The sensor you have probably uses Triac to turn the AC on and off(the same electronic device is used in typical dimmers).

Triacs create some sharp transitions in the current flow, and these can be damaging to any device not intended to deal with them (and anthing that uses coils is often vulnerable, like an AC motor).

The problem is that choosing a realy that can tolerate the sharp transitions is also an issue.
They will cause heating in the relay coil.

The ultimate solution is often to rectify and filter the bad AC waveform and then use a relay with a DC coil.

The problem is that NO transformers can be used to step down the AC before it is rectified, meaning you need either a relatively high voltage DC coil, or a way to safely reduce the rectified (and now DC) voltage.

You may be able to find some small AC-DC switching power supplies that can tolerate the waveform, fomr the triac.

Numerous under cabinet low voltage supplies exist that can operate from Triac dimmers.

Feeding one of them and then using their output to control a DC coil relay might be a good 'packaged' solution.

The problem is that some of them are poorly filtered and can also overheat a coil, so you need to actually observe their wave forms (both voltage and current under load) to determine what they look like.

You could always go the cut and try method, but make sure you observe the relay coil for a long enough time to make sure it is not overheating (think a lot of hours) and monitoring its temperature during the test and comparing it to how the temperature changes on a straight DC supply.

I would hunt for a motion sensor with contact closure outputs, then these issues have already been addressed in the design of the sensor.

A capacitor in series with the relay coil can add enough impedance to effectively reduce the power to said coil. The right value capacitor could be calculated but I hate algebra. Find the lowest-value capacitor that will still allow the relay to engage, and go about 1.5x that value.

I'm not 100% sure, but I suspect the same approach could work to drive the fan motor directly.

Now another approach.

Measure the DC resistance of the fan. AC impedance will be different but it will never be lower. Now use ohm's law and figure out the peak current it could possibly draw (about 180V peak, divided by the resistance you measured). Is this value less than 1/4 of the sensor's output current rating? If so, we're not worried about frying the sensor - only damaging the fan itself.

Fan thermally protected? Run it, check it every once in a while, I'm betting it's fine.

and now I'm on the phone, so I gotta cut this short.

"A capacitor in series with the relay coil can add enough impedance to effectively reduce the power to said coil. "

But it will do nothing to limit the heating caused by the higher frequency components of the waveform, and that is the problem.

if this were a dc fan, would it still be a problem?

But it will do nothing to limit the heating caused by the higher frequency components of the waveform, and that is the problem.

Perhaps not, but it WILL reduce the overall power supplied to the coil, and thus the overall amount of heat generated.

The idea is to reduce the power level as much as possible and still have the relay engage. We care about overall heat. Where there are multiple factors contributing to that heat, provided we still make the relay work, who really cares which factor we are reducing, as long as the overall amount of heat generated is brought to an acceptable level?

We could very well add a capacitor PARALLEL to the coil to shunt the high-frequency components, but the dangers of playing "guess-the-value" with a parallel component are far higher than those with one in series.

Gary,

For a fairly common example to illustrate the potential problem being described above, is if you ever had a light on a dimmer and when turned down low, the light bulb filament sings. This is caused by the triac in the dimmer switching on/off and causing a spike, square wave and high frequency harmonics.

A choke, or inductor, is commonly used in a filter circuit to eliminate this problem but to design one for an unknown type of load would require looking at the output on an oscilloscope and then doing some considerable work calculating and testing.

Anyhow, I hate to point out the obvious, but your first paragraph makes this all hypothetical: the PR150 requires, as you mentioned, a minimum load of 40 watts/VA to operate and your fan only draws 11 watts so, it ain't gonna work regardless.

However, after a quick look on their website, Leviton makes the ODS10 which is an output relay switched sensor that should work for you.

By the way, neat idea... but where are you venting this to anyhow?!! :)

yosemitebill,

Thanks for the input, I see where that unit has no min. load and is suitable for motors. I wasn't sure the significance of the min load spec for the PR 150, which was why I included that information. After all, I'm just the dumb homeowner. Which is why most of the above discussion is above my pay grade. Although I certainly appreciate the input.

While the ODS 10 is an obvious solution, I'd still like to know if it would make any difference if I were to use a DC motor. And also, would it make a difference if I put a light bulb in the load circuit?

Thanks,

Gary

"Perhaps not, but it WILL reduce the overall power supplied to the coil, and thus the overall amount of heat generated."
Not very effective at higher frequencies at all.
The capacitors impedance decreases with frequency, and the heating from the waveform discontinuities is NOT going to be decreased.
It is more than just total power, since the coil is already rated for the voltage it is intended to be used at.
"We could very well add a capacitor PARALLEL to the coil to shunt the high-frequency components, but the dangers of playing "guess-the-value" with a parallel component are far higher than those with one in series."
Not if you know what you are doing, and "guessing" is not part of knowing what you are doing.
The problem is that you will need a correctly rated film capacitor, not just any old capacitor.
"if I were to use a DC motor. And also, would it make a difference if I put a light bulb in the load circuit?"
Using DC through a Triac will normally not allow it to ever turn off.
To turn off a Triac you need to both remove the gate drive AND have the current flowing though pass though zero (there are also ways to force a Triac off by puling out gate current, but you do not want to even think about this mode of operation).
This happens 120 times a second for AC, but never for DC.
Once the Triac turned on it would not be able to turn off.

Finding a sensor with a contact closure is much safer and simpler.

I know what I'm doing, and I still avoid algebra as much as possible. I don't do electronics design on a large enough scale to make 'trial and error' inconvenient or inefficient, and I judge "unsafe" on a case-by-case (and relative) basis.

The whole idea was simply to insert an impedance in series with the coil. Any impedance. I know the heat from the HF components will still be there. If you run a relay from "proper" AC (a pure sinewave) you will still have some heat. Looking at the heat emitted from a "chopped" coil as normalheat+hfheat, they're added together to determine how hot the coil gets. Reducing either one will reduce the overall temperature.

Am I going to sit here and argue with you that a series capacitor is the BEST solution? Absolutely not. But it can work, and I've done it, and for the purpose of this thread, it seemed the easiest to suggest. I don't know about the OP, but I hate coil winding.

The discussion of using a DC motor involved putting a rectifier AFTER the triac. On the triac side, there is still zero-crossing, there's still a moment of no current, and it will still turn off. If we have a filter capacitor on the motor side, then the zero-current period is even LONGER since there will be no current flowing at all whenever the magnitude of the AC voltage is lower than the capacitor is already charged to. Of course, now we've got higher peak current and need to consider whether or not the triac's limitations are being pushed.
ÃÂ
ÃÂ
ÃÂ
ÃÂ
ÃÂ
When all is said and done, nobody will argue with you - buying an already-manufactured relay-output sensor is easier and safer.

What's your thought on my suggesting a 12V fan and PIR motion detector of the burglar alarm variety? (one of the crappy ones WITHOUT "pet immunity", of course)

There he goes again, disrespecting someone elses post over a small minute detail so he can spew forth a lot of technical data to come off as this oh so knowledgeable guru. Anyone else besides me notice how in just about every thread he participates, the thread turns from helping a poster to where he takes a tiny bit of information offered by someone else so he can quote a bunch of stuff that most times means nothing to the person with the problem.

Well excuse me!

If brickeyee feels disrespected I certainly hope he'll tell me so himself, as he's actually one of the most knowledgeable folks here, and I very much respect his intelligence and input.

I enjoy going back and forth with someone who knows what they are (and I am) talking about. Educates me more. Problem with that? I don't see one.

"There he goes again, disrespecting someone elses post over a small minute detail so he can spew forth a lot of technical data to come off as this oh so knowledgeable guru."

Just the little issue of starting afire by overheating a relay coil using a bad waveform.

But I'm sure the OP could care less if he burns his house down.

pharkus,

Brickeyee is the offender, not you. You are trying hard to help the OP, he is doing nothing much more than usual, taking a very insignificant statement and blowing it out of proportion to make you look insignificant. He does it often enough in just about every thread.

i've got your back brickeyee. i find your engineering expertise and perspectives refreshing. sure you talk in terms that most of the people on this forum don't understand, but you give correct answers, which is the point of this forum. and most of the time it is necessary to correctly explain the reasons why what other people are saying is wrong. most of the time there is really no way to "dumb down" the complicated nature of electricity, so unless you have a strong electrical background you wouldn't be able to comprehend the answer anyway. just because you can make something work, doesn't make it right. brick always sticks to the point and gives appropriate responses whether you are able to understand why or not. if you can't, then hopefully you are one of the ones asking questions, and not answering.

if there ever was an offender on this forum, it was a fella named bullheimer haha. you remember him brick? from about 4 or 5 years ago? i miss him he made me giggle.

... I am not offended by brickeyee - at least in any thread I can remember us both participating in - and it's far too late at night for me to scour the forum looking for threads where he might have said something that irritates me. Who does that anyway?

I do a lot more electronics than electrical. This thread is a case where I've done something on the electronics side of things and, while I still firmly believe it'll work just as well applied in the electrical realm, I certainly understand and respect his concern.

Most of what I do works. Sometimes it doesn't. If and when I'm wrong, it's just as important to me that I be corrected as it is to anyone I may be trying to help. Even when I'm not completely wrong, having to explain myself to someone who doubts me forces me to put things in different words, sometimes different perspectives, and helps me to better understand what I already know.

When I did this, the subject of this thread, I didn't even think about the fact that capacitive reactance decreases with frequency. I just hooked it up. It worked, and I never asked why. When I first read brickeyee's counter, I thought, "he's right. It shouldn't work. But it did. Why?" Now I've got a theory why, but I would never have thought about it had I not been questioned. That's how I learn things.

Of course, his point wasn't to further-educate me - it was to protect the original poster. I respect that, and I do not feel as though he was intentionally disrespecting me in doing so. It's not like he was calling me names. [He did raise issue with my "trial and error" component selection process, but I readily admit to being lazy when it comes to calculating things. I'll do the work if I have to, but if there's no danger that I'm going to blow myself up, I'll just try different things until I get it right.]

All in all, if the situation were reversed I'd probably do the same - ie, if another poster on here, even a regular such as brickeyee, was answering a question with information that looked potentially-dangerous to me, I would probably, at the minimum, ask for an explanation - and if I were confident that the information was completely wrong (or that there was a better way), I'd state so - and I hope nobody would feel personally targeted by my doing so. I don't.

An alarm system PIR makes a lot of sense, and most alarm co's make relay boards that go with them, and some may be mains-rated.

If you don't mind kludging something together, the relay in the PIR can probably switch a DC relay rated for mains on the contacts.

This whole triac thing makes a mockery of energy saving bulbs, I haven't seen many, if any dimmers, motion-sensing switches etc that can handle them. Kind of dumb and the regulating authorities ought to look at making it mandatory. I am sure there's already DC in most motion-sensing light switches so a relay is just a matter of cost I suppose.

Of course once you start modifying stuff the ASA/CSA approvals are void. I suppose home automation systems could do it, but it's kind of complicated....but cool.

The ones I used to install (can't remember the exact model but I was ordering them from MCM) had both NC and NO contacts.

I was simply proposing running a 12V computer fan through the NO contacts.

That's probably the simplest way, and computer fans are pretty quiet when properly mounted.