Relay

-- Randy M. Dumse

Caution: Objects in mirror are more confused than they appear.

Although it's a bit dated, I'd recommend "The Art of Electronics" by Horowitz and Hill.

Relays have a coil, which requires power, and contacts, which switch the output. Coils come in many voltages and current ratings. Most relays require more power to the coil than a signal from a computer, what's called a "TTL level" signal. So you usually need a "relay driver" IC to get enough power just to run the coil.

Many contact options are available, but the usual ones are "normally open" (N.O.), "normally closed" (N.C.), "single pole double throw" ("S.P.D.T"), which is like a 3-way lightswitch, and "double pole double throw" ("D.P.D.T."), which is like two 3-way lightswitches ganged together. If you're going to turn something on with a computer, use "normally open".

To switch AC power, you need a relay that's rated for it. A typical rating for something to control a light would be

5 to 20 amps, 120 to 240 volts AC.

Wiring on the AC side of the relay should be 18 gauge wire, bigger if the load is more than 5 amps. Get a relay with tabs for insulated crimp-on connectors on the AC side. That's what everybody uses today. You'll need a supply of insulated crimp-on connectors and a crimping tool; most big hardware stores have this. Provide strain relief on power cords, so if somebody yanks a cord, it doesn't pull wiring apart. The relay should be in the "hot" side of the AC line (which connects to the narrow pin on the power plug), not the "neutral" side (the wide pin on the power plug.) It is customary to use white wire for neutral, red or black for the "hot" wire, and green (always) for protective ground. If something has a metal case, it needs a 3-prong plug, and the case is connected to the green wire, which grounds the case when the plug is plugged in.

When working on this stuff, use a GFCI-protected outlet as your power source. This reduces the shock hazard, since the GFCI outlet will cut the power if you touch a hot wire.

Here's an example of a typical relay for switching an AC load given a 5 volt DC control signal.

Note the first relay in the table. It will switch 30 amps, and needs 900 milliwatts, or 0.9 watt, applied to the coil.

0.9 watts / 5 volts = 0.18 amp, or 180 mA. So you need a control signal with at least 180mA at 5 volts.

For practice, wire this up and use a 5 volt source (actually a 9V battery won't hurt the relay) to activate the relay. That's a good test.

If you look around on the web, you'll see lots of circuits for relay drivers suitable for connection to some computer output signal (the RTS line on a serial port being a classic example, because you can control it from software). In practice everybody uses a relay driver IC rather than discrite components.

John Nagle

Hi,

Maybe have a look at using solid state relays, they don't need much power to switch. Something like:-

Regards Ian Dobson

Home of the Atmel based UDP mobile web cam

mails scanned with av-filter.pl (F-Prot / perl)

"Its me Earnest T." schrieb im Newsbeitrag news:_D1fh.326$ snipped-for-privacy@tornado.southeast.rr.com...

Relays can be a lot of fun. And you can get started without even having one. Just look up "ladder logic" and start by learning just how they have been used to get full automation into the factory floor and sequencing the lights at your busy intersections, and all the controls in your furnace, washing machine, dryer, coffee maker, microwave etc. They are the true forerunners of computer automation.

Ladder logic is a superb primer for all electronics that drive motors, lights, valves, etc.

Wayne

Here's a good place to start with the really basic approach to what I was talking about earlier.

have you seen the Cubloc controller

aside from basic it allows Ladder programming.

regards, colin

Your question really hit me. Sorry to flood you with so much info... Just that I love relays!

I ordered a few timer relays some time ago for a project which quickly died and I'm sitting here with two that I am willing to give you on the condition that you respond from time to time with your experimental results so I can duplicate them with the ones I'm keeping.

They are Macromatic with specs you can get from their online site http://72.14.253.104/search?q=cache:3VANVNDztP4J:

they are standard 8 pin and clearly illustrated timing sequences with a variable knob and really, really neat!

If you have some alligator clips and a voltmeter and anything with a 24 volt supply you should be able to have a ball by one triggering the other, to trigger the first on x minutes of signal or not...

If you are interested, send me a private email ( snipped-for-privacy@att.net and I'll send you my mailing address to which you can send a self addressed and paid envelope from the post office and I'll slip them into the pack and they will be on the way.

Wayne

Thanks Colin. I went to their site to see what it's all about and really don't know what it is they are selling nor how they apply it to the real world. The words are there, but apply to anything from a single PLC type device to a Moon Rover. For example, the use of the word BASIC, to me that implies sequential programming and when they say ladder then I see conditional programming. But try as I might, I could not verify that opinion. \

But what the heck... this technology changes so fast that it's hard to keep up. When I started designing machines my basic control were cams... like the Swiss watchmaking machines. Now I'm into DNC.... maybe.

That was useful. Even though I've worked with digital logic for decades, I had never seen the term "ladder logic" before. I've even seen schematics in that format (like washing machines and pinball machines) but had never realized there was a term for it. It's always fun to learn something new....

snipped-for-privacy@kcwc.com You're my kind of person! So few of us around willing to admit that we are in a continuous learning mode if we are to survive in our profession.... even if retired. Wayne

Curt Welch wrote:

Heck, ladder-logic programming has been a mainstay of industrial automation control systems for decades, and is still going strong in tech-conservative industries. Even for newer systems that support more complex languages, Ladder Logic is still grandfathered in -- the IEC-1131 standard still supports it. It's looking a bit long in the tooth now, though -- a lot of brand-name LL setups support all sorts of exotic functions "buried" in LL function blocks, b/c the demands placed on industrial automation keep getting more and more complex, and LL has some real limitations, especially when you start getting into sequential logic and mathematical operations. Unfortunately, some industries have gotten into a trap where LL is the only programming language any of their personnel know, and they can't/won't invest in retraining. So their hardware vendors have to keep building larger and larger tower-of-Babel constructions of LL that are liable to collapse under their own weight... Whoops, sorry, drifted off onto a typical rant, there. :) Personally, I like IEC-1131 compliant systems -- you can mix&match multiple programming "languages" (LL, C-like ST, register-based bit-banging, and others) in the same project file, using whichever language you know best and/or works best for a particular function block. Hasn't exactly caught on like wildfire in the industries I work in, yet (the customers I deal with a *very* conservative), but it's pretty much inevitable.