A 60' flat top for 160 and 75m!

[Bacon, WA3WDR]

I think that bending the elements back on themselves like that is called 'linear loading'.  Someone suggested it to me some years back, and the only caveat was that you should separate the folded part by at least a few inches, or there is too much capacitance between the parallel sections.

[Tom WA3KLR]

Frank said “the coax jumper gets noticeably warm after repeated long "old buzzard" transmissions”.
Let’s analyze the 50 Ohm coax end of his antenna system now:

The 1:1 baluns probably contain 4 – 8 feet of miniature coax inside.  Then there is the 10 feet of RG-8 to the tuner output port.

I stated before that the antenna feed point impedance at 1.9 MHz. is 3.21 – j1548 Ohms.  At the other end of the 65 feet of 450 Ohms ladder line the impedance becomes 2.51 – j738 Ohms.  Plugging this load impedance into 8 feet of RG-58A (to simulate what may be inside the 1:1 balun) yields 0.33 dB of loss and an impedance of 1.96 – j628 Ohms.

Plugging that load impedance into 10 feet of RG-8A yields 0.29 dB of loss and impedance of 1.21 – j477 Ohms at the tuner port.  Yikes.

Frank runs a 4-1000A rig, so let’s assume that the output is 400 Watts average under modulation.  The 10 foot RG-8 coax with 0.29 dB loss dissipates 26 Watts.  Under an old buzzard transmission, no wonder it gets warm.

O.k., so 374 Watts is being delivered to the balun.  The balun having 0.33 dB of loss dissipates 27 Watts. Again, no wonder it gets warm.

The r.f. voltage on the balun coax is 18 kV r.m.s. on voice peaks (1500 Watts PEP) and this is also 25.5 kV peak.  Yikes.

The tuner sees 1.21 – j477 Ohms.  At 400 Watts, this is 18.2 Amps and on voice peaks this is 36.4 Amps.

As far as the ladder line, 600 Ohms is better, because the highest loss is at 160 meters and the feed point impedance there is 1550 Ohms, so the closer the Zo to 1550 Ohms, the lower the SWR and loss.

Dimensions with #10 wires = 7.6” spacing for 600 Ohms.
6” yields 572 Ohms though.

[K1JJ]

Wow - We be talking some low impedances, big currents, high voltages and big losses -

This is a job for SUPER-CONDUCTOR-MAN !!


T

[Bacon, WA3WDR]

SUPER-CONDUCTOR-MAN !!

-he ain't just some ordinary Mho!

[The Slab Bacon]

Tom (klr) FWIW the balun is one of those ones with RG-8 coax wrapped around 4" pvc pipe (I dont remember how many turns). I didnt use a toroidal one for this ap as I figgered it would be too brutal for it. I dont know if that changes your calculations or not.

I would feel that at these low impedances and brutal feedline currents, resistive loss in the feedline would be more of a problem than the swr losses.

FWIW I did a lot of experimenting with baluns a few years back. I built many different types and swept them to find some interesting results: I found that most of the 4:1's were much more accurate than any of the 1:1's. The 1:1's would get pretty squirrely when you got away from the charactoristic impedance, but the 4:1's seemed to give a pretty true 4:1 impedance ratio what ever was on either end. It was a pretty interesting experiment, i went from 1.5 - 15 Mc, and 5 to 500 ohms. The 1:1's seemed to be happier if you sinned to the low side but were still somewhat squirrely.

I knew that a 4:1 would be just about murder at these low impedances and currents, so I elected to use a 1:1 for the 160 testing. Coax cable also seems to handle the sin of high swr at low impedances better than high impedances. If one trys to use 50 ohm coass on a high impedance antenna the loss is unbearable, but it is not all that bad on a low impedance application. the capacitive reactance in the coax itself just eats you alive in a high voltage / low current scenario. Since my situation is just the opposite (lower voltage / highte current) I kinda feel that the warming of the coax in my scanario may also be a reistive loss problem.

                                                                       The Slab Bacon

[Steve - WB3HUZ]

Using Tom's numbers below (1.21 – j477 Ohms) and running them through W9CF's tuner applet shows the following losses.

T type tuner: (2) 400 pf Caps (Q=2000), 50 uH Coil (Q=shown as below), freq= 1.885 MHz, tuned for minimum loss

Q          Loss (dB)
100         9.8
150         8.1
200         7.0


So, unless that coil in the tuner is really high-Q, at least 6 dB of loss can occur there. Frank, did you say the tuner isn't getting warm? If so, something doesn't add up.

[Tom WA3KLR]

I mentioned previously that the tuner inductor coil would have to have high Q, greater than or equal to 600.  That was in the "160 Meters Transmission Line Study  - SWR and Losses" topic.  All of the changes proposed help to raise the impedance the tuner will see.

Frank,

You want to think of your 1:1 coax balun as just a line isolator balun between the transmitter output and the tuner with the system re-configured.  The tuner receives the ladder line directly and transforms the impedance to 50 Ohms or near that.  The isolation provided by your PVC balun allows the tuner to float r.f.-wise and even though it is a single-ended design, it should work o.k. with the balanced line, since it is floating.  With the balun coax and any other 50 Ohm jumpers actually operating at 50 Ohms impedance between the transmitter and tuner, there is no problem with 400 Watts or 1500 Watts at 1.9 MHz.

RG-8 (0.4” o.d.) wrapped around 4” PVC (4.5” o.d.) is approximately 4.9”diameter coax turn = 15.4” per turn.  Let’s say you have a little under a foot of PVC wound with coax.  This is 28 turns and is a total of 36 feet of coax.  Along with the other 10 foot piece is a total of 46 feet of RG-8.

The SWR the RG-8 coax is seeing at the ladder line point is presently 4360:1.  The transmission line program says that 46 feet of RG-8 with the load impedance from the end of the ladder line of 2.51 – j738 Ohms yields 12.9 dB loss.  (I don’t think the situation is quite this bad, perhaps you don’t have 28 turns?) 

If you had a perfect 1:1 match on that 46 feet the loss would be 0.11 dB (or 10 Watts lost out of 400 Watts applied).  Think of the r.f. energy as reflecting back and forth 117 times on that 46 feet of coax before it is essentially transferred, in the present configuration.  That is where the loss multiplication comes from.  Remember the SWR on the coax is 4360:1! 

Most of the loss on the 50 Ohm coaxes will disappear when they are between the transmitter and tuner with the tuner matched to 50 Ohms, operating with an SWR of 1:1.  Your tuner will see a higher impedance than before and that will lower tuner losses also.

[WA1GFZ]

Why not make a balun with heavy wire pair like a big common mode air wound choke. say it was wound with #6 stranded wire pair. The resistance would be a lot lower. It would be longer since #6 stranded is almost as big as coax.

[The Slab Bacon]

Why not make a balun with heavy wire pair like a big common mode air wound choke. say it was wound with #6 stranded wire pair. The resistance would be a lot lower. It would be longer since #6 stranded is almost as big as coax.

Hmmm........... I like when you talk dirty!! Tell me more. I have about 150' of #8 stranded wire laying around and looking for something to do. Is that like a parallel pair wound on a round form of some type?


Tom, I think what you are telling me is to ditch the coax isolator and wind up some kind of strapping 1:1 broadband transformer. The tuner input is running at 50 ohms (or somewhere close to it) As I tune the trans up into a 50 ohm dummy load and switch over to the antenna without changing the transmitter's final settings.

                                                        The Slab Bacon

[K1JJ]

Tom,

While we're at it...

I'm having trouble finding a matching method to go from my 75 ohm hardline in the shack into a 50 ohm load. This is for the 2el 75M quad arrray. I want 50 ohms flat throughout the shack.

 Right now at the shack end of the 75 ohm cable I see 45 ohms, j22.  I THINK that's inductive j cuz of the way the MFJ-259B climbs in j as the freq goes higher, but it might be capacitive - cuz of the squirrelly influence of the reflector, etc. But I'd say 75% probability it's inductive.

I've tried longer and shorter 75 ohm and 50 ohm combinations of coax cables, but each time I get it close to 50 ohms I see some j in there.

Using a vac cap and/or inductor, can you suggest a circuit config and values to match 50 ohms >>  to  both 45 ohms  plus OR minus j22  from the 75 ohm cable?

Or maybe there is a specific combination of RG-213 and RG/11 lengths that will do it.

BTW, this is for 3795 khz.

I already tried JSing in a series variable cap alone and no luck. I also tried tapping down an inductor and NG.  I'm thinking that maybe some kind of L network would be better, but want to get a simulation first to stop wasting time here...

****  Or a better point to start from may be when I add 8' of RG-213 to the 75 ohm line. I get 60 ohms +-j22.  I will need that coax anyway to connect to the network. I've been told that matching 50 ohms to another load that is close in impedance (40-60 ohms) is not easy in the real world.

Tnx.

T

[Tom WA3KLR]

Frank,

No, I like the coax balun very much, stick with it.  It is good.  I would like to know how many turns it is though.  Just put it between the transmitter and tuner and it's loss will go away.  There might be even a few more dB loss in that balun right now than we originally thought.  Put the ladder line right to the tuner.

I think your shack is in the basement?  Hang the tuner from the rafters or something. 
I'm wondering if just a separate fixed L-C tuner might work for the portion of 160 you work.  A big inductor with 1/4" copper tubing would be lower loss than a roller inductor.

Tom,

You should have started a new topic.  I'll look at it later, getting tired of this for the moment.

Bandwarmers,

What luck, my wife is staying with her sister tonight.  See youse on 3685 or wherever at midnight, I'm ready.
What luck, solar storm.

[WA1GFZ]

3685 cool. I tuned up on 35 last night with a 2:1 SWR 85 will be better.
I suggested a wire balun because the center conductor of coax is pretty small. I would stay with coax if the tuner is between the balun and antenna. Now to find the best tuner configuration with the least loss.

Tom vu a Tee tuner is the best for matching coax. 2 inductors and 1 cap to ground will be a low pass. A PI network will also work and be low pass. 

[The Slab Bacon]

Tom, for now I have to keep the balun after the tuna, if you saw the ergonomics of the shack you would understand why things are layed out the way that they are. At least it is working respectably for now.

Frank, I am very interested in the idea of a #8 wire commom mode choke balun. The #8 wire would have negligable resistive loss at those current levels. I think that I have a good feel for what you are talking about. Do you have a quickie recipe for one??
                                                          The Slab Bacon

[WA1GFZ]

Frank just take two lengths of wire the same length as the coax. or take 4 lengths to get even lower resistance. The key is to get enough inductance. Tom Vu put a table in a while ago for the reactance/ inductance you will need. fc

[The Slab Bacon]

Frank just take two lengths of wire the same length as the coax. or take 4 lengths to get even lower resistance. The key is to get enough inductance. Tom Vu put a table in a while ago for the reactance/ inductance you will need. fc

Frank
         Is what you are saying: Take 2 lengths of #8 wire, wrap them parallel (bifilar) around something (like 4"pvc pipe) somewhere around the same number of turns as the coass balun. Attach the coass to one end and the ladder line to the other?? If that is the case I will twist one up this weekend and sweep it to see how it reacts.

                                          The Slab Bacon

[WA1GFZ]

It will only work between the TX and tuner. Ladder line voltages will flash it over.
But yes just use the same number of turns for both conductors. 1 to the center conductor and the other to the shield

[The Slab Bacon]

With that said, it would be easy enough to build. I have everything laying around to do it.

I may be lacking a little knowledge in theory here, but, if the balun is acting as a 1:1 and the actual feedline impedance of the antenna is very low (a few ohms) how does the voltage get that high?? Is it simply a finction of the combined feedline / antenna lenght putting that point at the high voltage node? If it is 1:1, I would think that the voltage would be the same at the coass end of the balun as well. I may still twist one up this weekend to experiment. QTF?? Also if the voltage is that high, why doesnt the coass flash over?

                                                                     the Slab Bacon

[WA1GFZ]

I would think the voltage low on 160 but 75 the total length of 1/2 antenna plus feed line is greater than a quarter wave so the voltage will be higher. I'm not sure of the voltage due to reactance of the choke vs antenna.
Remember the old B&W air  balun with two inductors side by side. Maybe that is worth trying if it flashes. The balun needs good coupling between windings to force current balance.

Here is a thought. Run rig coax to a parallel tuned circuit large wire/tube inductor like the primary of a link tuner. Then just tap in the feeder between 1 or more turns.
I would think you would need a reactance of about 200 ohms for the inductor and tuning cap for a good load on the transmitter. Eliminate the balun to eliminate the loss. Might even series tune it. Maybe a second cap across the feeders.
cool science project.

[K1JJ]

Tom vu a Tee tuner is the best for matching coax. 2 inductors and 1 cap to ground will be a low pass. A PI network will also work and be low pass. 

Franz,

Tom sent me some info that should do the job using a variable cap to ground and a small inductor. I will make up a small box and mount a vacuum cap. This will keep it adjustable.

The reason I need 50 ohms flat is cuz of the new linear with the RF negative feedback. The flatter the ant load, the more feedback I can pour on before it takes off. When switching ants, I need them real close to the same.


Tom's solution:

"Another solution is a small L-C network added:

Coax (as is) gets about 623 pf. to ground, then 0.71 uH in series to output.
Should tweak right to 50 + j0.

This is more flexible to changing conditions in the future than just the cap. to ground on the coax."


Later -

T

[WA1GFZ]

Yup L network will work to convert 42 to 50 ohms. I suggested a T so if it ever goes the other way you have it covered.

[Vortex Joe - N3IBX]

Frank,
        You put out a consistently good signal on 160M, as you do on 75M. If you can hear 'em, you can work 'em with the same signal strength you copy them. In anybodys' book, that ain't too shabby!

Looking forward to hearing a lot more of you on 160 Metros this season.

Very Best Regards,
                         Joe Cro N3IBX

[WA1GFZ]

Joe,
We are just egging Frank on to produce the best signal he can with a small antenna. He has a nice test bed we can all learn from.

[Tom WA3KLR]

Frank,
You mentioned that I had stated 18 kV on the balun and that did strike me as odd since we know the impedance is progressively lower as you get to the transmitter end of the system; good catch.  So I looked things over again.  (You do have loss of power in the tuner and coax before getting to the balun also.)

Re-checking the numbers I found that in some instances I did not enter the negative sign in the “Load reactance” box of the transmission line calculator program.  This can make a big difference!  The impedance at the tuner is even lower than previously stated, the ladder line loss is higher, and the balun loss is lower.

With 1500 Watts, the most voltage I see now is 10 kV r.m.s. at the feed point of the dipole.  The balun has 2700 Volts r.m.s., this occurring at the ladder line end.

I’ve been presuming you have 36 feet of RG-8 wrapped up in the PVC balun.  Where the balun meets the ladder line, the impedance is 3.5 – j 222 Ohms.  For comparison, at this load impedance this is what 4 types of 50 Ohm coax losses are for 36 feet at 1.9 MHz.:
RG-8 has 4.4 dB. of loss,
RG-8 foam has 3.5 dB. of loss,
9913 has 3.0 dB. of loss,
¾” 50 Ohm hard-line has 1.1 dB. of loss.  A tough situation, and on 40 meters, the situation is much worse, 13 dB. of loss with 600 Ohm ladder line!

The impedance leaving the balun for the tuner is 0.78 – j41 Ohms.  If you deliver 400 Watts average to this point, the current is 22.6 Amps.

65 feet of the 450 Ohm crappy brown ladder line is 10.5 dB. loss.   If it were the beefy 600 Ohm line, 5.3 dB of loss.

Checking 40 meters, feed point impedance will be higher, countering higher transmission line losses.  The net result:
65 feet of the 450 Ohm crappy brown ladder line is 4.0 dB. loss.   If it were the beefy 600 Ohm line, 1.2 dB of loss.

(40 m.) I’ve been presuming you have 36 feet of RG-8 wrapped up in the PVC balun.  Where the balun meets the ladder line, the impedance is 352 + j2567 Ohms.  RG-8 has 10.4 dB. of loss with this load impedance.

The PVC balun may have 105 uH of inductance if it is 28 turns, but the 36 feet is a lot of loss.  If you have to operate with the balun between the tuner and ladder line for a while, a toroidal core with 5 feet of RG-8X would have less loss.  Hopefully you could get at least 25 uH.  More is better but you can’t stand more than 4 or 5 feet of RG-8X.  At 40 meters, the 5 feet would be 0.33 dB loss with the 450 Ohm line, not quite as low as putting the coax balun at the transmitter output.

Magnetics is one of my weak areas, Frank ‘GFZ can probably pick a core.  The only flux in the core is the imbalance current.  How much would there be??  The PVC balun is o.k. for when you can install it between the transmitter and tuner.

I found that my 75 m. lazy C dipole had been at 40 feet.  I bought it down 5 feet to 35 feet to match Frank’s present dipole.  My dipole’s radiation is no longer 1.5 dB better at 160 m., just 0.2 dB better, but the higher feed point impedance reduces losses through the rest of the system.  The AHE antenna radiation is 0.5 dB better at 40 m.  I picked 60 degrees elevation at 45 degree azimuth for comparisons.

To summarize the losses, except for the tuner:

Ant.                         AHE 40 m. folded dipole                    KLR Lazy C 75 m. dipole
Freq. MHz.              - - - - 1.9 - - - - ]    [  - - - 7.29  - - - -           1.9           7.29
Rel. ant. loss               0.2                            0                         0 dB.       0.5 dB.
450 window line  10.5  10.5   -     -       4.0   4.0    -      -
600 ladder line        -       -   5.3   5.3      -      -     1.2    1.2         1.9           0.2
Balun tuner out    4.4     -     7.2     -     10.4    -   13.0    -       
Balun at tx out       -     0.1    -      0.1    -      0.2     -     0.2         0.1           0.2
Sub-total losses  15.1  10.8  12.7   5.6  14.4   4.2  14.2  1.4         2.0           0.9
Tuner                 ?         ?       ?     ?        ?     ?     ?       ?            ?             ?

The biggest incremental change is to first get the balun on the transmitter output.  This is a 4.3 dB improvement for 160, and a 10.2 dB improvement on 40.

Changing the ladder line only is a little help on 160, none on 40.  What happens here is that you lower the ladder line loss, but the change in impedance to the coax balun causes the balun loss to increase.

Changing the balun location and ladder line gives a 9.5 dB. improvement on 160 and 13 dB. on 40.

Ultimately you get a little over 13 dB on 160 and 40 if you make all the changes.

I’ve ignored 75 meters.  The antenna program says the folded dipole is 0.3 dB better gain!?

Ant. at 75 m.       AHE 40 m. folded dipole     KLR Lazy C 75 m. dipole
Rel. ant. loss                        0                                   0.3 dB.
450 window line   2.9      2.9     -       -
600 ladder line        -         -     1.0    1.0                     0.12
Balun tuner out     5.5       -      8.8     -     
Balun at tx out       -       0.13     -      0.13                  0.13
Sub-total losses     8.4     3.0    9.8    1.1                    0.55
Tuner                    ?          ?       ?       ?                      ?

Moving the balun to the transmitter output is 5.4 dB. improvement.  Just switching the ladder line is 1.4 dB worse, because the balun loss increases more.

Changing the ladder line and balun location is 7.3 dB, improvement.

7.8 dB. ultimate improvement possible with all changes implemented.

Final summary – you get most of the improvement with just the balun and ladder line changes which is 9.5 dB. on 160, 7.3 db on 75 m., and 13 dB. on 40 m.  Changing the antenna configuration after that gives little improvement to 75 and 40, but 3.6 dB more on 160 meters.

So says the p.c.!

[WA1GFZ]

After giving different core materials some thought I think it may be a problem at high power.
Say your transmitter makes 1000 volts peak to peak. 0 volts on the shield and 1000 V P-P on the center conductor. Now put it through a balun with 500 volts p-p on the shield at the output. This means you will need a stack of cores that will support 500 across the stack.
I helped QIX pick the core material and size for the class E stuff when he first got into it.
We all found it took 6 cores for 130 volt peak modulation on 160 meters. The drain voltage swings up to about 350 volts The cores stay nice and cool under those conditions.( meaning low loss) Extrapolate that out for Franks and he will need 18 to 24 cores. We used type 43 material that works great at the low end of HF.
Above 40 meters I would use type 61 with a perm. of 125 down from 800 (6 times 18)
So bottom line it would cost a fair amount of cash to build a balun with ferrite that would be low loss. How about RG17, RG393 or 1/2 inch andrews. OR #10 teflon wire pair with teflon tubing over eack conductor.
You quickly see this balun needs to be on the low voltage side of things or it will grow even more.
I wonder if the balun could be eliminated by driving a parallel tuned circuit with the feeders tapped off the coil. Maybe a second cap across the feeders if required.
Look at like a link tuner backwards with the output 2 taps. The inductor will be a small value to support 50 ohms on the primary. ( like the primary of a link tuner) make the conductors large say 1/2 inch copper tubing to handle the current. A vacuum variable to handle the current.
KISS means low loss

[The Slab Bacon]

Guys,
        I somewhat have to keep the basic mechanical configgeration the way it is or change the layout of my house. It would really be difficult to get the ladder line any closer to the tuna / operation position than it is. That is why I did things the way I did, it is really an issue of ergonomics / building construction.

I am pondering the idea of Frank's #8 common mode choke bakun on pvc pipe, that is easily doable and will definately reduce the heating / resistive loss. keeping in mind that the center conductor in RG-8 is prolly #14 at best.  I can possibly get the balun a few feet closer to the tuna, but nothing dramatic. Also have some LDF-450 hardline to feed the balun with as well. I have everything here to twist one up just for the hell of it.

Instead of a torroidal balun, how about some type of step down matching transfoma?? Since the antenna impedance is way lower than the feedline how about some type of step down transformer, just a thought.

I have been on again since last post, Still getting the same signal reports that I am giving. With the current configuration, it still sits at "if I can hear them, I can work them"!! anything from here is now onto optimization and improvements.

Per Tom's engineering calcs the whole thing is really in the Balun / feedline configgeration. this is going to make for some fun experimentaton!!

                                                                                The Slab Bacon