Spontaneous lamp flashing. Hunter fan limiter failure

[Back2Volts]

I have a Hunter fan with a four lights lamp kit.   Installed a few years ago.   Currently with three incandescent buklbs and one compact fluorescent.   No dimming.   Minutes ago, as it was getting darker, I turned it on.    To my surprise, it took a split second to go on, which was abnormal.  Then after another split second it went off, and then on, off... with a steady flashing pattern.   A bit faster than once per second.   It is not flickering.

Wondering about the fluorescent bulb, I removed it, making no difference.   My wife removed the rest of the bulbs, swapping the last one, with no difference.

I believe the same circuit has two smoke alarms in it, but I do not see how they can have any part in the magic.   

Any ideas as to what the cause would be?     Am I in an encounter of the third kind?    Can you point me to a paranormal detector circuit?   I'd hate to have to call Ghost Busters!

[Ian.M]

Bad contact + thermal expansion, probably of the copper wire, due to the contact resistance.  When it gets hot enough its moved enough to make full contact till its cooled some.

Its currently a fire hazard so do *NOT* use again until you've dismantled it and checked for overheated wires and loose connections.

[Back2Volts]

I have it off.    I thought of the fire hazard.   I was thinking of some thermal effect, but I was thinking that it would be more like current goes through and goes on first, then warms up and disconnects.  It cools off and reconnects.    The opposite threw me off.

What you are saying is that there is a poor contact, not enough to bring the bulbs on, but enough to provide heat for the cycle.

I will look at the connections on the hanging box.   I was wondering about the breaker as a possibility.    I will recycle it first. 

[Ian.M]

Its unlikely to be the breaker, but probably not impossible.   It wouldn't hurt to check it with a known good bulbholder and bulb directly connected to it, and as access to the panel is a lot easier than taking the fan apart, its a good place to start.  However, the fan/lamp wiring has been vibrating due to any slight imbalance in the fan itself, probably for thousands of operating hours if you've been using it much, so is a prime candidate for bad connections.  You'll have to take it down to work on it, so at that point patch in a bulb + holder to its ceiling junction box to eliminate any faults in the circuit up to that point.  It also gives you light in the room so you can take your time to check the fan/lamp assembly thoroughly. 

[Monkeh]

Is it by any chance a remote controlled fan and light?

[Tomorokoshi]

Ceiling fans are usually installed sometime after the house was built. In either case, the wiring in the fans is usually somewhat marginal, and the wire nuts supplied with the installation kit are usually on the cheap side.

Check the wire nut connections. Some fan designs have the lamp assembly as a separate unit that screws in underneath the fan motor. There is a space there for the wire nut connections. There is a separate set of wire nut connections in the ceiling box the fan mounts to. This is trickier to service, but the wires can usually be pulled out from the side depending on the support structure without taking the fan off the ceiling.

If the nuts of not sized and used properly the relatively small fan or lamp wire can slip out, particularly of the solid copper 14 or 12 AWG service wiring was not twisted together before the wire nut was put on.

As for Monkeh's point, I've had to replace the remote control receiver assemblies a couple times. I don't recall how they behaved when they failed.

[MK14]

This video explains why your specific fan unit flashes like that, and how to fix it.

WARNING:
I'm NOT sure how safe or if it is a good idea, to perform the repair as shown in the video.
Best to get professional/electricians advice or leave it alone.
Video only to help you understand what the problem is (i.e. why it is flashing).

[Monkeh]

Oh, yeah, that'd do it.. I'm not sure it's strictly legal to modify it by removing that, but I'd have had it out by now..

You can, apparently, obtain replacements: http://www.ceilingfanremotecontrols.com/fan_limiter_.html

Repair is, I imagine, fairly trivial also.

[Back2Volts]

Interesting!.    I will try to find out what limiter it has and look for a replacement.

Thank you all!    I'll post the outcome.

[Ian.M]

From the cycling behaviour, the limiter could simply be a self-resetting thermal over-current CB.  If so, if you can get at its contacts and clean them with 600 grit or finer wet&dry paper followed by a strip of thin card wetted with contact cleaner to get any abrasive dust out, it will probably cure it as the usual cyclic failure mode for these is excessive self-heating due to contact burn raising the contact resistance.

I'd also be reluctant to remove it without providing appropriate circuit protection, but alternatively an inline fuse holder with a ceramic slow-blow fuse of an appropriate rating could be used to protect the wiring against idiots overloading the fitting.

[amyk]

http://diy.stackexchange.com/questions/1511/what-is-the-purpose-of-a-watt-regulator-in-my-ceiling-fan

Regulatory bullshit... along with a touch of planned obolescence

[CatalinaWOW]

There is no real safety issue in bypassing the regulator.  Might or might not be a legal issue in your location, but generally no.  All you have to do to keep safe is not put too many watts of lighting in the sockets.  Should be a trivial problem now that LED lights are available.

I have found the replacement regulators hard to find and extremely expensive.  Nearly high enough to justify buying a new low end ceiling fan to salvage the part.  Which is why I have bypassed them in all of the cases I have encountered in being the family handy man.  Three failures out of less than a dozen installations in less than a decade.  Totally unacceptable reliability.

[Monkeh]

I suspect they're very easily repaired. There won't be much to them.

[CatalinaWOW]

I thought I should update my comment from above.  The comment reflected my experience about 2011.  Did some quick googling and found that replacement modules are now widely available on the internet (though I still haven't seen them at local hardware stores and the like).  Prices have also come down.  What I found then was going for over $65 as I remember.  Now I found replacement modules available for well under $20.

The hidden message here is that the failure rate has been high enough to quickly create a market for replacements, and that competition in this market has brought prices down.  Not a great message.

I remain unconvinced that these modules provide a safety benefit.  You can easily find reports on line of these modules catastrophically failing, smoke, flames and everything.  But it may be wise to replace them for legal reasons.

[IanB]

Replacing them with a wire nut seems like a good idea. But 4 x 40 W bulbs is 160 W -- surely that is enough light for anyone?

In most of my multi-lamp fittings I have replaced the 40 W bulbs with 25 W bulbs. Still plenty bright enough.

[Monkeh]

They are not for safety reasons - they're designed to limit the power to 190W maximum to comply with energy saving legislation. They'll just be a current-controlled dimmer. Most likely just a cap gone bad inside causing this behaviour.

As I said, I'd just throw the thing away, but that seems to be illegal (modification of a listed fitting).

[Back2Volts]

I have removed the lower fan can very easily: three screws and undo a 3x3 wire connector.    The can is really packed!   It contains the rather large limiter, one 3.5uF capacitor, one dual 5uF+5uF capacitor, three switches (lamp, fan, reverse), five wire nuts and a zillion wires.

Peeking inside the limiter through vents it seems to be mostly through hole components.   The exception is a 5x5mm DPAK? with markings T4 1060 7PAAT843, tagged as SCR1 in the PBC.   One MOV disk, couple of resistors, couple of yellow case caps and couple of small electrolytic caps.     

I assume the SCR1 is probably a TRIAC, and since it goes on and off, it is probably ok. 

[Monkeh]

T410-600

[Ian.M]

Odds are replacing the small electrolytics will fix it.   However I wouldn't hesitate to rip it out and fit a fuse - which still meets the requirements of the legislation cited.

[Back2Volts]

Odds are replacing the small electrolytics will fix it.   However I wouldn't hesitate to rip it out and fit a fuse - which still meets the requirements of the legislation cited.

I will detach it and look at the insides tomorrow.   If I have comparable caps I will replace them.   

Attached a couple of pics

[Back2Volts]

I opened the thing.   It does have quite more SMD components under the yellow caps.   I think the electrolytic cap is shot.   Must be very poor quality.   470uF 6.3V, 105.   Brand LM, series GR.   Anybody know what LM is ?

I know I have good 470uF/50V on hand but are too big.   I may have to order some small 10V ones. 

[Monkeh]

Mmmm, electrolytes. What plants crave.

Bet a 125C cap in that would last a few decades.

A wire would last a few centuries.

[Ian.M]

So patch in your 50V cap and run it open on the bench with a 100W bulb as the load.  If it works you know you are on the right track and can order the right part.  Make sure you get a minimum of a 105 deg C rated cap.

If it doesn't and you cant find any open resistors or shorted diodes etc. then its time to look at scrapping the limiter (maybe trace as much of the circuit as you can and send it to Dave for a teardown?) and patching in an inline fuseholder, which should be a high temperature rated one (usually black or brown phenolic) not a cheap nylon one, with a ceramic body T3A or T3.5A fuse. 

[Bryan]

Watching some stupid show on the History channel. Ozzy Osborne and his son on some road trip. Ozzy was complaining about his motel room and the lights in a fan flickering off and on. Bet it was the exact same thing.

[Back2Volts]

Bet a 125C cap in that would last a few decades.

Unfortunately is not that easy to find a 470uF/6V3/125C in 6.3mm diameter.    The best I have found is 6000hrs at 105C.   Still, probably much better than the LM it would replace.