RAMPS 1.4 Mosfet issue

Been runing a RAMPS 1.4 board for about a year but recently run in to trouble, The mosfet controlling the heated bed suddenly started drawing heaps of power enough that it was starting to smoke the wires on the power supply odly though it sort of still worked it would only draw this power when it was turned on in software and the bed still heated up. So I replaced the mosfet and same thing, Lifted the drain pin and checked the current being drawn by the bed about 3 amps, hmmm. Re soldered the the drain and lifted the source to check the current there, turned it on and before I could even check I noticed the mosfet was insanley hot so cleary the issue is one between gate and drain, I would assume this is just a dead mosfet but I have replaced it with 3 known good fets and same issue. I am really lost on this one, any ideas would be really helpfull...

Did you look for shorts on the RAMPS board?

Yes I have looked and looked but there is nothing I can see, as I said it was working but just stoped.

If the bed draws more current than it should it must be a problem with the bed. The MOSFET's job is to connect the full 12V to it and the current drawn is then down to Ohm's law.

Putting a meter in the source connection is a bad idea, it probably made the MOSFET oscillate at RF frequency or at the very least stopped it turning on fully.

Edited 1 time(s). Last edit at 12/03/2012 10:45AM by nophead.

---

[www.hydraraptor.blogspot.com]

RAMPS has 2 x PTC polyfuses for over-current protection. One is 5A (Motors, Nozzle heater and Fan) and one is 11A (Heated bed).

So far, all experiences I've had with PTC polyfuses have proved that they're unreliable. Especially the 11A one for the heated bed.

Many heated beds will draw very close to that current limit (someone reported 12A on their bed). This is just over the 11A that the PTC polyfuse supposedly will give allow you to draw.

Internally the polyfuse works on heat (much like a normal fuse). The current passes through the polyfuse, and heats up. Once it internally reaches the specified temp, the circuit opens. HOWEVER, as soon as it cools, the current is restored. It's quite possible that the inherent hysteresis of this temp will cause the output to oscillate fast, which in itself could cause the MOSFET to heat up unduly.

If you don't decide to replace the polyfuses with wire links (and then put proper in-line fuses on the power leads to your RAMPS board - NEVER run without some sort of fuse), then make sure of the following things:
1. Make sure that the two polyfuses aren't touching. The 5A and 11A polyfuses trip at different temps. If they're touching, you can end up with one of them tripping before the current reaches the required amount, simply due to temperature transfer. It's usually the 5A one that trips, but that in itself can lead to confusing results testing other things.
2. Consequently, make sure they're also sufficiently cooled (with a fan), as they will then not overheat unduly. I've seen setups that print fine in normal temps, but get a hot day and the things just don't want to work without a fan (or in some cases, at all).

Also things to note/check/look at:
A. Some firmware used to run high frequency switching on the heated bed. This would lead to the FET running much hotter than it should. It's also not required. Using a PWM frequency of 78 Hz seems to be a fairly good choice. Search the wiki and the forums for more info on this, and how to change it in your particular choice of firmware.
B. The STP55NF06L's FET's aren't exactly much chop. I personally prefer replacing them with the FDP8870, or the IRLB8743's (at least for the heated bed output). They both have very low RDS(on) values, which helps cut down on the FET getting too hot.
C: If you have access to one, put an oscilloscope on the Arduino MEGA's output for the heated bed (D8). Also check that R4 hasn't somehow come off the board, as this could lead to the FET input not shutting down properly when the output is switched off, which could also be your issue.

Good luck!

Thanks for all your replys, I checked the supply and with the bed on it was drawing 5 amps which is ok. I ended up getting some cheap mosfets rated for 70 amps put them in with a big heatsink runs fine now, very cool. Curious why the new fets I tried did not work they were the original part number but just got too hot. Thanks again for the help.

The current rating of the MOSFET doesn't determine how hot it gets. People mistakenly believe that a MOSFET with a higher current rating will run cooler.

The current rating should be higher than the load of course, but the important parameter is RDSon at VGS = 4.5V. That determines the amount of heat generated. Modern MOSFETs are available with 3 milliohms on resistance. They don't need a heatsink and only get slightly warm at 10A.

---

[www.hydraraptor.blogspot.com]