
places like Mouser or DigiKey. The problem is that you probably don't need
a whole pound of wire, so the calling around to motor rewinders is a handy
solution.
You need to choose a core with a hole big enough to feed all the wires
through. Yes, bigger cores cost a bit more, but, trying to get wires
through too tiny a hole is a real pain, and you really, really don't want to
nick the insulation in the wire, because it will give you a really tough
intermittent failure to track down.(imagine what the effect of having the
turns ratio changing only on voice peaks, when the voltage gets high enough
to cause breakdown, or breakdown occurring only when it rains, but only on
certain bands, where the mismatch is just right, etc.)
The number of turns is determined by not saturating the core with the design
frequency and voltage.The turns also affects the desired inductance. In
turn, the saturation flux is determined by the cross sectional area of the
core (you can stack multiple cores to increase the total area). There are
design charts, etc., on the various mfr websites to help choose the number
of turns, etc.
In general, you want relatively few turns, because you use less wire (less
resistive losses) and the core is physically smaller (cheaper). However,
few turns makes it tougher to get enough inductance and to get good magnetic
coupling between windings. (You want the reactance of the inductance to
dominate the resistive losses.) Fewer turns also means more volts/turn,
which means more potential saturation effects. You can trade off bigger
cross sectional area (bigger core overall, and lower flux because the field
is spread out over more cross section) with fewer turns (which increases the
volts/turn and reduces inductance), etc.
You want to evenly space the turns, and to have the two windings neatly
overlapped, to improve the magnetic coupling, and reduce the parasitic C
effects. It's helpful to wind some of that nifty fiberglass tape over the
core first to keep the sharp edges on the core from nicking the insulation
on the wire. I've heard that regular old filament packing tape works ok for
this, and so does masking tape. It's really a matter of expected
temperatures (i.e. PVC tape might melt if you're pushing a kW through the
transformer, but, if you're only using it for receive, then who cares...)
The ARRL handbook does cover quite a bit of this.
Really.. what you do is start out with an approximate size of core (say with
a 1" hole), run the numbers, using reasonable wire sizes, etc., and see if
it works. If not, change to another core size and iterate. Be bold, spend
the $5-10 on parts, wind and try it. Cores are cheap and so is wire. From
practical experience, it's easier to just spend a bit more and build it with
a bigger core and bigger wire than you strictly need, because you're
probably not concerned about ekeing out the last bit of profit and volume,
as a manufacturer would be.
----- Original Message -----
From: "Lee Buller" <k0wa@swbell.net>
To: "TowerTalk" <towertalk@contesting.com>
Sent: Wednesday, December 31, 2003 7:46 AM
Subject: [TowerTalk] Information Please?
> Hi Everyone....Happy New Year....
>
> I need to know where I can either purchase an UNUN or an article on how to
make an UNUN and where to get the cores. I have seen ByteMark but I wonder
if there is a way to construct one? I cannot seem to find a QST article and
I don't need to buy a book. I just need parts and insructions. Yes, I am
cheap....but in HR you can be cheap
>
> Lee Buller
> K0WA
>
> _______________________________________________
>
> See: http://www.mscomputer.com for "Self Supporting Towers", "Wireless
Weather Stations", and lot's more. Call Toll Free, 1-800-333-9041 with any
questions and ask for Sherman, W2FLA.
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
See: http://www.mscomputer.com for "Self Supporting Towers", "Wireless Weather
Stations", and lot's more. Call Toll Free, 1-800-333-9041 with any questions
and ask for Sherman, W2FLA.
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|