EEVblog 1489 - Mystery Teardown!

Teardown of a bizarre looking Banshee 343 Ultrasonic Gas Leak Detector

Thumbs up this video and pinned comment if you want to see a countdown timer refit project!

Support the EEVblog on:

Buy anything through that link and Dave gets a commission at no cost to you.

Donate With Bitcoin & Other Crypto Currencies!
#ElectronicsCreators #Teardown

[EEVblog]

Thumbs up this video and this pinned comment if you want to see a countdown timer refit project!

[Satelitko]

Finally, a proper case for a Raspberry Pi! :D

[toddphelps]

I was an engineer at Net Safety for over 10 years (a Canadian company in fact) - it's a blast from the past seeing the "Banshee" here! I worked on the electronics and firmware for every gas detector we designed from 2004-2015 except, ironically, this one - it was a rebranded product. We all had the same reaction to the looks!

[mikeselectricstuff]

The principle of this type of protection is that they assume the thing will fill with gas and ignite, and the enclosure must not allow any flame to escape. (During the approvals process, they test this using either hydrogen or acetylene + oxygen, with expensive gas analysers to get an accurate mix to meet the standards)
This is why it has a long, tight-fitting flange on the main enclosure joint. The terminal area has less stringent requirements, so no flange, but also no active components.
I think the external glands ( called line bushings) specced for use with these are designed to be gas-tight and not allow gas to go through the cable, hence the lack of per-core sealing here - they typically have a bulkhead filled with epoxy, with the wire stripped bare where it passes through the resin.

This is not the same thing as intrinsic safety - with IS, the protection is provided by limiting voltages and currents by design, whearas explosion proof (Ex-D and Ex-E from my hazy memory) is used where this is not practical due to power requirements.

The soldered fuse is not uncommon - a major part of design for Ex approvals is not so much designing it to be safe, but designing it to be easy to prove it's safe. It is standard practice to include fuses, zeners etc. that have no chance of ever acting in normal or even fault conditions, just to make it easy to compartmentalise parts of the circuit to make it easier to evaluate for approvals.

Including a fuse is a simple & cheap way to prove that the current can never exceed a certain amount under any circumstance - the standards define how much current can pass a fuse of a particular rating, the number 1.7x rings a vague bell. It also reduces the risk that some idiot replaces the fuse with the wrong value ( "less danger of it blowing guv....").

[bobcunningham]

This construction brings to mind the commercial nuclear plant radiation sensors mounted within the concrete reactor containment that had to survive a LOCA (Loss of Coolant Accident). The sensors needed to survive not only the pressure pulse of the explosive release of steam, but also the more troublesome thermal heat pulse, not to mention the radiation pulse it was intended to measure.

Those same standards applied to everything within the containment connecting to the sensor, including conduits, cabling and most importantly the connectors. The main approach was to use lots of secret-sauce epoxy everywhere, so the housings themselves didn't need to be built so robust, which particularly mattered for the sensor, as excess material affected its sensitivity, and a large neutron release could make the housing itself radioactive.

Even the heat-shrink enclosing the connectors was beyond special, where a 10 cm length would cost $200. The coax was an entirely different story, as no manufacturer was willing to certify their coax for LOCA. So we had to do the certification ourselves. While we had access to a physical LOCA simulator for temperature and pressure (which sounded like a bomb going off), we did not have one for radiation. The highest radiation flux we could find was behind the beamstop of a small linear accelerator, where we made a small closet and hung coils of the coax candidates, to be pulled out and tested/characterized every month or so, along with measuring the total radiation dose.

To keep the cabling to a minimum, we sent 2-4 KV down the coax, with the radiation level encoded on the return. During the LOCA event, the pressure and temperature pulses would significantly change the capacitance of the coax, an effect the overall system had to both survive and account for with great precision. The coax runs could be up to 1 km in length, to get the measuring instrument far enough away from the containment for it to continue to function through the event.

[ACElectrode]

I design industrial explosionproof electrical systems that something like this would be used in. Aside from the obvious heavy duty nature, there is a whole field of study you could jump into regarding flame paths through all the seals, gaps, conduit entries, etc. That is a HIGHLY engineered piece of equipment and probably costs as much as a house! Good on ya for getting one in your hands!

[baronbarbaron]

Hello :) In the factory where I work, there is a unit where we produce ethyl alcohol. We use Ɛx rated appliances there. Many of them have similar soldered fuses. When I asked one of the manufacturers why they were like that, they told me that it was a requirement of the insurance companies and a policy of our group as a customer. The idea is that in the event that this fuse burns out, the device should not be repaired, but replaced with another one. It's simply not worth the risk of putting an appliance that has had a technical problem in front of an explosion that will surely cause massive damage and kill people.

[ddEEE]

You also need to put an eccentric motor vibrator inside that triggers when the countdown reaches 5 seconds or thereabouts, so as to intensify the effect.

[tddotnet]

"It's not a bomb it's a clock!" (also, Dave that would make the IDEAL kitchen timer for the Mrs...) 🙂

[TheDefpom]

You've got to modify the display to count down... maybe it can do blanking too so it flashes when it gets to about 10 seconds or something too.

[mikefochtman]

Worked on 'explosion proof' fans and such in the Navy. They aren't completely 'gas tight', but the idea is that any explosion inside won't ignite the atmosphere around them. So for example, submarine battery fans (H2 hazard) had to have very close tolerance fits. H2 might actually get inside, but if it ignites the flame can't 'blow out' through any of the tiny cracks/ crevices. Sort of like how a steel screen placed over a bunsen burner, the flame won't go up through even though hot gases flow up through the screen.

And of course the casing is strong enough to not rupture if the inside is filled with the 'perfect' mixture of explosive gas/air.

So that interface cable might let gas seep along it between all the individual conductors, but that's fine.

[SeanBZA]

Probes into the sensor heads are there to have the heater blocks screwed in instead of the blank probe, to heat the actual sensor itself in use in icing conditions. Self cleaning probably uses those probes, and then a separate port into the case that provides the air pressure for cleaning, via stainless steel pipes and a spiral tube that connects the ports to the top. That is why there are 2 recesses in the probe, one for the cleaner, not fitted and drilled to spec, or the heater, fitted to all but in temperate use only the dummy heater block.

[seeigecannon]

At 16:00 you mention seeing the home-made optio-isolators before, but didn't remember where. They were in the electric fence power supply you cracked open a few years ago.

[rubikmonat]

You will probably find the TO220 devices on the heatsink are the heater, for preventing icing in cold climates. Fuse is soldered in so people can't replace it with a wrong item reason is:. That opto isolated area is all about limited energy so there's not enough power there to generate a spark at the microphones in any possible conditions. The power will get across there via that fuse then there will be zeners to limit voltage and blow the fuse. There is also a mandated clearance distance around that intrinsically safe circuit, which is why the homebrew optocouplers exist.

[zaphodb]

Ya need to have a big motor and weight in there, so when the counter reaches 0, it vibrates like mad, dancing across the floor, then stops, and a little voice says "bang."

[TheCritterWindow]

Make the count down timer light up and start counting when touched or moved. Just think the fun you could have.

[paulpanda]

I swear this looks like the training drone that Luke skywalker and obi wan kenobi use in Star Wars Episode 4: A New Hope

[stevekemble]

Please do! Replace those sensors with red LEDs that light up when "armed". It would be nice if it had an audible hum too.

[NoName-efjq]

I love how excited he sounds through the whole video.

[davidellis]

Hi Dave, The Key to EP is not to pressure seal completely, but let the hot gases cool as they leak out so they can not ignite the atmosphere outside.