Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: Prev 1 [2]   Go Down
  Print  
Author Topic: What's wrong with stored energy, exactly?  (Read 4570 times)
STAYVERTICAL
Member

Posts: 854




Ignore
« Reply #15 on: October 30, 2012, 08:57:03 PM »

Hi JAHAM2BE,

I just had another thought about your strip based magloop.
You will probably wonder what the equivalent tubular dimensions are for a particular strip of copper.
From my investigations online, and looking up books and articles, it seems that the equivalence is around 4 to 1.
So a 200mm wide copper strip would be about equivalent to a 50mm diameter tube.

I came across a book online called:
"Electrically small, superdirective, and superconducting antennas" (Wiley publishing 2006) , by R.C. Hansen which had this and more information.

Another thought is that if your loop is too large for resonating with a practical capacitor, it is possible to add a parallel loop.
If you space a 20mm tube for example, around 4 inches (100mm) away from the first loop and put it in parallel, the overall inductance will drop.
This is of course the well known inductors in parallel formula.
You will then be able to use a higher value capacitance to resonate it at the high end of the frequency spectrum.

I tried this, and it did work, giving me 10 metres where I was only able to get to 12m formerly.
Unfortunately it will mean your capacitor must have enough excess to resonate the combo loop at your lower end (40M).

The reason I said to space 20mm tubing 100mm is that this spacing is one that has been found to minimise proximity effect but still give good results.
Technically, it should be 80mm, but I used 100mm and it worked well.
The spacing is determined by the number of parallel loops and their diameter tubing.
In a nutshell - for two parallel loops the centre to centre spacing should be 4 times the loop tube diameter.
I obtained this information from:
"The proximity effect in systems of parallel conductors and electrically small multiturn loop antennas" (Office of Naval Research) by Glenn Smith.

So, as you can see there is certainly hope for getting a wide range of bands with a little ingenuity.

Happy experimenting,

73 - Rob
« Last Edit: October 30, 2012, 09:21:30 PM by STAYVERTICAL » Logged
JAHAM2BE
Member

Posts: 276


WWW

Ignore
« Reply #16 on: October 31, 2012, 04:08:25 AM »

I just had another thought about your strip based magloop.
You will probably wonder what the equivalent tubular dimensions are for a particular strip of copper.
From my investigations online, and looking up books and articles, it seems that the equivalence is around 4 to 1.
So a 200mm wide copper strip would be about equivalent to a 50mm diameter tube.

W9CF has done some of the most thorough analysis of strip conductivity at RF that I've seen available freely online. At the following links you can see, respectively, a formula and approximation for the diameter of a tubular conductor having equivalent RF resistance to a given flat strip conductor; a Java applet that computes this equivalent diameter with (it appears) more accuracy than the approximate formula; and a formula for the diameter of a tubular conductor having equivalent inductance to a flat strip conductor.

http://fermi.la.asu.edu/w9cf/articles/equiv/index.html
http://fermi.la.asu.edu/w9cf/equiv/index.html
http://fermi.la.asu.edu/w9cf/articles/conform/index.html

From memory, for some various and plausible width:thickness ratios that I tried, I got equivalent-resistance diameters anywhere from 1/5th to 1/3rd of the strip width, which is in the ballpark of the 1/4th figure you mention from your reference.

The way to use W9CF's resources to design a small loop is: pick the desired sheet dimensions (width, thickness, and length), compute the equivalent tubular diameter for resistance, calculate the RF resistance of the equivalent-resistance tube having the previously-chosen length (by using either a small loop calculator program/formula, or by directly computing the RF sheet resistance of the equivalent-resistance-tube's surface area - they will be the same), compute the equivalent tubular diameter for inductance, compute the RF resistance of the equivalent-inductance tube, make a note of the additional RF resistance between the equivalent-inductance tube (which will be lower) and the equivalent-resistance tube (which will be higher), and finally to use a small loop calculator program or formula with the equivalent-inductance tube dimensions and the just-calculated additional RF resistance to determine final values for the loop efficiency, resonating capacitance, loop current, etc.

By doing the above, you will at the end be calculating the small loop values for a tubular conductor that has the same RF resistance and the same inductance as the flat sheet, according to W9CF's approximations (which he says were documented at least in part in previous literature; see http://fermi.la.asu.edu/w9cf/articles/update/index.html).

If you find my explanation difficult to follow, N3OX also explains the same thing here: http://forums.qrz.com/showthread.php?317014-Magnetic-Loop-Calculations&p=2361920#post2361920

Having gone through the quite interesting calculations as described above for my particular strip dimensions, and comparing the efficiency at 7 MHz of the flat strip used flat vs. the flat strip rolled into a tube, I can say that the difference was small enough not to matter to me - less than half a dB (in favor of the tubular conductor) as I recall. That possible half-dB improvement didn't outweigh the increased constructional challenges of making the tube-based loop.

This all of course ignores any calculation of capacitor losses. Electromagnetic field simulation or field strength measurements of the completed loop-and-capacitor assembly seem to be the only way to get an idea of these.

Your suggestion about an additonal loop is indeed interesting, but I feel the issues surrounding multi-turn or multi-conductor loops are a bit too hairy to tackle right now. Since coming to the recent realization that it's actually entirely feasible to home-fabricate gigantic-diameter copper tubing by rolling flat sheet metal into tubes, I'll probably (one day) pick up on the large-diameter-tubing loop idea before I venture into multiple conductors.

The nice thing about starting with the flat-strip design is that I can re-use the same material to form copper tubes, whenever I feel like tackling that.
Logged

STAYVERTICAL
Member

Posts: 854




Ignore
« Reply #17 on: October 31, 2012, 12:01:46 PM »

Thanks for the info and the websites.
I will certainly investigate them and see what else I can learn (a lot I am sure).

A few notes from the trenches:

From my experiments, it does not seem to matter much whether you use strip or tubular conductors, taking into account the equivalences.
My large copper strip loop (almost a foot wide strip) worked well.
It had about 90 percent greater bandwidth at 12m than a single 20mm conductor.
At the lower end (10Mhz) it had about a 20 percent wider bandwidth.
So there was a difference, but I determined the difference was not great enough to warrant the extra effort in my case.
This was because I already had a pretty efficient system/capacitor to begin with.

As you imply, the benefits of large diameter conductors/strips is primarily evident at 40m and 30m.
Above that the incremental benefit is not that great if you have an optimised 20mm copper tubular system.
From my experiments the two things which deliver the greatest increase in efficiency are:

1. Using copper rather than aluminium.
2. Making the loop circumference/span greater.

Also, I found that it is best to avoid wires for interconnecting capacitors.
Using 12mm copper tube to directly connect the capacitor to the main loop/loops gave me far better results.
It is a case of the weakest link causing the whole structure to become inefficient.
That's about it for small loops.
They really are just a big tuned circuit with losses in the form of R.F. radiation (in addition to the other dissipative losses).

Happy experimenting,

73 - Rob
Logged
JAHAM2BE
Member

Posts: 276


WWW

Ignore
« Reply #18 on: October 31, 2012, 09:50:46 PM »

My large copper strip loop (almost a foot wide strip) worked well.
It had about 90 percent greater bandwidth at 12m than a single 20mm conductor.
At the lower end (10Mhz) it had about a 20 percent wider bandwidth.
So there was a difference, but I determined the difference was not great enough to warrant the extra effort in my case.
This was because I already had a pretty efficient system/capacitor to begin with.

Your results bring to light an interesting subtle aspect. The strip exhibits wider bandwidth. Why? Because it has less self-inductance per unit length than the tube. In other words, it has less inductive reactance, leading to lower Q=X/R and broader bandwidth.

And as we see in this thread, less inductive reactance means less energy storage and - that's right - less loss!

So it seems for a given conductor length and sufficient strip width to ensure good conductivity, a strip is actually better than a tube! Shocked
Logged

W5DXP
Member

Posts: 3546


WWW

Ignore
« Reply #19 on: November 01, 2012, 04:36:28 AM »

So it seems for a given conductor length and sufficient strip width to ensure good conductivity, a strip is actually better than a tube! Shocked

Would that also apply to magnetic loop antennas?
Logged

73, Cecil, www.w5dxp.com
The purpose of an antenna tuner is to increase the current through the radiation resistance at the antenna to the maximum available magnitude resulting in a radiated power of I2(RRAD) from the antenna.
N3OX
Member

Posts: 8853


WWW

Ignore
« Reply #20 on: November 01, 2012, 07:26:46 AM »

And as we see in this thread, less inductive reactance means less energy storage and - that's right - less loss!

You know, this is probably an example where lowering the reactance by itself probably doesn't lead to LESS loss, at least not in any meaningful way.

If the loss resistance and enclosed area (and hence the radiation resistance) of the two loops is the same, and they only differ in their reactance because of the extra self-inductance of the conductor, then the balance of loss and radiation will be the same.  I don't think the resistance of a good air or vacuum capacitor changes much with its reactance.

From the Jennings vacuum variable literature:
Based on actual tests, the ESR value is not affected
by change in capacity, other parameters being fixed. The value
of ESR varies over a range of 2 to 20 milliohms from 2 to
30 MHz

So tuning out some extra reactance with a good vacuum cap, for example, is simply irrelevant to the loss.  The cap adds 2-20 milliohms in series, period, no matter the capacitance.

----------------

The bandwidth will be improved with lower conductor self-inductance.  But the equivalent circuit, at least for a small loop, is more like your series LCR circuit

o--Rloss--Rrad--Cap ESR--Loop Inductance--Conductor Self-Inductance--Tuning Capacitance--o

Rrad and cap series resistance are probably about constant for a good cap, Rloss (the conductor) is chosen by choosing strip width for good efficiency, so let's say it's the same to put things on equal footing.  So there's nothing you can do with the conductor self inductance to change the power dissipated in the combination of Rloss+Rrad+Cap ESR at resonance.

In a purely series circuit the only reason to get more loss with more loading inductance is the highly non-ideal nature of an inductor; seems hard to get a weakly radiating inductive structure with a Q exceeding 1000.  But if you talk about keeping resistance equal and lowering conductor self-inductance, and you don't have some kind of parallel or distributed transformation like my example, all you will change is bandwidth.  

These bandwidth issues are kind of insidious for people who like to use bandwidth as a proxy for efficiency.  It's plausible that you can increase conductor loss resistance and self-inductance simultaneously, leading to a narrower bandwidth magloop with more loss.  It doesn't have more loss *because* you added more inductance; it has more loss because you added more resistance.  But the added conductor inductance masks the bandwidth widening you'd expect if you just added loss resistance.
Logged

73,
Dan
http://www.n3ox.net

Monkey/silicon cyborg, beeping at rocks since 1995.
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #21 on: November 01, 2012, 09:31:08 AM »

If the loss resistance and enclosed area (and hence the radiation resistance) of the two loops is the same, and they only differ in their reactance because of the extra self-inductance of the conductor, then the balance of loss and radiation will be the same.  I don't think the resistance of a good air or vacuum capacitor changes much with its reactance.

From the Jennings vacuum variable literature:
Based on actual tests, the ESR value is not affected
by change in capacity, other parameters being fixed. The value
of ESR varies over a range of 2 to 20 milliohms from 2 to
30 MHz

So tuning out some extra reactance with a good vacuum cap, for example, is simply irrelevant to the loss.  The cap adds 2-20 milliohms in series, period, no matter the capacitance.

The resistance stabilitly occurs because the capacitor is compact and has the same path length through the component, and nearly the same current distribution and density in conductors, regardless of capacitor setting.

However, the capacitor Q changes. :-)

It would take a good sized book to cover this stuff.

:-)

73 Tom
Logged
JAHAM2BE
Member

Posts: 276


WWW

Ignore
« Reply #22 on: November 01, 2012, 03:25:19 PM »

The resistance stabilitly occurs because the capacitor is compact and has the same path length through the component, and nearly the same current distribution and density in conductors, regardless of capacitor setting.

However, the capacitor Q changes. :-)

It would take a good sized book to cover this stuff.

Might you suggest any references on capacitor analysis and design?

Here's a 2009 article that indicates some analysis techniques for inclined plate capacitors: http://www.icrepq.com/ICREPQ'09/452-barrachina.pdf
Logged

W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #23 on: November 01, 2012, 05:03:48 PM »

I doubt you will find a detailed article on capacitor design, unless someone did their doctorate thesis on the topic. It is probably like RF inductor design, although there was one good inductor design paper many years ago. 
Logged
Pages: Prev 1 [2]   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!