Marlin/RAMPS heater bed PID vs. bangbang settings

I have a reoccurring problem with my ramps board overheating when driving a heater bed.

I have a 12V 30A power supply and the resistance across my heater bed and leads is about 4.8 ohms. I buy the ramps boards off ebay: [www.ebay.com]

The problem Im having is that I keep burning up the MOSFET on the ramps board that supplies the heater bed. I see in the configuration.h file that there is a way to limit either the duty cycle to the bed and or the total current aloud to be sent to the heater bed (1-255 fashion).

So does anyone have this happening as well? I have tried restricting the current and raising the duty cycle, but then the bed wont get up to 100C or will take forever to get there. Ive tried working my way up to values that reach the appropriate temp without burning up the mosfet, but this is a pretty frustratingly slow process of trial and error. I have done the auto tune for heater bed PID constants but I believe these also depend on the duty cycle and max current constants chosen, which I haven't decided on.

Any preference to PID vs. Bang BanG?? PID seems smarter to me, but I suppose they could be essentially the same if your messing around with the duty cycle and max current constants.

also, shouldn't the fuse have tripped before the mosfet gets so hot it melts everything around it??

as you might have guessed, this is getting frustrating and expensive,
any insight would be appreciated.

thanks and merry christmas!

The ptc fuses on a RAMPS works like a very slo-blo fuse.

One fix is to use a solid state relay for the bed. I use them on my printers, one printer has an ac heater and the other operates a qu-bd 12v heater.

so...it is common to burn up your mosfet/board trying to power your heat bed?

It's interesting, on my Sanguinololu I'm having the heatbed wired up with a relay, it's fast, it's painless. However, I have never tried it without a relay.

Today, I want to start working on another printer with a Ramps 1.4 board, however, it's not using a relay, instead the heatbed is connected directly to D8.

In theory it should be no problem right? I mean, without a relay? I don't know that many people that use a relay, and I never heard about the MOSFET burning on their Ramps boards!! My MOSFET also has a cooling plate on it and the Ramps is cooled by 2x 50mm fans on top.

The Ramps board is a G3D one and the cooling plate is also theirs.

Are you using active cooling Archy?

In theory, yea, that D8 power supply should be able to carry as much current as the bed will pull, the re-settable fuse should blow before the mosfet ever does, and if your blowing the fuse, you should be able to tweak your firmware settings to accommodate.

I likely have the same cooling plate as you and I have two 30mm fans over my ramps board. They don't move a whole lot of air, but it's better than nothing...

I've probably burnt up 3 ramps boards with this setup, so I just switched to a rambo board, but I'm pretty sure I'll have the same problem there. Maybe its my heater bed.

Care to share a part number of your solid state relay?...its definitely the safer way to go, but if a lot of people are getting away without a relay, I'd like to figure out what I'm doing to have this problem...

Here, this is what I am having:
[gadgets3d.eu]

The relay part number, I can't see unfortunately, it would mean that I have to disassemble part of my machine.

However, I can ask someone to send me another one maybe. The first one I also got from him and it works perfectly!

Something I see all the time is that the soldering on cheap RAMPS boards is pretty poor, and around D8 (the heated bed) this can cause REAL issues. Things like the FET not having enough connection to the pads on the bottom and on the TOP of the board. It doesn't need to be excessive, but you really shouldn't be able to see a gap between the top pad and the pin of the FET.

As for the PTC fuse, the 11A PTC fuse can handle a huge amount of current in short bursts, which means if you accidentally short something, it DOES NOT PROTECT YOU. For example, the 11A fuse will handle 40A for anything up to 20 seconds before it will trip (the trip time for just greater than 11A is 22 seconds). If you have a short, that's easily enough to melt most of the tracks on the bottom of the RAMPS board before you get anywhere near the 20 second mark, depending entirely on how much current your PSU can provide and the nature of the short.

This is why, even if you keep the PTC fuses on your boards, I always recommend using a fuse in-line with your PSU as well. PTC fuses are not a replacement for a standard fuse.