开云体育

Thanks Alan, and thanks to everyone else for sharing your knowledge and experience.

But now I must move on. The sun is shining and the coil is up in the air. I have optimised the antenna wire lengths for resonance and I'm looking forward to making some good contacts.

I have another similar coil and will continue the measurements on that when I get a moment.

Mike

--
Mike

I should add, strap a bar or plate where the LC is located to get a sense of the fixture R. Back out that mesured loss to get Rs fixture. You
can use that mesure to provide some correction.

In the same fixture, check the inserted 15 ohm only. The uniform loss should be ~ 8.5 dB.
It is possible that not all the current with the LC inserted is passing via the inserted R.
Since this is a distributed loss,

Hi Brian,

Thank you for your explanations.
That is what I love about this forum, I learn something every day.

The coil length explanation is really interesting and makes sense given your explanation.
As for the resistance issue, I felt a niggle as I wrote it. I would have put it down to skin effect. But I was not aware of the proximity effect. Makes sense.
Again thanks for the excellent explanations.

Cheers...Bob VK2ZRE

On Wed, Jul 5, 2023 at 09:26 PM, Bob Ecclestone VK2ZRE wrote:

But in your 2nd iteration, I notice the coil length is significantly longer
(156.8mm vs 80mm) with the same coil diameter (37.5mm) yet the inductance
remains the same at 97.8uH.
Surely the inductance should increase.

Bob, for a given number of turns, inductance decreases as the coil is lengthened due to lower coupling between turns. The optimizer maximizes Q while keeping inductance constant. It increased the number of turns just enough to compensate for the smaller coupling in the longer coil.

And finally, the Resistance (of the winding?) for the 2nd iteration is lower
than the 1st despite using a longer length of wire (6 vs 7.39).

Resistance is the RF resistance of the coil plus leads due to all causes. It is R in the R + jX model. Mike close-wound his coil with no space between turns. This minimizes coil length but maximizes the proximity effect. This magnetic effect bunches current inside a turn due to the proximity of adjacent turns. It increases RF resistance in the same way as the skin effect, by reducing available copper. The proximity effect is strong. I've never seen a close-wound coil where spacing the turns didn't increase Q even if it required more wire. The optimizer increases coil length to the point where any additional drop in the proximity effect is offset by the increase in wire resistance.

Brian

On Wed, Jul 5, 2023 at 02:55 PM, Mike wrote:

It's solid, not foam.

Actually foam, being mostly air, would be better. Another approach would be to drill many holes in the material, leaving a skeletal frame which, again, is mostly air. This is essentially what B&W does with their Miniductor stock, plastic rods are melted into the wire coils, stabilizing them physically while ensuring that the electric fields mostly encounter air.
73, Don N2VGU

On Wed, Jul 5, 2023 at 08:30 PM, alan victor wrote:

Hi Mike.

I suspect that the dielectric loss of your coil is secondary. A Qul of nearly
350 implies to me that the primary losses are likely copper wire R and
fixture.

A couple of useful things to do. Place a known R in series with the coil.
Deliberate de-Q and return to a measurement on Rs.

The number calculated should make sense. Example, suspect 7 ohms is the "real"
number, place a 10 ohm (low inductance R) not a WW, you should
see say 17 ohms. This is simply a sanity check.

Second, in looking at your solenoid coil, the L/D ratio looks wrong for
optimum Q. Something like L/D of 0.5 to 1 per Terman is better served for Q.
There is quite a few papers on this relation. But worth investigating if Q is
the big deal.

Hi Alan

I tried the experiment with a 15 ohm carbon resistor but it didn't produce the expected result. The calculated resistance went up from 7.5 ohm to 12.6 ohm.

I'm aware of the effect of L/D on Q. On a previous antenna I used a coil diameter of 48mm but I want to reduce the visual impact because our garden backs on to a footpath around a lake.

--
Mike

Hi Brian,

I do not pretend to be any sort of expert here, so I have probably missed something vital.

But in your 2nd iteration, I notice the coil length is significantly longer (156.8mm vs 80mm) with the same coil diameter (37.5mm) yet the inductance remains the same at 97.8uH.
Surely the inductance should increase.

Even if the coil windings were spaced in the 2nd iteration, the Wire Length indicates just more turns as do the Turns (105 vs 79).

And finally, the Resistance (of the winding?) for the 2nd iteration is lower than the 1st despite using a longer length of wire (6 vs 7.39).

Just wondering...

Cheers...Bob VK2ZRE

All,
What about radiation from the coil? Another basic definition for Q relates to the energy stored / energy lost. If energy is lost to radiation that needs to be considered as well.
Very interesting about the PVC. Thanks for that and other aspects of Q. This discussion makes me want to go back and tinker with this stuff but not now when I can still paddle my kayak:)

73, Pete

Thanks.

On Wed, Jul 5, 2023 at 12:30 PM, alan victor wrote:

Second, in looking at your solenoid coil, the L/D ratio looks wrong for
optimum Q. Something like L/D of 0.5 to 1 per Terman is better served for Q.

Rules of thumb are handy, but today it's easy to directly optimize a coil design. The first image show's Mike's coil. The second shows it after optimizing turns and coil length. The third is after optimizing turns, coil length, and coil diameter. Q could be further improved by using thicker wire. The optimizer uses the Differential Evolution algorithm.

I suspect that the dielectric loss of your coil is secondary.

Form loss shows the loss in the coil form as a percentage of total loss.

The coil program is here:



Brian

Hi Mike.

I suspect that the dielectric loss of your coil is secondary. A Qul of nearly 350 implies to me that the primary losses are likely copper wire R and fixture.

A couple of useful things to do. Place a known R in series with the coil. Deliberate de-Q and return to a measurement on Rs.

The number calculated should make sense. Example, suspect 7 ohms is the "real" number, place a 10 ohm (low inductance R) not a WW, you should
see say 17 ohms. This is simply a sanity check.

Second, in looking at your solenoid coil, the L/D ratio looks wrong for optimum Q. Something like L/D of 0.5 to 1 per Terman is better served for Q.
There is quite a few papers on this relation. But worth investigating if Q is the big deal.

On Wed, Jul 5, 2023 at 05:57 PM, Jim Lux wrote:

On 7/5/23 7:58 AM, Mike wrote:

After all that debate about PVC I've just discovered that my coil former is

actually ABS ...

Apologies!

Ohh. Black ABS? As in Drain Waste Vent (DWV) pipe?

The black is from carbon black as a pigment - yeah, as in resistors
Also, most DWV is what is known as "cellular" which means that the walls
are actually foamed ABS, with smooth ABS skin on them. Makes it nice and
lightweight, since it doesn't have to hold pressure.

Some is lossy, some isn't.

No I wouldn't use black for that reason. It's white (see photo further up this thread). It's solid, not foam.
--
Mike

On 7/5/23 7:58 AM, Mike wrote:

After all that debate about PVC I've just discovered that my coil former is actually ABS ...
Apologies!

Ohh. Black ABS? As in Drain Waste Vent (DWV) pipe?

The black is from carbon black as a pigment - yeah, as in resistors
Also, most DWV is what is known as "cellular" which means that the walls are actually foamed ABS, with smooth ABS skin on them. Makes it nice and lightweight, since it doesn't have to hold pressure.

Some is lossy, some isn't.

After all that debate about PVC I've just discovered that my coil former is actually ABS ...

Apologies!

--
Mike

Thanks for all your comments, Jim. I think I've seen different dissipation factor values for PVC depending on filler amount. I believe I used values for pure PVC, but I can't remember. In any event, with other dielectrics so readily available, I think PVC pipe should be avoided. By the way, cardboard is the lossiest dielectric my program lists. I included it for the oatmeal boxes builders traditionally used for crystal sets.

Some dielectrics like PVC are lossier at HF than UHF and some show the opposite behavior. That's why I think the microwave oven test can be very misleading.

Brian

On Tuesday 04 July 2023 02:54:32 pm Jim Lux wrote:

On 7/4/23 11:16 AM, WB2UAQ wrote:

Mike and Alan,
I measured the loss factor for PVC and it was about 0.005 or so at 1 kHz. I did more measurements at RF but will have to go back to old note books. I made a simple fixture to test a number of plastics. I rebuilt a coil that plugs into a BC-610 transmitter and from that sprung all kinds of questions about dielectrics. Many local guys are using 3D printers and using PLA, I think it is called. They printed out a plate of it for me to test. I would not say that all of this material is made equally and permitivity and loss factor are probably not well controlled. Won't be able to get back to this for awhile as it is Independence Day and a lot going on to take care of this week. Can only say right now that none of this plastic seemed to jump out as being a terrible dielectric. I also tested plate glass because back in the early days of radio capacitors for spark transmitters used what I think was regular glass available at that time.

PVC pipe has two potential problems as a coil form (based on experience
of folks building tesla coils, so 100-500 kHz):
1) it's hygroscopic so the dielectric loss varies with the water content
- nylon has the same problem, BTW.
2) most pipe is made from recycled material, and it can have metal
fragments and other debris in it. The specs for pipe are about holding
pressure and being "safe for drinking water", not dielectric properties.

Tesla coil folks solve #1 by drying it and then coating with something
like lacquer, glyptal, or urethane

#2 is a bit trickier - it's not like you can take a 4 foot long piece of
6" pipe and throw it in the microwave and see if it arcs or has hot spots.

One V/U dual-band antenna that I used for a while was totally enclosed in PVC. But it was a particular kind, my brother made it and he had to get this specific kind to not have the pipe affect the antenna much. I did indeed get out with it, reaching as far as 80 miles (!) with the output of a Baofeng...


--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, ?a critter that can
be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin

On 7/4/23 12:24 PM, Brian Beezley wrote:

On Tue, Jul 4, 2023 at 11:54 AM, Jim Lux wrote:

PVC pipe has two potential problems as a coil form (based on experience of
folks building tesla coils, so 100-500 kHz):
1) it's hygroscopic so the dielectric loss varies with the water content

I asked Google AI whether PVC is hygroscopic. Here, in part, is what it said
Polyvinyl chloride (PVC) is a hygroscopic powder material. Hygroscopic substances can take and hold moisture from the surroundings.
PVC is non-hygroscopic, meaning it does not absorb moisture internally into the pellet. However, moisture can be collected on the surface of the pellet.
Other non-hygroscopic polymers include polypropylene, polystyrene, and polyethylene.
I thought that was an interesting distinction. Google AI generates very useful summaries, but I don't fully trust it yet, so I searched onward. I found two more references that said PVC is not hygroscopic.
I'm sure your remark about Tesla coils is based on practical experience. Any idea why it might differ from what these references say?

Probably the structure and composition of PVC pipe - unlike a pure bar of PVC, pipe (and must plastic products) are a combination of PVC and some fillers to get the right mechanical properties, color, etc. For instance, in larger sizes it might actually be a foam with solid inner and outer surfaces.

The measurements of loss were with actual secondary coils measured over time with both swept Q measurements and ring down tests (put an impulse in and see how it decays). They put it in environments where the humidity was high (garage or barn on rainy days, I believe) and then when it was dried (warm dry air)

There was also a comparison with other materials - Cardboard tubes (def hygroscopic), acrylic tubes (clear, so no filler), etc.

It was a long running discussion about "good materials" for secondary forms - partly inspired by radical differences reported with large cardboard Sonotube type forms. Turns out some concrete forms have an aluminum foil layer in them.

probably >10 years ago on Tesla Coil Mailing List.

By the way, I've collected dielectric data for years for use in my coil inductance and Q calculator. The PVC dissipation factor it uses is 0.016 at 1 MHz and 0.0055 at 3 GHz. (This is why testing PVC in a microwave oven does not indicate its HF properties.).

Indeed - and that is kind of weird - *usually* tan d gets bigger as frequency goes up.


My coil inductance and Q calculator logarithmically interpolates these values. PVC is not too bad at HF, but much better dielectrics are readily available. Try any round plastic container from your refrigerator or kitchen cabinet. It is likely to be one of the non-hygroscopic polymers Google mentions. Read the recycling code on the bottom to determine just what it's made of.


Yes - clear plastic is likely to be more like the "handbook" values.

Pipe (or tube), not so much. I was surprised the first time I was turning a piece of white 3" PVC pipe on the lathe to "clean up the surface" and discovered that right below that surface was blotches of gray and black.

On Tue, Jul 4, 2023 at 11:54 AM, Jim Lux wrote:

PVC pipe has two potential problems as a coil form (based on experience of
folks building tesla coils, so 100-500 kHz):
1) it's hygroscopic so the dielectric loss varies with the water content

I asked Google AI whether PVC is hygroscopic. Here, in part, is what it said


Polyvinyl chloride (PVC) is a hygroscopic powder material. Hygroscopic substances can take and hold moisture from the surroundings.

PVC is non-hygroscopic, meaning it does not absorb moisture internally into the pellet. However, moisture can be collected on the surface of the pellet.
Other non-hygroscopic polymers include polypropylene, polystyrene, and polyethylene.


I thought that was an interesting distinction. Google AI generates very useful summaries, but I don't fully trust it yet, so I searched onward. I found two more references that said PVC is not hygroscopic.

I'm sure your remark about Tesla coils is based on practical experience. Any idea why it might differ from what these references say?

By the way, I've collected dielectric data for years for use in my coil inductance and Q calculator. The PVC dissipation factor it uses is 0.016 at 1 MHz and 0.0055 at 3 GHz. (This is why testing PVC in a microwave oven does not indicate its HF properties.). My coil inductance and Q calculator logarithmically interpolates these values. PVC is not too bad at HF, but much better dielectrics are readily available. Try any round plastic container from your refrigerator or kitchen cabinet. It is likely to be one of the non-hygroscopic polymers Google mentions. Read the recycling code on the bottom to determine just what it's made of.

Brian

On 7/4/23 11:16 AM, WB2UAQ wrote:

Mike and Alan,
I measured the loss factor for PVC and it was about 0.005 or so at 1 kHz. I did more measurements at RF but will have to go back to old note books. I made a simple fixture to test a number of plastics. I rebuilt a coil that plugs into a BC-610 transmitter and from that sprung all kinds of questions about dielectrics. Many local guys are using 3D printers and using PLA, I think it is called. They printed out a plate of it for me to test. I would not say that all of this material is made equally and permitivity and loss factor are probably not well controlled. Won't be able to get back to this for awhile as it is Independence Day and a lot going on to take care of this week. Can only say right now that none of this plastic seemed to jump out as being a terrible dielectric. I also tested plate glass because back in the early days of radio capacitors for spark transmitters used what I think was regular glass available at that time.

PVC pipe has two potential problems as a coil form (based on experience of folks building tesla coils, so 100-500 kHz):
1) it's hygroscopic so the dielectric loss varies with the water content - nylon has the same problem, BTW.
2) most pipe is made from recycled material, and it can have metal fragments and other debris in it. The specs for pipe are about holding pressure and being "safe for drinking water", not dielectric properties.

Tesla coil folks solve #1 by drying it and then coating with something like lacquer, glyptal, or urethane

#2 is a bit trickier - it's not like you can take a 4 foot long piece of 6" pipe and throw it in the microwave and see if it arcs or has hot spots.