21
\$\begingroup\$

I want to install a relay to my fusebox which is controlling whole basement with machinery.

Relays that I have have got 2A limit each, and I have 32 relays. Can I just use all of them for the same line and assume the relay's limit is 64A ?

Is there any risk of doing this?

\$\endgroup\$
  • 45
    \$\begingroup\$ Do all of your relays make contact within the same microsecond? If not, what do you think will happen to the first one that makes contact... \$\endgroup\$ – PlasmaHH Jun 18 '17 at 19:29
  • 8
    \$\begingroup\$ what if one fails? or two? or x... dormant failures. Paralleling is ALWAYS risky \$\endgroup\$ – JonRB Jun 18 '17 at 19:39
  • 5
    \$\begingroup\$ worst case scenorio i can imagine is contacts in first relay getting fused. and it will pass current even if i try to open the circuits. Resulting bonus fireworks besides a simple pcb. \$\endgroup\$ – user2102266 Jun 18 '17 at 19:49
  • 10
    \$\begingroup\$ the downside is an extremely low MTBF. \$\endgroup\$ – dandavis Jun 18 '17 at 20:08
  • 2
    \$\begingroup\$ Making some own insertions into the fusebox is a change to the accepted distribution system. The approval certificate of the original system changes to a piece of waste paper, if you are not a certified contractor yourself, who has power to do the work and accept the changes. Check the legal side even if you happen to know enough to put together a working system. Otherwise the bill can be astronomical. \$\endgroup\$ – user287001 Jun 18 '17 at 22:03
44
\$\begingroup\$

Unless you can guarantee that all the contacts will close and open at exactly the same instant of time the only safe current you can assume is 2A - that is, the capacity of the first contacts to close, or the last contacts to separate.

\$\endgroup\$
  • 15
    \$\begingroup\$ Since they are mechanical devices where an armature has to physically move, and the speed of that movement depends on factors outside of your immediate control, no you cannot assume that anything you do can make them all switch at the same instant of time. I'm talking femtoseconds or faster. \$\endgroup\$ – Majenko Jun 18 '17 at 19:42
  • 3
    \$\begingroup\$ Or a high current contactor \$\endgroup\$ – Majenko Jun 18 '17 at 19:48
  • 5
    \$\begingroup\$ I am struggling to find a 35kVA cutter that isn't 3-phase. The best I can find is 3.7kVA. Are you sure it's 35kVA and not 35A cutting current? \$\endgroup\$ – Majenko Jun 18 '17 at 20:11
  • 16
    \$\begingroup\$ This is totally a job for a contactor. It's simply not in relay territory. If you're controlling machinery you're dealing with inductive loading which is going to seriously downrate your relays as well. You simply can't do this with parallel relays. It won't work. \$\endgroup\$ – Ian Bland Jun 18 '17 at 20:13
  • 12
    \$\begingroup\$ I don't understand. To me, "basic home automation system" means stuff like turning the lights on/off/dimmed, maybe a music system or some motorized blinds, perhaps a smart thermostat. How possibly does a basic home automation system involve plasma cutters and CNC machines? Any equipment you're going to be using with a 35kw plasma cutter is of a completely different character than what you'd use to control the bedroom lights. \$\endgroup\$ – Zach Lipton Jun 19 '17 at 5:22
41
\$\begingroup\$

Sixty-four little relays were turned on by a key;
One soon welded shut and then there were sixty-three.

Sixty-three little relays started glowing bright blue;
One of them turned black and then there were sixty-two.

Sixty-two little relays bore of amps a metric ton;
One said he'd had enough and then there were sixty-one.

...

(I can't be bothered to write 64 stanzas. I hope you get the idea. It ends with your house catching fire.)

\$\endgroup\$
  • 9
    \$\begingroup\$ Not enough poetry on EE.SE. +1. \$\endgroup\$ – Transistor Jun 19 '17 at 22:06
  • 1
    \$\begingroup\$ I do like your song, but why did you start with 64 relays, when the OP clearly said they had 32? \$\endgroup\$ – Dmitry Grigoryev Jun 20 '17 at 8:49
  • 1
    \$\begingroup\$ As usual, a brain fart of mine confused relay count with amperes. Good thing I don't work with high-power circuits! \$\endgroup\$ – Kroltan Jun 20 '17 at 12:29
  • 4
    \$\begingroup\$ @DmitryGrigoryev With 64 relays OP would be twice as safe. Of course, twice zero is still zero. \$\endgroup\$ – TripeHound Jun 21 '17 at 7:49
  • 1
    \$\begingroup\$ @TripeHound rather, 64 relays is equally safe for at least twice the time. after that, well... \$\endgroup\$ – Kroltan Jun 22 '17 at 15:58
25
\$\begingroup\$

Here's what will happen. When you activate your relays for the first time, the first one to close will have its contacts welded together by a massive overcurrent. Soon enough, other relays will close and hopefully distribute that current more or less evenly.

When you deactivate your relays, they will all open except the one with welded contacts. Since it will now be alone under full 64A load, it will soon blow up.

Now you have 31 little relays standin' in a line.

\$\endgroup\$
10
\$\begingroup\$

You're not allowed to parallel.

You can't parallel through relays because you can't parallel generally.

In US NEC, it is disallowed except for currents so large that single wires are not readily available. I don't have a cite but I'm sure the EU model codes have the same rule, since it's a basic physics problem.

And even then, each paralleled conductor must be fused separately. (For instance I have an 800A 3-phase service with conductors paralleled, each on a 400A fuse. The conductors are 1000kcmil, the next size up is hard to procure.)

They must be the same material and length and terminated in the same manner, so they have the same electrical characteristics - that's to keep current from favoring one path over another. (which would result in a cascade of fuse blows).

This "favored route" problem will be even worse if you have contacts making (or breaking!!!) not quite simultaneously.

And remember, those small electronics relays aren't necessarily listed for use switching mains voltage. Their ratings as a component only mean they can be used as a component in an assembly/product/machine, which must then go back to the testing lab to be listed as an assembled machine.

Just use a big contactor. If your controller don't have the oomph to throw the big contactor, use a little relay to throw the contactor's coil current. This does an end-run around the snubbing problem; you only need to snub the little relay (though it will reduce arcing on the little relay's contacts if you snub the contactor's coil).

Or if one contactor is too expensive, you may be able to use several.

Multiple contactors are ok on multiple loads

I am inferring that you have more than one load. It's not paralleling to use one contactor per circuit or load. If each of your loads is <=30 amps, this is easy and cheap.

Suppose you have a 30A compressor, you can control that with a 30A contactor. As it happens, the HVAC industry has plenty of Code-rated 2-pole 30A contactors with 24VAC coils in the $12 range. 24V transformers are also in the $15 range. Here's a combo. Note how some of them are designed to mount in the cover or a knockout of a standard junction box.

If you have several loads on one breaker, you can just use another similar contactor on another circuit. If the contactor is 2- or 3-pole, you can switch one circuit per pole - that's not paralleling.

\$\endgroup\$
  • \$\begingroup\$ Really? You still have to snub the coil on a £300 100A contactor? I would have expected that to be built in for that price... \$\endgroup\$ – Paul Uszak Jun 19 '17 at 13:00
  • \$\begingroup\$ @PaulUszak You'd have to check the spec sheet. This is an electronics forum and to this constituency, that would make sense. The other users of that contactor might object to paying for snubbing they don't need. \$\endgroup\$ – Harper Jun 19 '17 at 13:48
  • \$\begingroup\$ @PaulUszak: The optimal characteristics for a snubber will vary depending upon the circuitry feeding the coil. If the feed circuitry won't mind being hit with 2x the normal drive voltage, the coil may be able to switch much faster that if the peak voltage must be kept to 1.1x the drive voltage. \$\endgroup\$ – supercat Jun 20 '17 at 22:48
  • \$\begingroup\$ How cant i thick multiple answers :/ \$\endgroup\$ – user2102266 Jun 23 '17 at 5:52
3
\$\begingroup\$

This is why they make large contactors (http://www.newark.com/c/automation-process-control/contactors-accessories/contactors) or for a good eg http://www.newark.com/eaton-moeller/dilm12-10-24vdc/contactor-24vdc-12a-din-rail-panel/dp/24M1369

Opto's traic/mosfet driven circuts or just throw on a solid state relay.

Now u can easily drive a large contactor like that any number of ways. I won't go into snubbers etc. But you had a good idea in theory to parallel the contacts but in practice you may/should see a cascade/premature failure, etc. These are the sorts of questions that lead people to eventually learn how to build & design professional systems :) Good stuff, But i'ma keep it short. Have fun! Feel free to ask more to clarify and i can drill down for ya.

\$\endgroup\$
0
\$\begingroup\$

The answer to your main question is yes, there is a downside to using multiple relays to increase relay capacity, and it is - multiple relay failures, and a high probability of melted contacts and cables/wires.
As far as your second question, you can assume anything, but that does not mean that it is safe. The only safe thing you can assume, is that the relay bank will handle 2 amps (the design spec of each relay)!
If you want to be able to safely handle high currents, you have to get a single relay/contactor that can handle the required current. Sell your 32 relays and buy one contactor.

\$\endgroup\$
-3
\$\begingroup\$

I would opt for PLC's. That is where they shine. You can do that many and more with them and they are easily programmable. I would stay clear of the Fuse Box/Circuit Breakers however.

\$\endgroup\$
  • 3
    \$\begingroup\$ Where did your find a PLC with a 64A switching capacity - 0.5A is typical? \$\endgroup\$ – Dale M Jun 21 '17 at 2:16
  • \$\begingroup\$ Put larger relays in between , on each branch circuit, that can handle it , Sherlock. Your thermostat controls a huge compresser/condenser and it can do it with a tiny relay. Same thing , just more branches of control, with a PLC. And if you need a contactor at the end of each one , use a contactor. The main thing here, is that you have control. Industrial control. The guy who asked the question should just contract it out to someone who has the necessary skills, for his own safety. \$\endgroup\$ – Tim Spriggs Jun 27 '17 at 19:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.