A little OT: Strange mains input circuit burning.

[Ethan]

I have a novelty bill collector/coin changer in my basement bar. Yesterday it stopped taking bills so I opened it up and saw that the current limiting resistor in series with the mains input burned up. At first I thought maybe the resistor was just underspec'ed (at 2W) and gradually burned up over time, so I replaced it with a 4 W--Still burned up almost immediately. Then I thought maybe the diode across the mains was shorted or leaky, so I replaced that, still the 220R burns up. My only other (layman) thought as to why that resistor would burn up is if the LED in the opto were shorted. NOTHING else in that area shows signs of burning but that resistor. Are there any other conditions that would make that resistor burn up (I measured the mains voltage at 118VAC)? I've never actually seen a regulator like this that uses an optocoupler to regulate mains input...very strange...

http://www.fairchildsemi.com/ds/4N/4N26.pdf

 

[Gus]

220 ohms?

120V/220 ~=.5amps

voltage x current =power 120 x .5 = 60watts?



I think the resistor might be more like a 22k for around 5ma

220k would be about .5ma

The circuit looks like it is used to make a 60hz pulse, it works on 1/2 of the input waveform turning on the photo transistor at 60Hz

 

[Ethan]

Yeah I know!
That's part of what confuses me...because obviously it worked for many years...
I got the resistor value from the fried resistor color code: RED-RED-BROWN-Gold.

Thanks Gus.

 

[Ethan]

Wait just a second...
I guess it could have been RED-RED-ORANGE-GOLD
Which would make it 22K!

Note to self: when a resistor fries color bands may change a little.

 

[Mark Burnley]

Yep,

That's got me before now

Ohms Law saves the day again!

:thumb:

Mark

 

[bcarso]

Still odd that a 22k 2W fried. The LED and reverse diode drop is negligible---there might as well be a short across them from the perspective of the resistor dissipation, for military work.

 

[Ethan]

I think the LED in the opto got fried too when I put in the 220R. The 22k resistor now doesn't burn up, but its still not working. I will change out the opto and see what that gets me. thanks guys!

 

[Svart]

what's that other diode? Is it a Zener?

LEDs rarely short closed. their junctions will totally burn out with a quickness with any overcurrent/voltage. I think your opto died when whatever it was burned that resistor, not after you changed it.

I think that this was a circuit to monitor if AC was applied? If i were you i would check the other half of this circuit for damage too.

 

[bcarso]

The other diode would be a conventional rectifier one, just to protect the LED from excessive dissipation in reverse breakdown.

Certainly the 220 ohm wold blow most any LED open as you say, leaving the other diode to get quite warm perhaps due to forward voltage and current. OTOH, if it is a relatively low breakdown part, normally "protected" by the LED's forward drop, then once the LED opened it could fail shorted after overheating due to the additional dissipation in breakdown.

But the mystery remains as to what caused the failure in the first place---a 22k R should take a licking and keep on ticking. If it was defective to begin with with some localized hot spot I guess it could go a while before catastrophic failure. Or maybe some horrendous line spike caused an arc-over and localized damage, followed by eventual failure.

Clearly the 220 ohm took out the optocoupler though, at least.

 

[Mark Burnley]

Ethan,

Not at all OT- an electronics puzzle is an electronics puzzle, no matter what the app :grin:

Did this happen recently? In another thread it was mentioned that the mains supply in some parts of the US has risen to above 160V. Normally, discounting the LED drop which is minimal compared to 110V, the forward current is approx 5mA. The power dissipation is 0.55W. So a 2W resistor is well specced for an always-on mains powered app.

However, if the mains supply rises to 160V, the current becomes 7mA, and the power dissipation becomes 1.16W (P= I^2 * R, so the power is increased in a square proportion...) This could have caused the R failure, especially if the resistor is located in an enclosed case, in a warm room etc etc.

Just replace the opto, and use a 3W 22k to be on the safe side!

(Haven't forgotten about the A80 BTW!)

Mark

 

[Ethan]

For some reason I periodically forget that I know Ohms Law . The resistor that burned originally was cooked a light brown rather than scorched black. So it looks like it was a gradual, slow sizzle. I suppose it could've been a 1W resistor (I've seen some resistors that looked like 1W actually be 2 W and 1/2W actually be 1W, etc...) But, still 1W would be overspecced for a little more than half watt of heat. Nothing else on the board *looks* like it was affected. The way the 220 ohm resistor turned charcoal black, I would've at least expected the opto package (plastic) to melt, but it looks fine. I just ordered replacement optos, so hopefully that gets it back to normal next week.

I have a mains voltage meter that's always plugged into the wall, and I've never seen it go above 130V. (insert that mystery tune here).

(Thanks Mark)