- On 03/08/2013 02:45 AM, ceo@... wrote:
>

Transformer (and balun) impedance ratios are proportional to

> I'm using http://www.mingtek.tw/pdf/B-4.pdf

>

> 100Î© balanced fed into pins 14 and 16

>

> 50Î© output to R8 from pin 1

>

> pins 4 and 13 grounded

>

> Feel free to tell me I got it all wrong, I would not be

> at all surprised.

the square of the turns ratios. Using only half of a

winding reduces the impedance by a factor of 1/4.

In concept, since a 1:2 impedance ratio requires a turns

ratio of 1:sqrt(2), that impedance ratio can be only

approximated.

A turns ratio of 3:2 gives an impedance ratio of 9:4 or

about 100:44.44.

A turns ratio of 4:3 gives an impedance ratio of 16:9 or

about 100:56.25.

A turns ratio of 5:3 gives an impedance ratio of 25:9 or

about 100:36.

A turns ratio of 5:4 gives an impedance ratio of 25:16 or

about 100:64.

etc.

--

Regards,

John Popelish - That coupled with the fact that the receiver probably doesn't have an

exact 50 Ohm input Z, and of course the loop antenna Z is very likely to

change over a fairly wide range anyway. So I sure wouldn't worry about

some mismatch.

I have a similar setup here where I have an 88 Foot long dipole fed with

ladder line (the best I can do here). It feeds an oversized 2KW 4:1

balun (oversized to minimize heating losses from changing the balun

characteristics) before hitting the transmatch (MFJ949E). The balun is

primarily there for balanced feed as I realize the impedance varies a

great deal over the 3.5-29MHz range I use it. But it DOES perform very

well on all HF bands (and I've even used it on 6M too). I've tried MANY

different HF antennas in this Mobile home park (NO real estate space to

speak of) and this one outperforms everything else I've tried.

Yes, some verticals were a bit better on DX (especially a Butternut HF2V

I tried on 80/40), but their noise was WAY too high here.

So, I'd suggest just trying whatever matching ratio you have and see

what works best for you overall.

73 de Phil, KO6BB

http://www.qsl.net/ko6bb/ (Web Page)

RADIOS:

Grundigs: S-350 (2006)& G6 (2011).

Icom: R-75 with Cascaded 250Hz CW Filters.

Kenwood: TS130S HF Transceiver (circa 1980).

Radio Shack: DX-380 digital portable (circa 1990).

Zenith: Royal-7000 Transoceanic (circa 1969).

Audio Interface: Two Mark One Earholes.

Decoder Software: Gray-matter between two ears.

ACCESSORIES: Homebrew LF-MF Pre-Amp, MFJ-949E HF Tuner

Homebrew 6 Hz Audio Filter.

ANTENNAS: 88' Long Ladder-line fed dipole at 35 feet AGL.

Ratzlaff 3 Foot Active Whip at 36 Feet AGL for LF/MW

4 Foot/side Tuned Rotatable Loop at 15 Feet AGL for LF

Merced, Central California, 37, 18, 37N 120, 30, 6W CM97rh

On 3/8/2013 5:02 PM, John Popelish wrote:

> On 03/08/2013 02:45 AM, ceo@... wrote:

>> I'm using http://www.mingtek.tw/pdf/B-4.pdf

>>

>> 100Î© balanced fed into pins 14 and 16

>>

>> 50Î© output to R8 from pin 1

>>

>> pins 4 and 13 grounded

>>

>> Feel free to tell me I got it all wrong, I would not be

>> at all surprised.

> Transformer (and balun) impedance ratios are proportional to

> the square of the turns ratios. Using only half of a

> winding reduces the impedance by a factor of 1/4.

>

> In concept, since a 1:2 impedance ratio requires a turns

> ratio of 1:sqrt(2), that impedance ratio can be only

> approximated.

>

> A turns ratio of 3:2 gives an impedance ratio of 9:4 or

> about 100:44.44.

>

> A turns ratio of 4:3 gives an impedance ratio of 16:9 or

> about 100:56.25.

>

> A turns ratio of 5:3 gives an impedance ratio of 25:9 or

> about 100:36.

>

> A turns ratio of 5:4 gives an impedance ratio of 25:16 or

> about 100:64.

>

> etc.

> - --- In loopantennas@yahoogroups.com, John Popelish <jpopelish@...> wrote:
> On 03/08/2013 02:45 AM, ceo@... wrote:

After installing a 11 MB Chinese font file :-),

> > I'm using http://www.mingtek.tw/pdf/B-4.pdf

I find there are _five_ transfomer diagrams there

- you don't tell us which you use.

> > 100ÃÂ© balanced fed into pins 14 and 16

Your pinouts only make any sense for two of them:

> > 50ÃÂ© output to R8 from pin 1

> > pins 4 and 13 grounded

HN1664CG & HN1674CG. Let's hope it's the former.

> Transformer (and balun) impedance ratios are proportional to

John has explained it well. A centre-tapped (usually

> the square of the turns ratios. Using only half of a

> winding reduces the impedance by a factor of 1/4.

bi-filar) balun transformer is the way to connect an

FM dipole made of 300R twin feeder to the 75R coax

antenna input on the typical home hi-fi.

4:1 impedance ratio: half the voltage at 2x the current,

or 100R:25R. Also, if you are unlucky enough to have a

HN1674CG, you have the inductance of a half-connected

and un-specified current balun to add into the mix.

If you had access to all the transformer terminals

you could achieve a 3:2 or 4:3 turns ratio for one

of John's 'near miss' matches, but I imagine these

commodity transformers are inscrutably potted in

resin for their Ethernet isolation task.

However, does it matter for a receiver? There are

no damaging consequences of mismatch as in a Tx,

the coupling loop probably isn't 100R either, and

the only result might be transmission line

transformer effects when the cable gets to be

near an odd number of quarter wavelengths.

Better an antenna that works (mostly) than none.

Regards. LenW - On 03/08/2013 04:30 PM, hejsek2005 wrote:

> John has explained it well. A centre-tapped (usually

I think you should compare the result you get with one of

> bi-filar) balun transformer is the way to connect an

> FM dipole made of 300R twin feeder to the 75R coax

> antenna input on the typical home hi-fi.

>

> 4:1 impedance ratio: half the voltage at 2x the current,

> or 100R:25R. Also, if you are unlucky enough to have a

> HN1674CG, you have the inductance of a half-connected

> and un-specified current balun to add into the mix.

>

> If you had access to all the transformer terminals

> you could achieve a 3:2 or 4:3 turns ratio for one

> of John's 'near miss' matches, but I imagine these

> commodity transformers are inscrutably potted in

> resin for their Ethernet isolation task.

>

> However, does it matter for a receiver? There are

> no damaging consequences of mismatch as in a Tx,

> the coupling loop probably isn't 100R either, and

> the only result might be transmission line

> transformer effects when the cable gets to be

> near an odd number of quarter wavelengths.

>

> Better an antenna that works (mostly) than none.

these Ethernet baluns with a home made unit made of 5

strands of something like 30 ga wire wrap wire, twisted

together and wrapped a few times through a ferrite core.

Then you connect all those strands in series. The first 4

of them form the 100 ohm primary and the last 3 form the

secondary. Two of those strands are part of both the

primary and also part of the secondary. The end of strand 2

(which is the beginning of strand 3) is the ground point.

This arrangement gives you a 16:9 = 100:56 ohm balun.

--

Regards,

John Popelish