17
\$\begingroup\$

I am developing a home automation project in which I'm using relays to control appliances. I need to control devices with 220V and 6A rating.

Should I use relays to control these appliances as a long term solution?

Relay I'm using is mechanical and is rated 220V 7A. If I keep it ON to control e.g. a fan, for more than a few hours on a daily basis, will the relay cause any problems? If yes then, what are other possible solutions?

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Check the datasheet, it should have a rating for the number of cycles, or coil hours; do you have the part number of the relay available? \$\endgroup\$
    – Matt Clark
    Commented May 2, 2015 at 18:48
  • \$\begingroup\$ For "long life" I would recommend 250v 12A. This way, you should not have any problems for a long time. In addition have a couple of these relays for future repairs and you will have a system that will be "up" 99.9% of the time. \$\endgroup\$
    – Guill
    Commented May 8, 2015 at 19:52

5 Answers 5

22
\$\begingroup\$

Relays tend to be quite reliable in benign environments, however they have a limited lifetime. Typically something like 50,000-100,000 operations at full rated load. At lighter loads, the life will increase, generally up to many millions of operations with a negligible load (the so-called mechanical life).

All this information will be clearly given in any decent datasheet. The markings on the relay are only limits for safety agencies and have little to do with the relay life.

Not all datasheets show the life vs. switched current, even for resistive loads, so you may have to test samples to determine that characteristic if you are say, using a 30A relay to switch 5A maximum. Inductive loads, incandescent lamps, and motor loads will also shorten the life.

Solid-state alternatives to relays have no easily defined wear-out mechanism, however they can easily die suddenly due to voltage surges, current surges (including momentary shorts) and from thermal cycling. They are also less resistant to heat, and tend to create a lot of it (a ballpark number is 1W per ampere of load current).

Most remotely switched outlets and similar consumer devices (where the consumer can plug anything into them) use relays. If the load is relatively light and well defined (perhaps a lamp) then solid state may be a superior solution.

\$\endgroup\$
4
  • 1
    \$\begingroup\$ what about this "Also, if i keep a relay ON to control (say) a fan for example, for more than a few hours on a daily basis. will a relay cause any problems?" ? \$\endgroup\$
    – narzan
    Commented May 2, 2015 at 19:22
  • 4
    \$\begingroup\$ No, the life of a mechanical relay is not generally affected much by being on or off, only by switching operations. In theory it might last only 100 years instead of 500 years because the coil is hotter, but that's likely not a matter of immediate concern. \$\endgroup\$ Commented May 2, 2015 at 19:23
  • 2
    \$\begingroup\$ Some solid state relays are also available with zero-crossing detection. Very nice if you want low disturbance during activation. \$\endgroup\$
    – Dejvid_no1
    Commented May 2, 2015 at 20:10
  • 1
    \$\begingroup\$ I had a similar question and was even looking for solid state switching options like using MOC304x series optotriacs. But then I found those suckers cost a fortune where I live :( Thanks a lot for your answer. This helped. \$\endgroup\$ Commented Apr 22, 2017 at 12:13
21
\$\begingroup\$

This is an old post but I am a controls engineer who programs industrial machines so I have 2 cents. I have a machine that does 20,000 cycles a day and I must use solid state relays even if they will fail in a short circuit situation and need to be replaced.

A general rule of thumb for me in my undergrad education many years ago is that if it turns on and off more than 100 times a day make it solid state. Typically mechanical relays are cheaper so I choose them by default.

You must also account for failure of the solid state relay when it has a short to add to that rule of thumb.

Also, someone said mechanical relays live to be 50,000 to 100,000... This is wrong I am pretty sure most of the datasheets I have seen with my mechanical relays are 500,000 to 1,000,000. The relays I use now are 2x10^7 in lifecycles so I typically say 1 million life cycles but it depends on what you buy. Mine are low end industrial relays.

\$\endgroup\$
3
  • 2
    \$\begingroup\$ Welcome on EE.SE! I downvoted your answer, because it doesn't really appear to add anything that wasn't already discussed in the existing answers. Your last paragraph would be better suited as a comment on the answer you are referring to. Not having the reputation to post comments is no excuse to add an answer IMHO. \$\endgroup\$
    – Mels
    Commented Jun 26, 2017 at 12:51
  • 8
    \$\begingroup\$ Voted up, I learned something. \$\endgroup\$
    – David K
    Commented Apr 25, 2018 at 20:05
  • \$\begingroup\$ For any casual reader of this answer, please, be aware of the key difference between mechanical (no load) vs electrical (some rated load) life of a relay. The mechanical life of a power relay is typically at least one order of magnitude higher than the electrical life. What mostly matters is the electrical life (which may be 50k to 100k, typical value for good relays, under the specified load). \$\endgroup\$ Commented Mar 3, 2022 at 14:35
10
\$\begingroup\$

My company uses relays in our HVAC products for several reasons.

1) They are reliable. Based on past and current experience, I expect them to last decades.

2) Relays tend to be significantly less expensive than triacs and their drivers.

3) They waste less energy as heat than solid-state devices such as triacs.

This is important for several reasons:

1) getting rid of excess heat is expensive.

2) energy wasted in the controls degrades the allowable 'green' energy rating that the unit has.

In general, I've had far more triac failures than relay failures.

We use quality relays from reputable manufacturers.

\$\endgroup\$
5
\$\begingroup\$

Normally the relay will be specified by number of contact operations. Apart from mechanical wear the contacts may become prematurely worn due to the nature of load (inductive, capacitive or resistive) that is being switched.

The mean time to failure (MTF) of most devices follows a 'bath tub' function - high failure rate at the beginning of life (due to faulty manufacture, poor assembly etc.), then a period of low failure and then a rising failure rate due to wear, heat fatigue and so forth. {see http://en.wikipedia.org/wiki/Bathtub_curve }

Failure of electronic devices tends to be sudden death (it works, then it doesn't). Relays tend to get a bit 'sticky' before failure (a temporary cure being a 'tap' in the right place.

Its one of those questions along the line of "How long is a piece of string?" so there isn't a precise answer yes or no. All I can say is that relays have been and are used quite successfully in equipment over a number of years.

\$\endgroup\$
5
  • \$\begingroup\$ @Agent_L That is your opinion. How can it be completely wrong (an absolute term) or do you simply disagree in specific cases? i.e the exceptions to this 'general' rule (which there are a few). My car repairs are predominantly mechanical not electrical and so fit this rule. The general rule applies to any kind of mechanical device (i.e. something with moving parts) not just a relay. I don't list (or argue) the specific advantages of solid state against relays (as your comment seems to imply) but point out that failure follows a well described (bathtub) path for any device. \$\endgroup\$ Commented May 3, 2015 at 16:52
  • \$\begingroup\$ @JImDearden how in the world do you think this is even subject to opinion? \$\endgroup\$
    – hobbs
    Commented May 3, 2015 at 18:48
  • \$\begingroup\$ @JimDearden: Mechanical devices are indeed more prone to wearing out... but you said failure in general, and Agent_L is correct that solid state devices also have failure modes, which depending on the application might lead to a much higher failure rate. \$\endgroup\$
    – Ben Voigt
    Commented May 3, 2015 at 18:58
  • \$\begingroup\$ @JImDearden We're talking about devices that can be built as either electromechanical or solid-state. Your car is not a valid example, all its subsystems are singular. As an example I suggest Amtrak overhauling their 1930's rotary converters simply because even today it's not economically feasible to replicate their transient fault resistance with solid-state. Engineer's role is to recognize best technology for given application, not to issue summary judgements. And for OP's application - can anyone guarantee users won't occasionally connect 2.5kW hairdryer to a 6A socket? \$\endgroup\$
    – Agent_L
    Commented May 4, 2015 at 10:58
  • \$\begingroup\$ @JImDearden I removed my comment and downvote. \$\endgroup\$
    – Agent_L
    Commented May 6, 2015 at 17:52
4
\$\begingroup\$

Advice overall is generally good.
Relays are 'hard to beat' if proper attention is given to ratings.
Note that resistive versus reactive (L or C) loads make a major difference and manufacturers specs must be carefully noted.
Note also (not applicable in this case) that DC is very demanding compared to AC. Manufacturers specify DC voltage ratings that are much lower than for AC.

Do take good note of Dwayne's comment re using quality parts from a known reputable manufacturer. In a serious application you MUST use a product of known quality. Unknown brands and equipment whose "provenance" is uncertain (ie may be fakes or out of spec parts) MUST NOT be used.

Specifics: A 7A rated relay at 6A load is probably OK, especially at low switching rates, but if possible I would use a higher current rated relay and/or look carefully at the specifications. When they say it is 7A rated, do they specify resistive or inductive load or other conditions?

\$\endgroup\$
4
  • \$\begingroup\$ it states 220v 7a for inductive loads.... what are differences in these two types anyways...sorry I have very little knowledge in electrical field. \$\endgroup\$ Commented May 6, 2015 at 6:27
  • 2
    \$\begingroup\$ @SurajBhawal resistive (eg heater) = easy to switch. Inductive (eg motor, transformer) or capacitive (computers) = difficult to switch because of arcing. Ratings for inductive are lower than for resistive. If it's 7A inductive loads than you don't have anything more to worry about. But yes, over-engineer to extend lifespan. \$\endgroup\$
    – Agent_L
    Commented May 6, 2015 at 17:50
  • \$\begingroup\$ Purely inductive loads can surge 10 times the rated current, for a short period of time, when switched ON. Conversely, abruptly stopping the current flow of an inductive load (during switch OFF), may generate flyback ("spike") voltages of up to hundreds or thousands (even for moderately low inductance and current values), which create damaging arcing of the relay contacts. The first effect requires you to select a relay rated (tested) for inductive loads, or else select a relay which a way higher rating than the steady current of your inductive load. The latter effect may require a snubber. \$\endgroup\$ Commented Mar 3, 2022 at 14:42
  • \$\begingroup\$ @jose.angel.jimenez An incandescent lamp often has a current surge as you state. An inductive load with DC switches with zero initial current. An AC inductive load can have high current if switched at zero crossing (!). A capacitive load can have high inrush current. | I agree re being rated for inductive loads - covered in answer. | I did not mention snubbing, as you say. \$\endgroup\$
    – Russell McMahon
    Commented Mar 4, 2022 at 2:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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