[Ben Jones] just started a new site called Maker-Guide, where he makes some very informative and well produced DIY videos on anything from homemade photography hacks, to controlling an outlet using a solid-state relay.
It’s not the first time we’ve seen a relay controlled outlet, or this one… but it’s certainly one of the cleanest jobs we’ve seen. In his video guide, [Ben] shows us exactly how to fit a standard solid state relay into a regular outlet box, and easily control it with an Arduino Uno.
It even looks like there might be enough space inside the box for a small wireless setup — maybe using a Trinket even? What about using Power-Line networking to control each box via LAN? Could be the easiest home automation implementation yet! Well, aside from certain NEC (national electrical code) concerns of running high and low voltage in the same box…
Anyway if that wet your whistle, check out the great video guides after the break!
Curious about home automation? Last month one of our writers, [Josh Marsh], wrote a great feature asking what is “Home Automation”?
I know this gets hashed over every time one of these get’s posted, but for my sake: if the code says you can’t have high and low voltage in the same box, just where exactly are you supposed to use one to switch the other, in the open air outside of the box?
Probably in a box that is partitioned for the two sections like this:
http://www.hubbell-automation.com/content/products/instructions/instructions_files/cx1624lightingcontrolpanelsinstructions72_00474reva.pdf
Not finger safe screw terminals. It’s in a metal box. Really?
Nothing in the code requires IP2X contacts for a device not meant to be opened while energised.
The metal boxed is earthed. More to the point, the metal box is an outlet box, the type specifically designed for mains wiring.
So yes, really! What would you prefer? A plastic box where a loose wire could potentially go unnoticed as there’s no contact to make with ground thus being unable to trip the earth leakage device?
What earth leakage device, aside from the breaker? There is no GFCI or arc fault circuit here.
“The metal box is an outlet box, the type specifically designed for mains wiring.”
Yes, but not designed to shove an unsecured, not heat sinked SSR into so that it freely floats around.
Low voltage so far as the code is concerned is anything below 700v. I know we’re all used to microcontrollers and use “low voltage” to refer to anything below 12v, but that isn’t really accurate. Medium voltage is below ~80,000 and above 700…
I’m not sure which specific part of the code you guys are referring to, but the only stipulation I’m familiar with is that all insulation must be rated at least as high as the highest voltage. So if you have a cable tray with 27,600v and want to run a 12v conductor, that 12v conductor would have to be a big thick cable with at least 27,600v rated insulation.
Maybe I’m incorrect. If so, please let me know because I’ve seen this pop up a few times and have never been able to find the actual part of the code people are referencing.
Same with our code, except we also recognise Extra Low Voltage as anything below touch potential (<50VAC or <110VDC). Hence same scenario. You run a 24V battery inside a conduit with a 240V wire then the insulation must be rated for 240V.
Not sure about where you live but every electrical code I’ve seen requires segregation, not a different box. This can be either by means of separate enclosures, double insulation, or physical barriers. You would need to carefully read the wording of your codes to understand what is permitted.
Funny legalese creates all sorts of interesting situations. For instance where I live we need double insulation for home wiring. Plastic conduit can work as one of those insulations so you can strip the overall sheath off the wire if you’re trying to cram it into a tiny conduit. Conduits are required for mechanical protection. Metal conduits are best for that, but then that doesn’t count as double insulated.
A frequent poor installation I see is thus single insulated cable running through a metal conduit where mechanical protection is needed.
Yeah, Europe is all about double-insulation (although you might be somewhere else) which I have always thought was kinda interesting. The whole point of metal conduit is that you ground that too, except it’s either in your walls or on your walls and well grounded and designed to contain any fire…so in the US, singleinsulated in conduit is the norm. Homes are wired without conduit and typically with extruded jacket over a bare ground and neutral and hot wrapped in thin paper, solidcore. I wish residential was required to do conduit.
The funny thing is that having to double-insulate things must whack out the ampacity, as you are really killing the thermals on bigger feeders.
Well, you don’t have much problem with that because in a 220 volt circuit you only need half the current and get 1/4 the heat. But even if you double the heat insulation, you can still push 3/4 the current through.
This is idiotic on so many levels. An SSR in a plastic in-wall outlet box with no room for a proper heatsink, whats the worst that can happen? http://www.moorepage.net/hottub2.html
Maybe that’s why he didn’t use a plastic in-wall outlet box. Look again.
Thats just a proof of concept genius. If you were to use this in the wall of your house as it is suggested its going to plastic.
You can use metal boxes in a wall genius.
And all of you have no idea what you’re talking about, as a steel box is a poor-ass heatsink.
For what it’s worth, I set up my baseboard heaters to be controlled using a solid-state relay. The relay was very conservatively rated, both in terms of both current and voltage, and mounted on a generously sized heat sink. The system worked slick for a good two years, then one day, no heat.
I checked the closet where the box was affixed to the wall, and found the ceiling coated with soot, and soot trails up the wall. The contents were burnt to a cinder, and I felt darn lucky I didn’t have a house fire.
I’m not sure why it failed, perhaps a line surge. I had left the lid off the box so I could easily monitor a led that indicated the output state. Big mistake. Make sure your setup is stays sealed in case something should go wrong
those SSR relays are awesome but I get the impression they are overpriced. I just get a hunch that they are more expensive than they really are
They are kind of, bear in mind they aren’t just a triac tho but that they come with all of the interface stuff between a high impedance ELV DC input and the AC output.
Opto-isolated, tested to agency standards, cheap insurance to protect the control side. I use these all the time in my designs, for AC and DC control and HV (up to 5KV) by connecting in series. You have to understand the limitations like switching time, isolation and on resistance.
Of course you have to watch out for cheap chinese knockoffs (CCK). They will stamp them with CE or UL marks without ever seeing the inside of a testing agency and enforcement is way behind the curve on this.
$5 for a chinese SSR? Thats cheap for opto-isolated switching. The good SSRs are much more expensive $40 to $80 easily.
DIY Outlets or: How I Lost my Insurance and Love the Fire
If you are going to pull anywhere near a full load for a 120v outlet you will need to have a heat sink on those SSR modules. They are not rated to be ran without being attached to a plate.
Another option is to use a smaller SSR and a 120v coil 20A mechanical relay, This will give you full load and no heat sink required.
Yeah, the omission of a heat sink really caught my eye!
The heat sinks are quite large too and not designed to be stuffed inside a box with no air movement.
Sometimes (often), HAD posts this kind of stuff to generate “discussion”, despite the risk of others replicating said bad or dangerous ideas.
How else is the gene pool supposed to self-chlorinate???
Note the derating from 25A to 3A if you use it without a heatsink!!
I won’t be surprised that the SSR is dropping 2V or so when switched on,
so expect easily 50W or so when it is used under full load.
From SSR 24-480V AC 25A datasheet (cheap Chinese $5 SSR)
– If SSR work without heatsink, the maximum continuous load current
never be over 3A.
– If SSR work with CD-RR012 heatsink, the maximum continuous load
current could be 5-10A.
– If you want to use those SSRs with over 20A current, you should be
careful that they should be installed on a ver good heatsink. It is
strongly suggested that add an additional fan to enhance the air flow.
Heat resistance must be less than 2.5C/W. Over load and over heat will
destroy the SSR.
I think the metal box is sufficient as a heatsink (at least for resistive loads). But, using a plastic box would be a mistake.
I’ve used the 25amp version of that brand SSR on several cooking controller projects which can draw up to 10 amps continuously. At 10amps continuous, using just the metal lid of a plastic enclosure as a heatsink, it gets a bit warm but not too hot to touch. At lower loads, it doesn’t even get warm. Using the 35 amp or higher version of the SSR, or putting in a lowish-amp fuse would probably make this a non-issue.
Other outlets done using a WiFi controller connected to the cloud …
This one uses the Electric Imp …
http://www.instructables.com/id/Impowered-electrical-outlets/
This one uses the Electric Imp …
http://www.instructables.com/id/Electric-Imp-to-control-8-or-more-electrical-light/
Mechanical relays are much safer than a SSR and much much much much safer than a SSR without heatsink. This project is a house fire waiting to happen.
Not true at all.
However, people using SSR’s without understanding them at all — and buying cheap no-name SSR’s instead of Potter&Brumfeld / Teledyne / Crydom / Omron / Opto22 / etc. — is indeed a recipe for disaster.
Both have advantages over one another. The lack of heatsink is stupid, and according to NEC, if you want to do this, it can’t be in the wall — and it shouldn’t be hardwired, either.
Relays can have coils blow up too due to surges and other things, and the contacts do get chewed up over time. SSR’s are not considered dangerous in industry.
The biggest issue with this install other than the heat sink issue is that NEC has a defined space issue for electrical boxes. Boxes are measured in cubic inches and the bare minimum for a duplex outlet with 12/2 wire is 18 cu in. Add that SSR to that box and you don’ meet code for space alone.
To second the above statements: A SSR for a wall outlet is a bad idea.
SSRs must not be installed without a /very/ good heatsink, designed for the full rated outlet current of the outlet in this case.
A metal box is in no way good enough to dissipate approx. 2 V times 25 amps power loss.
To properly install a switched power outlet, use a contactor inside an appropriate enclosing. Another option, a normal mechanical relay, is also very likely not sufficient, except if it can handle the switching of inductive loads which have to be expected.
Of course, you can use a smaller SSR to drive your contactor… just don’t try and PWM it; welding the contactor to itself, while fun, becomes expensive quickly :-)
When did he tell anybody to install this in their home??
I see a lot of people getting up in arms about NEC code and having something like this done in a home built in. All I saw was someone build an arduino controlled extension cord. I didn’t hear him tell anybody to do a permanent install like this in their home. Why would one go through all of that when you can just buy UL listed wifi devices.
This could easily be tweaked to be made a lot safer for art projects or commercial/holiday displays. Use 4×4 deep box for just the relay. Include heat-sink maybe even a fan. If you did all that it would be just as safe as powering that little arduino for any extended period of time, I mean I just read another HAD article today about how dangerous even our simple hobby power supplies and devices can be.
Question for you guys: Do SSR’s only heat up when they are active? I have a pump – 110V 7A that I want to trigger with DC and run for 2 minutes then be off for over an hour, and continue this process indefinitely (flood/drain irrigation). I am considering this type of build in a 2 gang steel box (perhaps with thermal paste on back of SSR). Do you think this duty cycle would be too warm to run without a heatsink (indoors, but not in-wall)?
Yes, SSRs only generate heat when ON. It is approximately 1.2 watts per amp being switched (when using SSRs from good suppliers). CHEAPER SSRs might use an internal Triac, or they have have a under-weight metal back plate. Then the thermal rise is much higher. So, your 7 amp load will generate 7 x 1.2 watts = 8.4 watts of heat. 60/8.4 = 7.14C/W is the correct size heat sink. A 5 C/W heat sink (or lower) is better. A 10 C/W heat sink is too weak. Your pump will have INRUSH at turn ON. For most AC motors = 7amps x 8 = 56 amps in rush. Sometimes called LRA (locked rotor amps). Buy a 50 – 75 SSR for this. Your pump will have VOLTAGE SURGE at turn OFF. Install a MOV 330LA40B across the SSR’s power connections.