Heatbed burned up. What happened?

Building my first printer, and ran into this issue. The board is a Sangrinololu and the firmware is marlin, heatbed is MK2a. Using pronterface, I tried the heatbed for the first time. Issuing a barely even warm command of 30 degrees C. Immediately I heard popping and this track burned up before I could do anything about it.

The bed seems to still be at least partially functioning, the LED comes on still. And when I tried setting it to 30 degrees a second time (which is 86 degrees F), and then measured it with a noncontact thermometer, it read 84 degrees F, so that seems to indicate the thermistor is working right.

Making a newbie guess, I feel like possibly the board may be getting too much current too fast, since when the heat kicks on, I really hear a strong whine (even though I'm only raising the temperature by a tiny amount, from about 70F room temperature to 86F. Also in the pronterface monitor, even though I just issue it a command of 30 degrees, in the monitor graph, it shoots up almost immediately to 50 degrees, before slowly settling down to 30. Unsure if this is normal or not, or maybe if it has to do with the fact I'm taking little baby steps using much lower than typical temperatures.

I see there are settings for heatbed duty cycle, as well as stuff like bang_max (though that seems to apply to the hotend). Should I be using the duty cycle? And if so, what sort of value should I use. Could any other firmware settings cause this? Or was I perhaps just unlucky and got a bad MK2a?

And is my heatbed pretty much toast now? It seems to be working, but I'm guessing its not heating evenly (I don't have a thermal camera to test), and I've been too afraid to turn it any higher than a measly 30C.

what powersupply are you using ?
and where did you get the heated bed?

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

Power supply is a corsair 500W ATX. Heated bed was on ebay, bought it quite awhile ago, around 6 months ago, so I don't remember the seller.

i could be that the copper is too thin on that particular pcb or it's been damaged at some point

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

Quote
scubaguy
Issuing a barely even warm command of 30 degrees C. Immediately I heard popping and this track burned up before I could do anything about it.

The rise in temperature being called for does not proportionally increase the current being applied to the heated bed. In other words, your bed will draw the exact same current at the same voltage whether you tell it to rise 1 degree over ambient, or 100 degrees over ambient. The different is the amount of time that current is on. The 1 degree rise takes a lot less time to reach the desired temperature than the 100 degree rise.

The picture you posted shows heating on what appears to be opposite ends of adjacent traces. My guess is you had a defective circuit board with a short between traces. This dropped the resistance from the normal ~1 ohm to essentially none. There likely was nothing that you could do to prevent it aside from measuring the resistance prior to energizing the bed. It wasn't anything that you did wrong. Measure it now to see what it reads. My guess is that it won't read within spec...and even if it did, I wouldn't trust the board.

I appreciate the help. And good to know as user cdru mentioned, full current is applied, no matter what temperature rise is applied for.

I think I figured out though why the temperature would jump so high so fast, basically the vast majority of the heat was going through that one area, right next to the thermistor, so it makes sense that area would overshoot temperature almost immediately, while the rest of the board saw almost no heat. I think there may have indeed been a slight short between those 2 traces, can't tell for sure because its all burned up, but theres one area where the space between traces seems a little thinner.

I've got a new board on order now, from a seemingly good seller, who specifically states the boards are quality, and measure roughly 2 ohms resistance. So I'll give it a try. Thinking i may eneble the duty cycle firmware function to begin with to be safe.

Quote
scubaguy
measure roughly 2 ohms resistance.

That's a little high. 12 volts / 2 ohms = 6 amps. The target usually is 10-11 amps which is right around 1.2 ohms. I guess it's better to be too high of resistance then be too low, but you may have trouble reaching higher temperatures or longer warm up times.

Yeah, I figured 2 ohms might be high, when I heard somewhere 1.2 was normal. I figured after burning up a board I'd err on the side of caution, especially since everywhere I look, nobody ever seems to list the resistance. At least this seller had a spec for resistance. I remember the wiki also says get etched rather than plated, but searching for etched ones I don't really find anything, they all seem to be plated.

I think that warning is overstated. Entirely clad is desired over plated, but a quality plated board I think is more than usable.