eHam

eHam Forums => HomeBrew => Topic started by: KE7BZI on June 01, 2011, 03:10:30 PM



Title: ethylene glycol as a dielectric
Post by: KE7BZI on June 01, 2011, 03:10:30 PM
All

I'm thinking about building a small loop antenna.  I've been looking into option for the HV tuning cap.  I see that some people are putting a standard air variable caps in oil to increase the capacitance and voltage.  I started looking around at other fluids that could be used. The dielectric constant of ethylene glycol is 38 instead of 3 for mineral oil.  Anyone see an obvious flaw in using standard air variable submerged in ethylene glycol?  It seems like it would provide an impressive increase in performance.  I found a reference that the breakdown voltage is very high (60Mv/mm?).  The auto combustion temp is about 750F.  I also found a reference to it's use in electrolytic caps, but not as the main dielectric.

I'm aware of the butterfly / fixed stator vs. cap with wiper issue. 

Thanks
Marc


Title: RE: ethylene glycol as a dielectric
Post by: W5FYI on June 01, 2011, 09:05:42 PM
Check its hygroscopic qualities; it may absorb moisture from the air and change its dielectric constant over time. It is also attractive to animals and insects, and is poisonous. Other than that, you may have hit upon a good idea. Let us know how it works out.


Title: RE: ethylene glycol as a dielectric
Post by: KA4POL on June 01, 2011, 10:12:14 PM
I do have some doubt on the results of your plan. In one point you are right, the higher the permittivity (=dielectric constant) the higher the capacitance. However, the lower the permittivity the higher the voltage. I.e. your best solution for high voltage would be a vacuum cap. It may not be the best solution for your wallet. ;)


Title: RE: ethylene glycol as a dielectric
Post by: W8JX on June 01, 2011, 10:31:39 PM
Check its hygroscopic qualities; it may absorb moisture from the air and change its dielectric constant over time. It is also attractive to animals and insects, and is poisonous. Other than that, you may have hit upon a good idea. Let us know how it works out.

Actually you might want to try propylene glycol. It is non toxic and it has it highest boil point and lowest freeze point at 100% concentration while pure ethylene glycol freezes at about 9F or so.  Liquid cooled aircraft engines have long used pure propylene glycol.


Title: RE: ethylene glycol as a dielectric
Post by: KE7BZI on June 01, 2011, 11:20:37 PM
I found a PHD paper that lists the breakdown voltage of 95% ethylene glycol 5% water as 270 kV/cm.
Seems like it might be worth a try.

Marc 


Title: RE: ethylene glycol as a dielectric
Post by: G3RZP on June 02, 2011, 02:54:22 AM
What about its dielectric loss tangent at radio frequencies? Unless it's very low (like Teflon) you may not win anything.


Title: RE: ethylene glycol as a dielectric
Post by: KE3WD on June 02, 2011, 06:48:44 AM
What power level do you intend to run? 

Use of oil or glycols may be overkill here. 

If wanting to design for more than about a hundred watts, I'd recommend biting the bullet money-wise and investing in a vacuum-variable cap suited to the purpose. 


73


Title: RE: ethylene glycol as a dielectric
Post by: KA4POL on June 02, 2011, 07:39:29 AM
What power level do you intend to run? 

Use of oil or glycols may be overkill here. 

If wanting to design for more than about a hundred watts, I'd recommend biting the bullet money-wise and investing in a vacuum-variable cap suited to the purpose. 


73

Good and justified question concerning the power levels. We have not been told the value yet. And thanks for supporting my suggestion  ;) on the vacuum cap: http://www.surplussales.com/vaccumvarcaps/VVC2.html would be one possibility. He may be lucky on some auction as well.


Title: RE: ethylene glycol as a dielectric
Post by: GRADY on June 02, 2011, 10:03:57 AM
I think just plain ethylene glycol is very corrosive as well. They have to add other properties with it for antifreeze purposes if I am not mistaken.


Title: RE: ethylene glycol as a dielectric
Post by: N3OX on June 02, 2011, 02:14:22 PM
What about its dielectric loss tangent at radio frequencies? Unless it's very low (like Teflon) you may not win anything.

Agreed.  

I found a reference (Zahn et al. Proceedings of the IEEE, Sept 1986, Vol 74 Issue 9 Dielectric properties of water and water/ethylene glycol mixtures for use in pulsed power system design) that puts the loss tangent of ethylene glycol at 0.03 at 1MHz and 0.008 at 10MHz.

That puts it somewhere around wood or dirt at 1MHz and somewhere around G-10 fiberglass (FR4 circuit board material) at 10MHz.  I cooked a piece of fiberglass that was probably like that by using it as a spacer between the plates of a 17m magloop I was playing with, and that was with 30W.

Your capacitor Q is going to be much too low if you use even something with loss tangent of 0.008.  That's Q=125 and most of your power will go into heating the dielectric.


Title: RE: ethylene glycol as a dielectric
Post by: KE3WD on June 02, 2011, 04:19:07 PM
Design is never a one-input problem. 


73


Title: RE: ethylene glycol as a dielectric
Post by: KE7BZI on June 02, 2011, 05:44:30 PM
Well I suspected there might be a flaw and ESR may well be it.  I don't have the math skills to understand tangent loss but I understand it is one component of ESR.
I'll see if one of the guys at work will walk me through it.

My rig is in the area of 100w so the standard solutions of a vacuum variable or an air variable are options. 
The affordable vacuum caps are either old and unknown,  or tend to be sold by people from the capital of internet fraud (Ukraine), and still cost $200 with shipping.
Butterfly or fixed stator caps in the 5KV range tend to be expensive. I'm not saying they aren't the best option by it seemed interesting to look for options.

Since I have a complete machine shop, I tend to like mechanical problems so building some kind a of an air variable may the way I go.
After all, when you have a hammer, everything looks like a nail.

Marc


Title: RE: ethylene glycol as a dielectric
Post by: N3OX on June 02, 2011, 06:29:38 PM
The affordable vacuum caps are either old and unknown,  or tend to be sold by people from the capital of internet fraud (Ukraine), and still cost $200 with shipping.
Butterfly or fixed stator caps in the 5KV range tend to be expensive. I'm not saying they aren't the best option by it seemed interesting to look for options.

I don't know your definition of "affordable" and I don't know what capacitance you need but Max Gain Systems has a couple of Comet vac variables (200pF max and 60pF max) that clock in around $100:

http://www.mgs4u.com/RF-Microwave/vacuum-variable-capacitors-200.htm

Don't know what shipping adds to that, but it would just be regular UPS.  Anyway, not trying to discourage you from other approaches, but if you weren't aware of Max Gain I just wanted to point it out.  I haven't purchased a vac variable from them, but have been happy with their service with fiberglass and other doodads and they'd be one of the first places I'd look for a vac. variable just because of their stated policies about hi pot testing and returns coupled with their reasonable prices.  They do have a lot of $200 caps too, of course.


Title: RE: ethylene glycol as a dielectric
Post by: KA4POL on June 02, 2011, 10:07:41 PM
Quote
My rig is in the area of 100w so the standard solutions of a vacuum variable or an air variable are options. 

Air is between 2 and 5 kV/mm so there is absolutely no need for any effort to improve the dielectric strength. Just take a capacitor with sufficient distance between the plates.


Title: RE: ethylene glycol as a dielectric
Post by: G3RZP on June 03, 2011, 04:15:43 AM
N3OX,

You just confirmed my suspicions......


Title: RE: ethylene glycol as a dielectric
Post by: KE7BZI on June 03, 2011, 04:26:50 PM
Ok, see if this makes sense or if I'm missing something (else).
According to a little calculator I found at http://www.standpipe.com/w2bri/software.htm

If you have an butterfly air capacitor (0.5 the pF, 2x the voltage) with:
10 plates of 4 in sq. meshed surface area.
0.03" spacing

You get:
135 pF good for 4500 volts.

It seems like more capacititance gives you more tunable bandwidth and 500 pF is not too much to want.  So we'ed like a cap 3 of 4 times that value.

Now if you use something with a higher dielectric constant and a larger break down voltage things would seem to get better.
The same cap with 10 X 4" sq. is 5000 pF and has a voltage rating of 40 kV.

There may be reasons why it won't work, like resistive loss,  but it seems absolutely worthy of investigation. :P

Marc 


Title: RE: ethylene glycol as a dielectric
Post by: KE3WD on June 03, 2011, 06:15:31 PM
At "only" the 100W level you have plenty of good and inexpensive options.

*Many hams devise a Trombone Capacitor.  Screwdriver on threaded rod for remote control.  Google, there are plenty of good examples about.

*Some use flat glass as their dielectric and slide one copper-sided piece of PCB over two others beneath the glass.  Coolbeans if the antenna is a permanent installation indoors, such as in an attic.  Again, the screwdriver motor and threaded rod are implemented to obtain remote control.

*MFJ sells two butterfly caps, the ones used in their magloops. 

#1 - BUTTERFLY LOOP AIR VAR CAP, 18-136 PF, $89.95

Quote
This is our MFJ-23, a larger low loss butterfly loop tuning capacitor. The capacitance is variable from 18-136 pF.

This low loss butterfly loop tuning capacitor has a voltage rating of 4200 Volts rms. Built-in insulator feet give 1-inch of space under the bottom of the capacitor allowing convenient mounting and high voltage insulation. Solder lugs are provided on the stators and rotor for easy wiring.

Longer than the MFJ-19, this butterfly capacitor measures 10 inches in length, 3 inches wide and 3½ inches tall. Included is a 2½ inch long ¼ inch diameter insulated nylon shaft so a knob can be easily added to aid in turning the capacitor rotor.

http://www.mfjenterprises.com/Product.php?productid=MFJ-23 (http://www.mfjenterprises.com/Product.php?productid=MFJ-23)

#2 - BUTTERY LOOP AIR VAR CAP,12-67PF, $69.95

Finally, as a lifelong homebrewer, take a look at this guy's brilliant approach and use of sliding coca cola cans: 

http://www.youtube.com/watch?v=gaJlrN3cElo (http://www.youtube.com/watch?v=gaJlrN3cElo)

If you don't appreciate the elegant answer of the readily available coated aluminum cans as trombone cap and his use of two ordinary plastic medical syringes plus a length of plastic tubing to make a *hydraulic* remote capacitor controller that even has CC markings as to position, what can I say aboutcha?  In my book, this kinda thing is what ham radio is *supposed* to be about. 

Outside the box,

73





Title: RE: ethylene glycol as a dielectric
Post by: KE7BZI on June 05, 2011, 01:43:04 PM
Thanks for the link to the video.  That is a very elegant solution.  And the good news is that you could make it out of bear cans :).  With the proliferation of odd can sizes I'll bet you could find several that would nest.  I'll have to see what I can find on the self.  If you used a metering pump you could "gear" it down to have more precision.

Marc


Title: RE: ethylene glycol as a dielectric
Post by: WB6BYU on June 05, 2011, 04:43:16 PM
Quote from: KE7BZI
...And the good news is that you could make it out of bear cans...


A can would have to be pretty big to hold a bear, even a young cub.  And I don't think I'd want to
deal with getting live bears out of their cans, unless I'd had too much beer.


Title: RE: ethylene glycol as a dielectric
Post by: KE7BZI on June 06, 2011, 07:49:26 PM
Oh sure, make fun of the hieroglyphical challenged guy.  ;)


Title: RE: ethylene glycol as a dielectric
Post by: W6RMK on June 06, 2011, 08:26:01 PM
liquid dielectrics always seem interesting at first, but ones with high dielectric constants (e.g. glycols) tend to be hygroscopic, and with water comes loss and radically lowered breakdown voltage.

If you must try oil, then castor oil is probably your best bet. It's around 4-5 dielectric constant, has the usual several hundred kV/inch breakdown, etc.

Low loss tangent too (they use it in pulse power capacitors, among other things).

The problem, in general, with oil is that it burns.

There are also some PCBs that make good dielectrics.  The challenge is in finding ones without dioxin contamination.  At least it's not flammable.

All liquid dielectrics are a pain to work with.. they're a mess, they leak, etc. etc.  Unless you're building a hermetically sealed enclosure.  If it's a variable capacitor, oil will leak out around the shaft bearing, or you have to mount the shaft vertically and watch the levels.


Watch out on the pulse power literature.. for 3 ns pulses distilled water makes an excellent dielectric.  For 100% duty cycle, not so wonderful.


Title: RE: ethylene glycol as a dielectric
Post by: KE7BZI on June 09, 2011, 12:28:04 AM
Thanks,  so much to learn.
For now I decided to build a simple loop with trombone cap to gain more experience.

One thing playing with the loop calculators points out is that when you attempt to extend the tunable bandwidth of a loop with a larger cap, loop efficiency starts to drop off quite fast.  I need to build some examples to get a better understanding of what the means for real performance.

As KE3WD pointed out earlier in this thread,  Design is never a one-input problem.

Marc


Title: RE: ethylene glycol as a dielectric
Post by: W8JI on June 09, 2011, 08:03:08 AM
The ideal capacitor for a loop antenna, or any other high Q capacitor requirement,  is small and boxy in shape.

As the shape gets longer the length of conductors adds series inductance, and that increases losses. The worse possible shape is a long cylinder.

Adding dielectrics is always a problem, unless the dielectric is carefully selected. Things like Teflon, mica, and some other plastics and ceramics can maintain fairly high Q.

I've yet to see a high Q liquid capacitor at radio frequencies, nor a physically long capacitor with high Q. The two best capacitors are vacuum caps which can have a Q >50,000 and butterfly air dielectric caps which can have Q>5000.  Adding a dielectric generally takes them down to Q<2500.

Most trombone caps have Q<1000 and I've seen them as low as 50 or 100. The few liquid dielectric caps I've tested had Q~100 or less.