I bought a couple of Sonoff switches earlier this year and had great difficulties getting them to pair to my network. I contacted customer support and they were little help so I gave up on them. Cheap devices just don’t warrant a lot of my time.
This weekend I had an idea on how to use them so I tried again. I was inspired by something I read about how the presence of a 5g network sometimes interferes with pairing on a 2.4 GHz network. In any case, I got one of the Sonoffs to pair but not the other.
To test it I hooked the input up to 120 V, 60 Hz, US. I hooked a DVM up to the outputs. On one of them I read 34 V in the OFF state. The other reads 16 V when off. I cannot turn on the 34V unit because it won’t pair. The 16 V unit reads 115 V when turned on with the app.
Does anyone have experience with these? While 16V-35V when off is not terribly alarming, it does seem to indicate a serious design or quality problem.
It’s not a measurement error. I posted because 100% open circuit is what I would expect so getting a consistent 16V through seems to indicate a fairly serious problem either with leakage through the relay or something else.
Something was definitely seriously wrong. I took one of the units apart today. I didn’t see anything obviously wrong (cracked board, loose wires etc., but I did notice I had wired the input incorrectly. I had line and neutral reversed. I know that makes a difference, but I don’t see why it would for this device since there is no ground. I would be more careful for an actual project but this was just an experiment.
I corrected the wiring, made sure everything was correct. Plugged it in (in the off state), and the output voltage was 1.4 V, much more reasonable. Then I switched it on with my phone. The thing BLEW UP! I’m not kidding. There was a flash, a loud pop and a little bit of smoke.
I’m an engineer (but not electrical). This is not the first time something didn’t work the first time. I didn’t panic. I left it plugged in and investigated a bit. The output voltage dropped to zero. The GFI circuit the input was plugged in to did not trip. So whatever the issue was internal to the Sonoff device. It’s possible I damaged it by reversing the wiring initially. If so then that by itself should be a warning to anyone using there devices.
I took it apart again and could not tell what had blown up. There were no arc char marks or anything like that. I could probably do a bit more thorough component by component investigation but its not worth it.
I guess I’d give these devices another chance because they are cheap.
I know what I am doing. With AC polarity only matters if there is a reference to ground. In the US this is a relatively new thing and many residences are not properly wired such that polarity matters. Selling a consumer device in the US that is not protected against reversing line and neutral may not be illegal but is incompetent.
In the US “reference to ground” is provided by the neutral line being connected to the ground bus bar in the breaker box. It’s been like that “forever”. By that I mean as long as standardized wiring has been around. My 100 year old house has a neutral wire that is tied to ground. And that’s literally what makes the wire a “neutral” in the first place.
I think you might mean that polarized plugs & outlets are a relatively new thing in the US. They have only been around for a few decades… And as far as the equipment cares it makes no difference which wires are connected to neutral or hot since the voltage (therefore, current) reverses direction and motors/resistive loads don’t care. They are used to make the equipment safer for people by forcing the switch to be in the hot leg so the equipment really is de-energized if the switch is off.
And most electronics likely doesn’t care either. The problem is that you don’t know how they hooked up the electronics internally to make the device work as it was designed. They did go to the trouble of specifically telling you how to hook it up right on the case of the device. So they probably had a reason.
And you don’t really know why it failed. It could have been because it was initially hooked up backwards but It could be just as likely that you hooked something up incorrectly the second time as you did the first time. Especially as you say you saw an arc but only after you turned on the output relay and that you couldn’t see the results of the arc on the device. If you saw electricity escaping from the confines of it’s insulation (an “arc”) then you should see effects of that somewhere.
Since you didn’t get any arc until you switched on the relay and the only thing that is on the output of the relay is the output traces and terminal it’s possible that you had an issue there.
Lastly as far as the GFCI not tripping is concerned, do you know how those work and what they are designed to protect? They aren’t designed to protect the equipment from you hooking it up wrong and causing an arc. They are designed to prevent you from getting electrocuted if you touch something that you aren’t supposed to (intentionally, through a device failure or inadvertently because you didn’t know enough to not touch “the bright shiny stuff”). The way they do that is they measure the current on the hot an neutral and if the difference is too much (just a few milli-amps will trip it since that’s all it takes to kill you) then it decides that the extra current is going thru you instead of the equipment and trips the internal GCFI circuit breaker before you die. If the current going in equals the current going out then it will never trip even if the entire world is arcing and melting around it.
You stated better what I meant. I know ground and neutral are tied together at the service entrance (could be breaker box). Until a couple of years ago I lived in a ~80 year old house and the only ground was at the service entrance and where I put in grounded outlets in the early 90s (inspected and to NEC). My point was that an individual device should not care about line/neutral if it does not also have an internal ground reference, which is essentially what you said.
Upon further investigation tonight I think I see a slight char around the edge of the Sonoff board. They run a couple of heavy (14 ga.?) wires through the unit one from the neutral input to neutral output and the other from the line output of the relay to the line out of the device. The solder on one of those wires is very close to the edge of the board. I suspect the short happened not so much from damage at having originally hooked them up backwards but rather from a stray strand getting too close to the thick solder traces. That would have been my mistake but also something others could easily make.
And yes, I understand how GFCI circuits work. My point was a short to ground should have triggered it. A short to neutral should have triggered the regular CB. Maybe I’m not explaining my logic well.
I was about to toss both Sonoffs in the trash but now I think further investigation might be interesting. Don’t worry, I always protect myself.
I agree. But as you said there is no ground lead on those devices do there is no path to ground except thru you or possibly the device that is connected to the output. But that is unlikely unless that device is somehow also compromised.
Not if the fault current didn’t exceed the trip set point of the breaker. That’s another misconception many people have. If you hook up a load to a standard 15 amp breaker but only use wiring that is rated for less than that yiu can still get into a situation where the wiring will burn up becsuse it cant dissipate the heat caused by the current before the breaker trips.
Or it could have had a high resistance short that caused very localized heating and could cause an arc but not be high enough to trip the breaker. Think of it like a welder where there is an intentional high temperature arc but the breaker doesn’t trip.
That is more likely the case since you said you thought a whisker was touching the solder joint. In that case the breaker wouldn’t trip because the melting/burning/arcing pint of that little strand of wire was way less than the breaker trip set point. IOW, it acted exactly like a low current fuse only the arc wasn’t contained inside the fuse body.
And the manufacturer can’t be held accountable for that mistake.
I thought about this long and hard before responding because quite frankly I don’t think you are worth the trouble but on balance the things you said were unsupported and quite negative to a whole section of our society that you are judging on the basis of a couple of run in’s (dunno, I’m guessing, but at least I state when I’m unsure of facts).
@finity, @tom_l and @wellsy gave you relevent advice. finity especially as he works in the US in the electrical field and has 34years (finity, correct me if I’m wrong) of experience at the sharp end of electrical installations, fault diagnosis and repair. So in my mind a) he’d said everything that needed saying and b) he had the most relevant experience to comment.
So I was just endorsing the already stated opinions.
You then chose to send me a PM (do you have some doubts perhaps where you could not air this in open forum ???) Where you actually make my point that clearly you did not understand from my earlier post regarding assumptions. As below : -
Just to clear up some of those assumptions (not that I really want to talk about this and hence my reluctance) I have a Degree in Electical Engineering Science, with a Control Systems Bias.
I have been employed in Heavy Industry (Steel / Coke Production (from Coal) / Petrochemical) Utilities (Water disinfection (Waste and Potable) using UV, Ozone, Filtration, Aeration and RO) Pharmaceutical (Ciba-Geigy, SKB, Astra-Zeneca …) Food Industries from biscuits to coffee.
Part of my work involved the development and deployment of Earth Leakage Protection Equipment (RCD, RCDO, ELCB … they go by many names). As part of my institute membership the local committee would put members forward with particular expertise when requested by local fire/police authorities. (I often fear it was mainly cos I knew how to operate the test equipment )
I had previously been on ‘teams’ (this is rarely a one man show) with industrial settings which also includes expert testimony should a prosecution be sought.
I was nominated for a few domestic ‘incidents’ and I attended.
I already had a resonably high opinion of the fire, ambulance and police services but they all went up a couple of notches in my book (especially the fire guys).
Industrial cases had never (and still don’t) affect me, but I ducked out of domestic cases following the one mentioned. These guys deal with this shit on a regular basis and they rarely talk about it with ‘outsiders’ because it’s not something we want to hear.
The inherent thing in ALL your statements is that “polarity doen’t matter with AC” which you still seem to maintain even though EVERYONE is telling you different.
You continue to assume other data, even drawing profiles of people based on 2 or 3 sentances.
Your conclusions are erroneous and cast ANY further argument into the same level of doubt.
Although I have done (professional) work in other countries, none of them have been the US Therefore my comments were justifiably limited. finity is far better placed and given even a cursory examination of the postings on this site you would have seen his expertise.
Just to spell it out, you crossed some wires because of an assumption, some ‘other’ guy could have made some equipment (that you plugged in) that earthed to ground using a borrowed neutral. Zap.
Two assumptions - One electrocution (possible ?)
I was making the same mistake! Mixed L - N, assuming AC shouldn’t matter. I attached some strip led lights to the Sonoff but they seem to have a faint glimmer when power was off. The box is difficult to access but I am pretty sure this should fix it.
)
