Anyone know if JB Weld conducts electricity? (2024)

Hi all-

So as I understand it JB Weld has tiny metal particles in it? Does anyone know if it actually conducts, either before or after curing? Thanks...

Sometimes the answers you need are already there:

https://www.jbweld.com/faqs(the heat question is the first one on the FAQ, it has the answer to being a conductor as well)

"WILL J-B WELD CONDUCT ELECTRICITY?
No. J-B Weld is not considered to be a conductor. It is an insulator."

Straight from the source.

Tony

Interesting note, though not related.

When I was wiring my model railroad layout I cut gaps in the rails for the block control of a reversing Y.
To hold the rails in alignment and to keep the gaps from closing, one places a tiny drop of gap-filling "Superglue" in the gap.

Years pass with no trouble when suddenly I get a short between blocks but only in one direction.
Out comes the multi-meter and after an hour or so of increasing frustration I discover than one of the ACC filled gaps in acting exactly like a diode.
It will pass current in one direction only, thus the short, and the gap even had a diodes .7 voltage drop.

Nobody has ever ben able to explain this phenomenon.

FAQs - JB Weldhttps://www.jbweld.com › faqs
Will J-B Weld conduct electricity? No. J-B Weld is not considered to be a conductor. It is an insulator.

EDIT: I'm slow on the draw, as usual.

I would be interested in actual resistance measurements on both uncured and cured JBWeld. It may be considered to be an insulator but with its metal content I'd hypothesize that it's not nearly as good an insulator as are unfilled epoxies. For most applications that's probably not an issue, but if it's being used on electronics...I'd be wary.

Best -- Terry
Who said it? "In god we trust, all others bring data"

boomtube-mk2 said:

Interesting note, though not related.

When I was wiring my model railroad layout I cut gaps in the rails for the block control of a reversing Y.
To hold the rails in alignment and to keep the gaps from closing, one places a tiny drop of gap-filling "Superglue" in the gap.

Years pass with no trouble when suddenly I get a short between blocks but only in one direction.
Out comes the multi-meter and after an hour or so of increasing frustration I discover than one of the ACC filled gaps in acting exactly like a diode.
It will pass current in one direction only, thus the short, and the gap even had a diodes .7 voltage drop.

Nobody has ever ben able to explain this phenomenon.

Click to expand...

Only because you didn't ask the right people. If you used a superglue gel (there are many formulations on the market), it probably contained pyrogenic silica. Pyrogenic silica forms long chain molecules containing high concentrations of silicon, a well-known semiconductor. You can take it from here.

RICHARD COLARCO said:

Tested cured JB Weld with my digital multimeter. Resistance: infinite. It's an insulator.

Click to expand...

Thanks, that's just what I wanted to know. I'm still a bit surprised, though. I expected resistance on the order of tens of megohms per cm.

Best -- Terry

RICHARD COLARCO said:

If you used a superglue gel (there are many formulations on the market), it probably contained pyrogenic silica. Pyrogenic silica forms long chain molecules containing high concentrations of silicon, a well-known semiconductor.

Click to expand...

That's a pretty good explanation, but why did it take years before the "Diode Effect" became an issue?

prfesser said:

Thanks, that's just what I wanted to know. I'm still a bit surprised, though. I expected resistance on the order of tens of megohms per cm.

Best -- Terry

Click to expand...

With metal particles JB Weld would probably have a breakdown voltage at some potential. Even common insulators do. Also, most meters just do not accurately measure very high resistance. Megohmeters are necessary to detect that high of resistance. Practically speaking it may be considered as infinite resistance for most purposes.

boomtube-mk2 said:

That's a pretty good explanation, but why did it take years before the "Diode Effect" became an issue?

Click to expand...

A diode is not one semiconductor, but two: one p-type semiconductor, next to one n-type semiconductor.

If the semiconductor was under voltage for years, maybe some kind of impurity within migrated to one end, resulting in a "one p-type semiconductor, next to one n-type semiconductor" configuration.

prfesser said:

Thanks, that's just what I wanted to know. I'm still a bit surprised, though. I expected resistance on the order of tens of megohms per cm.

Best -- Terry

Click to expand...

I suspect results might vary a bit with test equipment, the distance between the probes, and how deep / hard you bury them in the sample. Most multimeters are using a very low test voltage for measuring resistance, usually less than 1 volt on low ranges, although some meters are using 5 volts for higher range tests. If you were doing a high-potential test, it would probably have some small current flow.

Found the following on a TDS sheet for JB Weld - This numbers are quite high. Surface resistance is 3 x the bulk product. Doubt this is something you can measure outside of a lab bench. This is effectively a non-conductor unless you are playing with a Tesla coil. Still would not use it to pot a circuit board.

http://www.aeroelectric.com/Mfgr_Data/JB_Weld/JB-Weld-Technical-Datasheet.pdf

The TDS published on the JB Weld site says the epoxy is 11% iron (steel) by weight. And 50% limestone. That is just for the resin component. The hardener doesn't have any metal filler. So mixed product is about 5% iron. Not a lot of metal. Suspect it is mostly for color.

Funkworks said:

If the semiconductor was under voltage for years, maybe some kind of impurity within migrated to one end, causing the "one p-doped semiconductor, next to one n-doped semiconductor" configuration.

Click to expand...

That was always my belief but it wasn't as if the rails were always "Energized".
Maybe a few hours on the days I operated but then it might set for a few days before I powered up again so the total time the rails were actually carrying a current were minor compared to the hours they weren't.

And then there was the fact that this effect came about literally between one power-up and the next.
One operating session everything was copasetic, the next it wasn't.

Still, it had to have been something along the lines of what you described else I have to start believing in Magic.

"Everybody wants a magical solution to their problems and yet nobody actually believes in magic"

-Jefferson aka The Mad Hatter-
Once Upon a Time

boomtube-mk2 said:

One operating session everything was copasetic, the next it wasn't.

Click to expand...

Long term temperature and humidity variations have adverse effects on both electronics and organics (thus silica packs and sealed bottles). For the rails, you have one or two metals (if plated), and then superglue which can have one of many formulations (and that's a specialty all by itself). Who knows what goes on exactly at the interfaces where these materials all meet and react with oxygen and humidity. It can be interesting to investigate, but it's a lot of work. People can make careers out of this.

If JB weld if filled with iron, which is a ferrous metal, encapsulation conductors or electronics in it would definitely have an effect on inductive reactance. So, any RF or signals with AC components could certainly be affected.

You also have to remember a simple multimeter is using a DC current to measure resistance. So, the conductivity of metal filled epoxies like this may be quite different if we're talking about AC, RF, or pulsed signals. If the metal particles are encapsulated by the polymer, that would prevent them from conducting DC voltages below the breakdown voltage of that insulating barrier, but they'd still act as a chain of small capacitors. So, that could certainly make it conductive to non-pure DC signals. To what extent and what result would depend on many factors. So, if we're just talking about encasing the leads to an igniter, which is only going to see a relative low steady DC voltage that's probably fine, but things like flight computers and transmitters and receivers are probably not a good idea.

JB Weld is sort of liquid ferrite (and not very good ferrite at that). You could probably make a very poor inductor out of a hardened blob of JB Weld.

Even good ferrite (with a high concentration of Fe2O3 (better known as rust)) does not conduct DC. (I just verifies this on a home brew choke on one of my portable antennas.) Feed it AC at line frequency or RF through windings, however, and you have an inductor (choke or transformer).

boomtube-mk2 said:

When I was wiring my model railroad layout I cut gaps in the rails for the block control of a reversing Y. To hold the rails in alignment and to keep the gaps from closing, one places a tiny drop of gap-filling "Superglue" in the gap.

Click to expand...

So, what's wrong with these? I'm pretty sure you can count on them not to spontaneously turn into diodes.

RICHARD COLARCO said:

Only because you didn't ask the right people. If you used a superglue gel (there are many formulations on the market), it probably contained pyrogenic silica. Pyrogenic silica forms long chain molecules containing high concentrations of silicon, a well-known semiconductor. You can take it from here.

Click to expand...

What no one has mentioned is that silica isn't silicon any more than salt is sodium. Silicon, in point of fact, doesn't form long chains, as silica can.

A better hypothesis, I think, is something weird creating a structure akin to a crystal detector.

A crystal detector is an[1] electronic component used in some early 20th century radio receivers that consists of a piece of crystalline mineral which rectifies the alternating current radio signal and was employed as a detector (demodulator) to extract the audio modulation signal from the modulated carrier, to produce the sound in the earphones. It was the first type of semiconductor diode, and one of the first semiconductor electronic devices. The most common type was the so-called cat whisker detector, which consisted of a piece of crystalline mineral, usually galena (lead sulfide), with a fine wire touching its surface.

Click to expand...

That is, the junction of a metal crystalline, bearing compound with a metal conductor, rather than two differently doped regions of a pure semiconductor like silicon.

cvanc said:

So as I understand it JB Weld has tiny metal particles in it? Does anyone know if it actually conducts, either before or after curing?

Click to expand...

you never said why you're asking? If it's because you want a conductive glue, you should know that there are several available, both epoxy and other types. They have much heavier metal loading than JBW, and often the metal is silver. Just Google it.

Can't believe I missed the FAQ but thanks for the meaty talk. It's almost awkward to reveal how trivial the item at hand is, but thank you for helping me fix it (big thumbs up goes here)!