# Making a device explosion proof

I am working on prototyping a gas leak detector device to be used in homes, it should have two small windows in the casing to allow sound to go out and the other window allows gas to reach the sensor.

Is conformal silicon resin enough to prevent the PCB from initiating a spark causing fire?

What kind of classification does it fall under, is it ex d class 1?

• I think the term "spark free" is closer then "Explosion proof". The latter suggests you device must be able to withstand an explosion. But the current header might attract more attention :-) – Oldfart Feb 2 '19 at 15:30
• According to the title, you could design a device that would cause explosions and fullfil the requirements, as long as it survives them. – Wesley Lee Feb 2 '19 at 15:36
• instead say device that is certified for operation in explosive atmosphere – jsotola Feb 3 '19 at 1:06
• @Oldfart I agree the terminology of "explosion proof" is misleading and "spark free" would be much clearer, but unfortunately "explosion proof" is the standard phrase used in the field. – pericynthion Feb 3 '19 at 21:13
• Just no. It doesn't matter at all what anyone says here - you need an engineer to review this design; one with expertise in the field. You need to understand which laws apply to your device in the countries you will be marketing it and you need to ensure that the design complies with the correct device class requirements. You can't just crowdsource an opinion on something safety-critical like this. – J... Feb 4 '19 at 12:38

First you should read up on the different ways of mitigating the risk. There's a few different, with different implications:

• Ex d
• Ex e
• Ex i
• Ex m
• Ex n

Those are the main. Often a product will use many of the protection methods.

Ex d is an approved enclosure, manufactured to withstand an internal explosion. That means that you can use normal equipment inside; and explosion inside will not spread to the outside. This means heavy metal, flame gaps and so forth to ensure that an explosion does not propagate. It is worth noting that Ex d doesn't mean whats inside the cabinet will function afterwards. Typically it will be destroyed in the explosion.

Ex e is increased safety. This is basically using components with a lower risk of igniting anything. For terminals, this may for instance be spring terminals that will not come loose with vibration, or other techniques.

Ex i is intrinsically safe. It does not contain enough energy to ignite the gases. This is achieved by limiting the amount of energy in a device, so that a spark can not ignite any gases. This is usually achieved by limiting current and voltage, and controlling inductors and capacitors.

Ex m is moulding the device. This may be embedding it in resin, so that no gas can access components, or similar.

This is a brief overview. Your device would probably not be Ex d. In short, you should combine many techniques. You may for instance embed the circuit board in epoxy, making that part Ex m, use Ex i for the actual sensor, and perhaps Ex e for the sounder. A think conformal coating is probably not enough to achieve Ex m, as for instance a burning IC will put a nice hole in that conformal coating.

You will also have to define if you accept faults. As it's a leak detector, you will probably end up in zone 2, lowering the requirements. Note that the plant owner (e.g. home owner) is responsible for classifying zones - which I doubt many home owners actually do.

In addition you should look at gas group. Probably IIA (propane and so forth), and temperature T3 will be applicable to your use. This is especially important with protection method Ex i, as hydrogen for instance wreaks havoc with your available energy.

This PDF gives an broader overview of the protection methods than I have in this answer, but it merely touches the surface as well...

In short, based on the questions you are asking, you are not qualified to make such an device. The question basically tells us that you do not understand the Ex regulations, and have no insight in the different protection methods. I have some insight, and have engineered electrical systems in Ex environments, but I would in no way be able to safely design a complex device.

• You refer to "Ex d" and "Ex i" etc, which are presumably standards. I think it would improve the answer if you stated where the standards are from. – Hearth Feb 2 '19 at 16:04
• Great answer, though I'd add to it the need to protect against any internal battery or capacitor failure causing an ignition point. While Ex i would be a great goal for the OP, I'd assume that meeting Ex e is the most common in home environments. – Jack Creasey Feb 2 '19 at 16:28
• @JackCreasey Of course, all components needs some kind of protection. In home environments, I doubt any kind of Ex equipment is common. Tho, the easiest way to solve this is to make the sensor the only piece of equipment inside the zone... – vidarlo Feb 2 '19 at 16:33
• It's probably not EX approved in any sense, no. That doesn't mean it's not safe for its intendet application. A ignition source and gas leak at the same time is unlikely – vidarlo Feb 2 '19 at 22:04
• @MahmoudRagab Plug-in residential sensors like that are intended to be continually operating, and they alarm at something like 25% LEL. Therefore, the house should be ventilated and gas shut off before a spark can cause an explosion. – user71659 Feb 4 '19 at 2:42

I am working on prototyping a gas leak detector device to be used in homes

If it's for use in the home and, nothing else in the home is designed to be Ex rated, then what makes you think your device needs to be Ex rated? After all, if reasonably designed (non sparking or arcing and can't produce ignition temperatures on internal surfaces), then what you are providing is a safeguard to the home thus, reducing the possibility of gas ignition and not increasing them.

But if this doesn't convince you try looking up ATEX Ex zone categoriess such as this: -

And I think you'll agree that if it's in the home in falls into (at the very worst) into a Zone 2 operation. Then if you scan across to the level of protection column you'll see that it says "normal" for the type of equipment and this means: -

According to the ATEX directive, an EC‐type examination certificate is not mandatory for devices of category 3 as specified for use in Zone 2

The above taken from this document entitled "ATEX - Source IEx" and the phrases you are looking for a "self-certification".

In other words, basic good practice is required with documentation that is suitable for a regular CE technical file. No formal certification by law PROVIDING that you supply equipment into homes governed by ATEX regulations (a geographical thing that covers Europe and other areas).

• Thanks alot ,my concern was that a spark might be initiated when the buzzers turn on and off, but I saw this photo of a chinese device where it has a lot of windows in the outer casing and the whole circuit can be in contact with the gas as it flows inside,is such a design safe? ae01.alicdn.com/kf/UTB8bI1BG5aMiuJk43PTq6ySmXXaP.jpg – Mahmoud Ragab Feb 2 '19 at 21:51
• As I said in my answer, a gas leak detector improves safety so, providing it has been reasonably well designed, it is a net benefit to the environment. I can't assess a design remotely on minimal documentation. – Andy aka Feb 3 '19 at 9:59
• The OP said "for use in homes", which could be subtly different from "for use in the home". e.g. Perhaps they are designing a piece of portable survey equipment to be used by gas company technicians responding to calls from customers who think they smell a gas leak. If that were the case, one of the Ex categories could certainly be applicable. Perhaps @MahmoudRagab can clarify. – pericynthion Feb 3 '19 at 21:17
• @pericynthion Here I am taking about unit designed to be installed inside the home permanently and not portable units, however what will be wrong about taking the device from one home to another ? – Mahmoud Ragab Feb 4 '19 at 22:29
• @MahmoudRagab what pericynthion might be alluding to is that a design for a piece of handheld portable equipment may be subject to a more stringent Ex classification because it may get used (contrary to handbook instructions and labeling) in a zone where the average probability of an ignitable gas and oxygen mix is higher than that of a home. – Andy aka Feb 5 '19 at 11:45

Explosion Proof in electronics means to withstand 🧨 electrolytic capacitor or arcing switch in a vacuum sealed container but basically means different grades of pressure and strength but may also mean non-arcing with combustible gas, so this may involve gas tight seals that do not allow penetration unlike Teflon seals which release H2 from sealed boxes with SLA batteries to relieve pressure during controlled charging. You can imagine an outdoor application with electronics with arc suppression and SLA battery backup for a Wireless repeater might be explosive without some venting or a Teflon seal to resist water but allow H2 to be released.

There are several design-specific criteria; Rigidity, moisture seal, explosive gas seals, arc avoidance, partial discharge from ESD externally, or Partial Discharge (PD) internally from humidity and contaminants reducing the breakdown threshold <1V/mm.

# Gas Leak detectors do not imply explosion-proof

Normally even expensive home gas detectors warn you to keep gas venting batteries away for combustible gases !!,

# So your requirements are vague

What spec’s do you want to meet?

The ratings for safety tradeoff the risk for Partial Discharge or ESD due to dust and static generation and the level of exposure to combustible gases.

Which gasses do you want to detect? A combustible gas leak detector may not detect toxic Carbon Monoxide as the sensors are different. To prevent a gas leak requires a solenoid before the flex hose and not after as inside the furnace, if the flex hose was damaged by impact of moving heavy equipment, the gas alarm may not prevent a leaking house from blowing up!! Although a gas leak inside a furnace may be possible detect and sound alarm and shutoff the combustible gas source.

However, semiconductor combustible gases can detect many including any or all the following:

Acetone Alcohol Ammonia Benzene Butane Ethylene Oxide Gasoline-Petrol Halon Hydrogen Sulfide Industrial Solvents Jet Fuel Lacquer Thinners Methane Naphtha Natural Gas Propane Refrigerants Toluene

For Hydrogen gas vapours H2 has a lower explosive limit (LEL) of 5 % so up to 1,000 ppm or 0.1% “May” be considered the safe yet warning limit may be 10,000 ppm and >=4% with any static discharge may explode. Other gases may be more volatile. So accuracy is not uniform for all gases.

Normally “any” conformal coating will not do, for preventing flashover as most plastics are hygroscopic , although they extend the life in some harsh environments.

Even epoxy sealed plastic IC’s once failed below freezing. They eventually absorbed moisture and failed when frozen so ceramic IC’s were offered until Sumotomo’s epoxy formulation and process was developed. When Plastic IC’s first came out, they were only rated 0 to 70’C , now the improvements from Japanese R&D made it possible to cover the wider temp range.

# Hygroscopic resins

Nylon, ABS, Acrylic, Polyurethane, Polycarbonate, PET, PBT

# Non-Hygroscopic resins

Polyethylene, Polypropylene, Polystyrene, PVC

Normally an explosion proof container is a rugged sand-cast aluminum case design to withstand high pressures. Better products use an epoxy coating. So moisture sealed alone is not adequate to prevent any possible explosion from an electronic failure.

If you need the best low capacitance moisture blocking conformal coating then in Aerospace they use Paralene, with vapor deposition, IC’s use special epoxy formulations and clean room procedures. The other coatings when thick enough may extend poor performance life such as silicates , acrylics and silicone but may not perform as well, if thin yet too much can cause crosstalk and capacitive loading.

The science behind explosion proof is determined by the contamination level of a good insulator degraded by moisture /and or dust where the low dielectric constant contaminant breaks down by accepting charges faster than the medium of higher dielectric content resulting in what is well known to those familiar with Partial Discharge , PD which is the precursor to an ionization Discharge or arc or dielectric breakdown of the insulation.

The test method depends on environmental stress levels of humidity and hygroscopic rates of various plastics with contaminants that may absorb moisture which has a polar dielectric constant about 20x greater than most plastics. The contaminant levels only need to be in the parts /million or PPM for PD to occur and this leakage rate with dielectric constant creates a unijunction like oscillator that may Discharge at low ratios of the expected breakdown kV/mm or V/um or mV/nm. With cycle times of many minutes, becoming more rapid with excitation ratio relative to Vbreakdown.

The test method simple using worst case ambient contamination (dust, moisture, salt spray) with a slow ramped voltage and determine the spark noise on an AM or SW radio nearby or using a scope probe shorted to its ground clip, wrapped around the conductor to detect the PD current pulse. The derating factor of stress voltages either conducted or induced to PD activity determines the safety margin after high temp/high humidity soak to accelerate moisture ingress.

The specific test procedure may vary from this, but the science of determining margin to trigger threshold is the key safety factor.

Exactly the same science is used in power transformers whether dry or oil filled and yet they only test to BDV or breakdown voltage instead of the optional test for PD. PD activity is monitored by H2 dissolved gas and yet so many transformers blow up every year that could be prevented with PD monitors and often only installed in million dollar transformers, yet it is so cheap to monitor.

• "most plastics are hydroscopic in the seals or plastic" I'm not sure what you mean by that -- is it an editing error? And are most plastics really hygroscopic? (Note the spelling, by the way: it's a "g", not a "d".) – David Richerby Feb 3 '19 at 23:30
• TY I’ll add references – Tony Stewart Sunnyskyguy EE75 Feb 3 '19 at 23:32
• @SunnyskyguyEE75 Thanks for the great answer,what about silicon resin ? – Mahmoud Ragab Feb 4 '19 at 22:41
• It depends on dust during application and thickness. Dip is better than spray. – Tony Stewart Sunnyskyguy EE75 Feb 4 '19 at 22:45