Wax Paper Capacitor Test

[R.G.]

The wax+paper and wax+oil dielectrics were later used because they provided more capacitance per volume, obviously.

I'm not sure that's obvious. It may well have been for the higher voltage possible with impregnation, or the ease in cooking out residual water instead of higher dielectric constant. It is also possible that paper plus wax or oil provided a way to use thinner paper for a smaller cap of a given voltage.

What is curious is that either paper or wax(paraffin) or oil (assorted phthalates, monoglyeride esters, and PCBs) all have a dielectric constant ranging 2-3; in combination with paper, the net dielectric constant was approximately additive, i.e. impregnated paper dielectric constants seem to range 4-6.

I'd very much like to read a reference on this. Can you post a link? It seems unlikely that paper with a dielectric constant of 1.7-4 would be additive with oils and waxes at 2-3; see

Dielectrics - The Physics Hypertextbook

. If mixing dielectrics were additive, the way to get a very high dielectric would be to go on adding more things with low dielectrics.

It seems more reasonable to me that the dielectric constant of impregnated paper would be a volumetric average in some sense, instead of additive. Dielectric constant reflects polarization of the insulator by the E-field, and this would seem to depend on the amounts of the relative molecules between the plates.

Purchasing and testing a box of ~100 pulls revealed the variability in wax paper cap quality.
Tested at 1kHz, about 1/2 of them have a DF <1%, 1/3 range DF ~1-2%, and the final 1/6 range DF ~ 3-25%.
The low dissipation factor ones are usually within 5% of their marked values while some of the others are as much as 50% high.

I believe this.

Wax-dipped capacitors dominated the defect pile while caps with plastic/epoxy/polymer cases age well.

I believe this too. Wax is much more permeable by water than plastic encapsulation is.

So it's not that all wax paper caps are good or bad, but that there are many good and bad instances that you won't find without testing.

I wonder if the insides of wax dipped caps is wax impregnated paper, necessarily, or if there is some other impregnant.

[salvarsan]

I'm not sure that's obvious. It may well have been for the higher voltage possible with impregnation, or the ease in cooking out residual water instead of higher dielectric constant. It is also possible that paper plus wax or oil provided a way to use thinner paper for a smaller cap of a given voltage.

Search term "impregnated paper capacitor permittivity"
Search term "dielectric constant tables"
Use scholar.google.com for more authoritative references.

There are many many tables and the same values appear a lot -- various papers: 2-3.7, paraffin wax: 2-3, oils are usually 2-3 whether paraffins, naphthenes, monoglycerides, or PCBs. Some patents extol the virtues of nitrobenzene over oil, but we should save that discussion for the energetic materials forums.

Paper's reported permittivity is problematic because the details aren't usually available. Was it bone dry refined cotton rag capacitor paper or air dried brown kraft paper? Most likely, the refined dry capacitor paper has the lower permittivity.

2.6.5 Dielectric properties of materials lists several kinds of paper values, and is the best ref I've found, listing density, test frequency, epsilon, and tan delta.

Dielectric Properties of Metalized Paper-Film Capacitors unambiguously states that paper+oil has a dielectric constant of 5, absent any details on the oil.

[R.G.]

Thank you. I'll go read up a bit on it.

[salvarsan]

Be selective. It's easy to get lost in the research papers and patents, and some of it is truly eye-glazing-over stuff.
Impregnated paper dielectrics are still researched today.

DEIS, the IEEE Dielectrics and Electrical Insulators Society, is probably the mother lode but needs some exploration since it's so large.

FaradNet is slightly dated and focused on electrolytics.

[Rhodesplyr]

In my testing of caps from a Hammond Tone generator, there were consistent patterns in terms of brands. Some resisted aging effects better than others. Throughout the 1950s, there was a lot of research and development going on trying to produce more reliable capacitors. I read that in some early TVs in humid environments, techs had to add always-on lightbulbs inside in an attempt to control water absorption related failures.

Both CDE and the John E. Fast Company made some special-order, high grade paper capacitors that Hammond used in their preamps and amps. These were well-sealed in ceramic or hard plastic tubes. These rarely fail, even today, and still measure like new. CDE's specialty cap was the "Budroc" line. And, of course, Sprague Vitamin Q capacitors seem to last forever.

[R.G.]

Be selective. It's easy to get lost in the research papers and patents, and some of it is truly eye-glazing-over stuff.

I'll try not to get too lost.
My idea of a good book to read for relaxation often includes textbooks, digests of technical journals, and other eye-glazing stuff. If my eyes were going to get very glazed, they'd already be there.

In my testing of caps from a Hammond Tone generator, there were consistent patterns in terms of brands. Some resisted aging effects better than others.

That is exactly what I would expect. Some of my friends in my previous career were manufacturing line support engineers, and I got a lot of inculcation into what to expect from different manufacturers, batches and engineering changes in parts. It's sometimes like buying matching draperies or upholstery - if you want a match, buy from the same dye lot and run.

[salvarsan]

Development of Specialty Papers is an Art: Electrical Insulation Papers From Indigenous Materials is a straightforward 2003 article on making capacitor paper from hemp. It also thoroughly characterizes the product. The dielectric constant (relative permittivity) is given as 1.11, an unusually low figure for any paper, usually spec'd at 2-3.7.


60 years ago, scientists were still scratching their heads about how papers made on the same machine could have such different properties.
Microstructure of Capacitor Paper,
D. C. McLean , H. A. Birdsall , C. J. Calbick
Ind. Eng. Chem., 1953, 45 (7), pp 1509–1515
DOI: 10.1021/ie50523a040

[Leo_Gnardo]

Been enjoying this thread Salvarsan. Never knew there was so much to capacitor paper. Nitrobenzene !?! Wow, that stuff is bad for ya. Right away, not like PCB's. I read 2 ppm in breathing air is enough to poison someone, back when I was reading up on deadly chemicals in the college library. Hey it was 1973, maybe NB been given a pardon by now. And hemp paper, with a low DC of 1.11? How relaxed n mellow. Wonder that the capacitor companies who sell to well-funded "golden ears" types don't put this one to work and charge a fortune. What could be better to use in gear that will be playing back old stoner rock recordings? Courses for horses.

@ 1970 some friends of mine attended Newark College of Engineering. They didn't worry much about the fine-tuned properties of capacitor paper. Just made their own caps out of aluminum foil and wax paper from the grocery store. Not little ones. Two rolls of foil and a roll of wax paper. What for? They used them to fire their home made rail gun. Sent magnetized nails flying across the city. Had a spotter scope on the Bell Telephone building and were able to hit it from their lab. Knew enough to not stand behind the thing, because firing direction was unpredictable. Ah, those crazy kids and their wacky japes.

Warning everybody, don't try this at home. Or anywhere else.

[salvarsan]

Been enjoying this thread Salvarsan. Never knew there was so much to capacitor paper..

It's a big topic. As I read it, they never stopped using it in the power grid, but those caps and wire sheaths are made to different quality and durability standards than anything you'd normally put on a circuit board.

Nitrobenzene !?! Wow, that stuff is bad for ya. Right away, not like PCB's. I read 2 ppm in breathing air is enough to poison someone, back when I was reading up on deadly chemicals in the college library..

It's not quite that toxic, but it's certainly bad news.

And hemp paper, with a low DC of 1.11? How relaxed n mellow. Wonder that the capacitor companies who sell to well-funded "golden ears" types don't put this one to work and charge a fortune. What could be better to use in gear that will be playing back old stoner rock recordings?.

When ya 'member that any cannabis can be bred for either fiber, seeds, or resin, it's unsurprising that they used the fibrous cannabis sativa cultivar to make a refined paper slurry.

Capacitor paper impregnation oils have been broadly researched, too. Apparently, a little dissipation factor is acceptable so that dielectric absorption is controlled; paper-oil caps are less microphonic+piezoelectric than most film caps, too. The move away from PCBs and mineral oils in oil caps and transformer oils is possible with monoglycerides and mildly exotic glycols, both biodegradable.

In short, any place you'd put an Orange Drop or Mallory 150 polyester cap, you could as easily put a modern impregnated paper cap.

[salvarsan]

Too much information, part 1:

Capsite 2009 is a terrific overview of all capacitor technology, more to-the-point than FaradNet for the broad sweep.
It has gems like links to practical modeling of dielectric absorption.

[R.G.]

There are many many tables and the same values appear a lot -- various papers: 2-3.7, paraffin wax: 2-3, oils are usually 2-3 whether paraffins, naphthenes, monoglycerides, or PCBs.

Yes, those are consistent with what I would expect.

Some patents extol the virtues of nitrobenzene over oil, but we should save that discussion for the energetic materials forums.

If you like energetic materials, look up dioxygen difluoride (known affectionately as "FOOF") and chlorine trifluoride, which was described by John Clark in "Ignition":

It is, of course, extremely toxic, but that's the least of the problem. It is hypergolic with every known fuel, and so rapidly hypergolic that no ignition delay has ever been measured. It is also hypergolic with such things as cloth, wood, and test engineers, not to mention asbestos, sand, and water-with which it reacts explosively.

Materials which are hypergolic with asbestos and sand are energetic.

Paper's reported permittivity is problematic because the details aren't usually available. Was it bone dry refined cotton rag capacitor paper or air dried brown kraft paper? Most likely, the refined dry capacitor paper has the lower permittivity.

That suits my expectations.

2.6.5 Dielectric properties of materials lists several kinds of paper values, and is the best ref I've found, listing density, test frequency, epsilon, and tan delta.

That suits my expectations as well.

Dielectric Properties of Metalized Paper-Film Capacitors unambiguously states that paper+oil has a dielectric constant of 5, absent any details on the oil.

It actually says this:

Fluids normally have much greater conductivities than solids as a result of much greater ionic mobility. Thus at very low frequencies, the fluid-impregnated paper looks like a conductor while the polymer still looks like a dielectric. This results in all the field falling across the polymeric dielectric which results in the effec-tive dielectric constant increasing from roughly 3 to 7 through the relaxation peak.

and this:

If we assume that the relative dielectric constant of the impregnated paper is 5,

The research paper is looking at conditions where capacitor paper is metallized on both sides, and insulated between paper layers by a plastic film. The specific condition is that one side of the paper or a region of the side fails to connect to the other metalization. The conditions are such that the isolated area causes a secondary leaky capacitance to be inserted into the lumped model of the capacitance in question. The comment about dielectric constant refers to the condition where the impregnating fluid (oil, usually) in the paper fails to act like a dielectric in comparison to the polymer film between layers, and the composite dielectric then "fails" back to the dielectric constant of the film.

The unambiguous 5 is actually an assumption of some dielectric constant between the roughly 3 to 7 from the first paragraph, and no particular original source is given for that, either.

The paper doesn't seem to say that the dielectric constants of the paper and impregnant add, that I can find. Did just miss that?

From my reading so far, it still looks like the properties of the paper and paper filled with impregnant still look like a volumetric effect, not an additive effect of the dielectric constants.

[salvarsan]

If you like energetic materials,

Nope, not a one. I think it was reckless that both nitrobenzene and picrates, at a time when their energetic hazards were well known, found utility in capacitor technology. The Kerr cell, an electro-optic shutter, was a calculated risk but not a commercial concern. Fortunately.

The paper doesn't seem to say that the dielectric constants of the paper and impregnant add, that I can find. Did just miss that?

I observed that the individual dielectric constants of paper and its impregnant are approximately additive. Paper does not appear to diminish the dielectric constant of its impregnant, not do the impregnated paper DC values look like averages of the components, nor is the final DC less than that of any one component. Given these enumerated possibilities (less than one, less than any, average, sum) I think they look more like rough sums. It doesn't seem like a novel or contentious observation.

One of the articles I cited states that paper is not a passive matrix for dielectric impregnant and its dielectric contribution can't be reasonably ignored.

Now that is an interesting observation because ...

As I understand it, refined capacitor paper may be considered mostly cellulose. Cellulose's structure is roughly described as a glucose linear polymer. Each glucose monomer in the strand has three hydroxyl groups available so it is both energetically (in a thermodynamic sense) and entropically favorable that the polymer strands align and stack from hydrogen bonding as well as space-filling effects at the molecular level. The importance of hydrogen bonding can't be understated since, though weak, it is ubiquitous. From the molecular level up to the microscopic level, cellulosic paper is weakly self-polarized and polarizable. I mean, c'mon, it clumps together, dudnit?

This also means that paper, as a dielectric, has a small intrinsic loss factor that tends to disperse localized charge when it relaxes. My suspicion is that the cellulose clump granularity strongly affects the relaxation period and may define its useful bandwidth, but that's only a suspicion.

Someone else needs to talk about how paper, a hydrophilic substance, absorbs hydrophobic substances into its voids.

[Leo_Gnardo]

Nope, not a one. I think it was reckless that both nitrobenzene and picrates, at a time when their energetic hazards were well known, found utility in capacitor technology. The Kerr cell, an electro-optic shutter, was a calculated risk but not a commercial concern.

I'll be danged, that's the first mention I've seen of Kerr in double donkey's years. Was assigned the task as a college junior, of setting up an experiment that would be run by underclassmen, demonstrating the Kerr effect. When I looked into what it would entail, that's when I spent library time looking up nitrobenzene toxicity (and finding many more hair raising accounts of toxins.) Scientific supply houses had NB at @ 66% purity so it would have to be distilled to beyond 99% in a fume hood I suppose. Not wishing to be exposed to this stuff, nor ensuing physics students, or blow the fumes out the roof and give EVERYBODY a dose, I found a non toxic solution. Bentonite suspended in water. Worked great, as non-toxic as you can get, and with a field of under 100V/cm would handily demonstrate the effect. Instead of 30,000V. Took some doing to convince the head of the department, and for all this safety concern earned myself a B. I had to make the argument that the purpose of the experiment was to demonstrate the Kerr effect, not subject newbie students to hazardous chemicals and voltages. That didn't impress him much.That department head is still around, a patron of the arts and you can look him up (Dr. Heinrich Medicus.)

For those wondering, the Kerr effect is optical birefringence (beam splitting) induced by an electric field. Haven't heard of its being used in optical logic chips - would be a natural fit - but who knows maybe it is.

Pardon my digression Salvarsan, you hit an old nerve. Good to wake it up every now and then. Could you point me to some mention of the Kerr cell? What was it used for in practice? Thanks in advance, LG.

[salvarsan]

The Kerr Cell was used as an ultra high speed shutter in photography and in "Q switching" of lasers to generate powerful pulses from continuous wave devices. Search on Q switching.

[R.G.]

I observed that the individual dielectric constants of paper and its impregnant are approximately additive.

Doing some more reading, yes, it appears that you are the one saying that.

Paper does not appear to diminish the dielectric constant of its impregnant, not do the impregnated paper DC values look like averages of the components, nor is the final DC less than that of any one component. Given these enumerated possibilities (less than one, less than any, average, sum) I think they look more like rough sums. It doesn't seem like a novel or contentious observation.

Nope, not particularly contentious. I was just struck that if they were additive, the result would always be more than any one constituent. I've been reading a bit, and near as I can tell, that's not the case either. In the absence of any other info, I'll continue to think it's a volumetric combination, possibly enhanced or detracted by the psychosometric icosahedrality of the monopenes in the matrix, but only on alternate hydrogen bonds. You can continue to believe they're additive, and the answers both come out much the same. That's fine.

One of the articles I cited states that paper is not a passive matrix for dielectric impregnant and its dielectric contribution can't be reasonably ignored.

That is correct. I do firmly believe that whatever you stick in the gap will affect the capacitance, because practically any solid or liquid will have a DC bigger than one. There are probably exceptions, but then I'm not a chemist nor a physicist, so I'd have to go graft on those bits of information if I needed them.

As I understand it, refined capacitor paper may be considered mostly cellulose. Cellulose's structure is roughly described as a glucose linear polymer. Each glucose monomer in the strand has three hydroxyl groups available so it is both energetically (in a thermodynamic sense) and entropically favorable that the polymer strands align and stack from hydrogen bonding as well as space-filling effects at the molecular level. The importance of hydrogen bonding can't be understated since, though weak, it is ubiquitous. From the molecular level up to the microscopic level, cellulosic paper is weakly self-polarized and polarizable. I mean, c'mon, it clumps together, dudnit?

Yeah. I'd be a lot happier about thinking about paper dielectrics, impregnants and voids if there was any evidence that there was any consistency to be had when it is made. A lot of my samples of the literature is focused on it varying all over the map. I'm guessing that subtleties of hydroxyl groups making it favorable for polymer strands to stack from hydrogen bonding may be washed out entirely by the equivalent of the old story about the night watchman spitting in the vat.

This also means that paper, as a dielectric, has a small intrinsic loss factor that tends to disperse localized charge when it relaxes. My suspicion is that the cellulose clump granularity strongly affects the relaxation period and may define its useful bandwidth, but that's only a suspicion.

Sounds like a great place for you to go do some novel research!