Polyfuse in Encapsulation help

[ajukarandikar11]

Hello,

I am working on a project that should qualify EN60079-18 (Ex-m encapsulation standard).
In order to limit the device current(compulsory by the standard) I plan to use surface mount poly-fuse, like 2016L,  in series with the supply and communication lines.

The device needs to be completely enclosed by epoxy potting compound.
I came across this "warning" note mentioned in the datasheet of the fuse: (attached image)

Can anyone suggest poly-fuse that could work under potting compound?
Or is there any other current limitation method that I could implement on the device?

[Siwastaja]

Series resistors on communication.

On power, just design your circuitry so that it doesn't go in short circuit or use overcurrent in normal operation. Use traditional fuse to protect against fire or excessive thermal dissipation; in this case, your electronics have failed anyway, so a self-resetting fuse is not an advantage.

[Zero999]

This sounds like a bad idea to me. Even if the polyfuse can survive the potting compound, it will alter the transfer of heat the the environment, thus altering its tripping characteristics. It will trip at a lower current, or higher current than specified, depending on whether it's better or less thermally insulated from the environment, as a result of potting. I believe most potting compounds are better thermal conductors, than air, but they will block radiative and convective cooling, so it's difficult to predict.

[Giaime]

I had (bad) experience with fusible PTH resistors in epoxy potting compounds. They tend to NOT fuse and start a fire, by carbonizing the epoxy immediately in contact with the resistor body.
If you have to use a fuse, use a fully incapsulated one, such as:
https://www.digikey.it/product-detail/it/bel-fuse-inc/RST-500-AMMO/507-1722-3-ND/814043

[Zero999]

I had (bad) experience with fusible PTH resistors in epoxy potting compounds. They tend to NOT fuse and start a fire, by carbonizing the epoxy immediately in contact with the resistor body.
If you have to use a fuse, use a fully incapsulated one, such as:
https://www.digikey.it/product-detail/it/bel-fuse-inc/RST-500-AMMO/507-1722-3-ND/814043

That's interesting. How does it burn? I'd expect there's very little free oxygen inside the potting. Or do the gasses build up, causing an explosion and fire?

[floobydust]

A polyfuse is not considered a "fuse" under the 60079 standard for Hazardous Locations.
Because there is no physical disconnection or break in the current carrying element, as I was told.

Careful on your design, if you are relying on a polyfuse to protect against fire and ignition, it's won't work.
You can get a crappy certification agency such as Baseefa to certify your hazardous locactions product but later find another agency fail the certification and end up doing a product recall.

You can use a polyfuse to protect your circuit, though.

A polyfuse trip/hold current is imprecise, and one that has tripped once or twice is "softer" for life, it trips at lower points.
You can encapsulate many makes of polyfuse, but the trip/hold current goes up due to the cooling effect the potting compound has on the element.
Hard epoxy potting compounds I wouild not use.

I have done designs for Zone 0, 1, 2. I'm don't know details of your device, location, gas group etc.

[Seekonk]

I can get a polyfuse to reconnect just by touching it.  Think how potting contact will change that.  Perhaps if you can place it in closed cell foam it might work.  Current rating will be a lot higher regardless.

[ajukarandikar11]

Hi,
Good info. Thank you.

[nctnico]

Use traditional fuse to protect against fire or excessive thermal dissipation; in this case, your electronics have failed anyway, so a self-resetting fuse is not an advantage.

A polyfuse is nice to protect ports which supply external circuitry. If there is a short in there a polyfuse ensures the rest of the circuit doesn't go down. Still you always need to use through hole polyfuses. The SMD ones will just burn.

[CopperCone]

Polyfuses amuse me because their electrical characteristics and tollerances are so different then all other components. The fuse seems like a metrological standard compared to them.

[nctnico]

Polyfuses amuse me because their electrical characteristics and tollerances are so different then all other components. The fuse seems like a metrological standard compared to them.

The basic rule is to use a polyfuse which is rated for twice the actual current.

[SiliconWizard]

I wouldn't use rigid potting with components that can significantly expand during their lifetime.

With polyfuses, that may either modify their tripping characteristics or end up making cracks in the epoxy encapsulation.

I don't know much about EN60079-18. Wouldn't it be possible to use more flexible compounds such as polyurethane?

[Zero999]

I can get a polyfuse to reconnect just by touching it.  Think how potting contact will change that.  Perhaps if you can place it in closed cell foam it might work.  Current rating will be a lot higher regardless.

Wouldn't closed cell foam reduce the tripping current? It will insulate better than air, by preventing convective losses, hence why it's used for thermal insulation.

Yes, other non-foamed potting compounds will increase the tripping current, as they will conduct heat, better than air.

[jbb]

I once considered what would happen if you sealed a small can over the polyfuse to keep overmould compound out.  In principle, it might work.  However, experience suggests that it's very hard to consistently get a seal, and some of them - which you can't inspect - will flood with compound.

So PTCs are probably out.

Depending on the situation, some kind of current limiter might be in order.  Two come to mind:

• Bourns make some interesting non-linear series elements, the TBU and TCS series.  These look like resistors until you reach the trip current.  Once the trip current is reached, they turn into current limiters and can stand off large voltages (up to 800V!).  I've never used them but they look interesting.
• An Electronic Fuse (EFUSE), e.g. the TPS26624 (note: data sheet describes several devices...) can provide overvoltage, overcurrent protection (some also provide reverse current protection)

I don't know what this will do to your safety certifications, though.

[ajukarandikar11]

Thank you jbb.
I will look onto the alternate options.

[Giaime]

I had (bad) experience with fusible PTH resistors in epoxy potting compounds. They tend to NOT fuse and start a fire, by carbonizing the epoxy immediately in contact with the resistor body.
If you have to use a fuse, use a fully incapsulated one, such as:
https://www.digikey.it/product-detail/it/bel-fuse-inc/RST-500-AMMO/507-1722-3-ND/814043

That's interesting. How does it burn? I'd expect there's very little free oxygen inside the potting. Or do the gasses build up, causing an explosion and fire?

The polyurethane we used was hardly a perfect gas seal. So there's oxygen, and due to leakages there's no explosion, but on some samples there was evidence of carbonization. On some samples where the failure of the circuit was "soft" (so not very high current) the fusible resistor did get very hot, it didn't open and started a fire.
After this experience and some other products I realized that the only useful potting compound is asphalt, because it gets soft at hot temperatures and avoids mechanical stresses (broken solder joints, etc...). But it can get so soft that it leaks out, so the product casing must consider this as well.
Even with this kind of potting material, incapsulated fuses are better in my opinion and on the economy of scale sometime the cost is not prohibitive.

[qwaarjet]

Series resistors on communication.

On power, just design your circuitry so that it doesn't go in short circuit or use overcurrent in normal operation. Use traditional fuse to protect against fire or excessive thermal dissipation; in this case, your electronics have failed anyway, so a self-resetting fuse is not an advantage.

I tend to agree with this statement.  Did exactly this in design I had. I'm assuming you referning the fuse-voltage limit(zener)-current limiting resistor set described for power. However  by looking at the voltage current tables in 60079-18 (dont remember the appendix off hand) you can probably just put a high value resistor in series on you data line. If the impedance is a issue you can buffer it on the non classed side, or if you need it on the the classed side  power the buffer with   protected power lines using the aforementioned barrier design.

[ignator]

You can not encapsulate a polyfuse in any material. The way they work is to expand and open the signal path of the carbon conductive material. Encapsulating them prevents this. The failure mode most likely is to get hot, and cause the encapsulation material to fail, probably via exceeding the flame temperature.