[SOLVED] Triac's gate resistor blew up

Hi all,
I've built a little power controller, intended to control the speed of a 120VAC fan. It's a pretty standard setup; I'm using an H11AA1 for zero crossing detection and a MOC3023 isolator to trigger the triac. The triac is a BCR5LM-12LB#B00.

So it all works... almost. For testing I have an incandescent lightbulb hooked up to it instead of a fan, and once I got it all working I just made it ramp up and down the output power forever. It's run for maybe 20 hours or so without skipping a beat (that I've noticed), but today I walked away from it, and when I was walking back towards it I noticed the light was off. So, I scooted over and started looking at things, and saw that my status LED was still blinking away happily, suggesting it was still sending the triac gate signals to the MOC3023. Then I tried gently shaking the lightbulb to see if I could hear the filament rattling around, indicating it was burnt out. At this time I noticed that one of the 1/2W resistors started to blacken and smoke, so I immediately unplugged it. I looked around and didn't see any other damage. I poked around for a bit but couldn't find any issues. The resistor is pretty well cooked, it was nominally 1K but now it's 150 ohms. I decided to go for broke and plugged it in again to see what would happen and... the light just keeps happily ramping up and down as it did before.

The resistor that cooked is between the triac's gate and MT1 terminal.

When I review the schematic I see that for some reason I didn't match the recommended application exactly; in particular I see that I put the load on the 'wrong'(?) side of the triac. I'm wondering if this might have something to do with it.

So, I'm wondering: Why did that resistor blow up? Why was it dissipating so much power that it cooked? I'm having trouble imagining a situation where the current through it would get high enough to cause trouble, short of the triac failing in some fashion. But since when I plug it in it still works I'm assuming the triac is still OK.

I've attached the recommended circuit from the MOC3023 datasheet, as well as my circuit for reference. It was R12 on my schematic that blew up (and the resistor on pin 6 in the MOC schematic).

Not sure what else to do to try and nail down the problem; I can't get it to repeat. I'm also more of a digital guy so I'm hoping someone with more experience in this domain might find something obvious that I'm doing wrong...

Thanks!

Can only happen with missing gate connection (or - less likely - internally disconnected gate terminal). Probably a problem of your experimental setup.

Oh my god, you are amazing.

If someone was suitably masochistic and detail oriented they may have seen that the triac leads in my schematic were mixed up - they are actually mixed up on the PCB too. The result is that the gate was being biased to MT2, so it was always on. I realized this a while ago and I had desoldered the triac, turned it around, cut the gate pin, and soldered a jumper to fix it.

Just now I powered on the unit and 'very roughly' wiggled that jumper wire. I was able to get the light to flicker and was eventually rewarded with the (usually dreaded) smell of burning electronics. I think you nailed it. I just put on a pot of coffee for nothing!

Thank you so much!