Hot bed heated up than cooling down

Hi,
My setup:
Kossel mini, Arduino mega 2560, Ramps 1.4, ATX12V P4 PSU, heatbed resistor reads 1.6 [Ohm], latest firmware [MOTHERBOARD is set to 33]

For testing I'm using pronterface - all works.
Every time I turn on the heat-bed for 110[C] it reaches this temp (I feel the heat and see it on the temp' graph), but than temp drops down.
It doesn't stay 110[C] temp & immediately after arriving to that temp it drops down the same speed it heated up..
Even after cooling if I re-set the temp to 110[C] it ignores the command & temp continue to drop.

Any ideas what is wrong?

Edited 1 time(s). Last edit at 08/30/2014 04:22AM by polkary.

Have you got a heatsink on your D8 mosfet chip and do you have a fan blowing across your Ramps board?

---

_______________________________________
Waitaki 3D Printer

Thank you for your replay.
No(for both of them), I wasn't aware that it is needed.
Any recommendations about fan size & sync type?
(is same sync as on the motors is ok?)
Edit:
I've checked and I don't know which will be better, should I go with a 120mm 1800 RPM fan or 80mm 2000 RPM?
In kossel mini usually the ramps are below the heated bed so more heat is added up while heating the bed.
I'm considering to add a bigger fan to cool down the ramps & printed area, or simply pulling out the ramps from below the bed ...

Edited 1 time(s). Last edit at 08/31/2014 05:56AM by polkary.

What is the part number of your MOSFET? The built-in ones might be too crappy and generate too much heat.

You can either add heatsink and add cooling to it, or you can change to a MOSFET which generate less heat.

thank you for your repaly.
My MOSFET number is P55NF06L is it one of those generating heat?
Thanks!

A 40mm or 50mm fan should be sufficient - it cools the fuses and the mosfet so should be located near them. You must get a heatsink on your mosfet - they are very cheap.

---

_______________________________________
Waitaki 3D Printer

Thank for the replays!
I've bought a 80[mm] 2000 RPM just to be in the clear zone
But heatsink will arrive only next week [nothing near me :\]
Crossing figures..

Seriously, the P55NF06L is not going to work nicely for the heated bed. Mine had gone up to 180degC and above without a heatsink and fan. I don't think it can be cooled down a lot even with a huge heatsink.

I just switched to an AOT240L (some guys on #reprap suggested that) which only go up to 50degC even without a heatsink and no fan cooling.

Seriously, why do you say that? There are thousands of printers out there using that chip, including mine - my printers have been running nearly 2 years with no Mosfet problems and I only have a small heatsink and a 40mm fan on mine.

---

_______________________________________
Waitaki 3D Printer

Quote
waitaki
Seriously, why do you say that? There are thousands of printers out there using that chip, including mine - my printers have been running nearly 2 years with no Mosfet problems and I only have a small heatsink and a 40mm fan on mine.

Yeah it may be fine with heatsink and fan, but still it is better if the MOSFET produce less heat. With a heated bed running at 10A the P55NF06L really dissipate a lot of heat.

BTW, have you measured how hot your MOSFET gets to without cooling?

Thank you guys, I've tested outside + fan + heat sink.
still having issues with slow heating (more than 10 minutes to arrive to 95 [c] while temp goes up than down than up again)
since the heat was sounded like a real issue I've tested that & with the fan [2000 RPM] I don't think that the MOSFET over-heating and I would expect to feel the heat when I'm near it but the fan does it's job & I fell the heat ONLY from the heated bed.
I would prefer to make sure that my setup is OK before switching to another MOSFET.

I'm not sure that my 5 years old PSU which suppose to deliver 450W (and up to more than 16 [A] according to it's spec) is giving the 10-15 [A].
How can I test that? my DVM is up to 10[A] is there a way to test that without new DVM?

Edited 2 time(s). Last edit at 09/03/2014 10:56PM by polkary.

Quote
polkary
Thank you guys, I've tested outside + fan + heat sink.
still having issues with slow heating (more than 10 minutes to arrive to 95 [c] while temp goes up than down than up again)
since the heat was sounded like a real issue I've tested that & with the fan [2000 RPM] I don't think that the MOSFET over-heating and I would expect to feel the heat when I'm near it but the fan does it's job & I fell the heat ONLY from the heated bed.
I would prefer to make sure that my setup is OK before switching to another MOSFET.

I'm not sure that my 5 years old PSU which suppose to deliver 450W (and up to more than 16 [A] according to it's spec) is giving the 10-15 [A].
How can I test that? my DVM is up to 10[A] is there a way to test that without new DVM?

It may be the same issue I am experiencing. I want my heated bed to go up to 120C but it will only reach 116.7C and then it starts to fluctuate between 114~117. It looks to me like an "Ohm's Law" issue. The bed resistance, 12V and whatever amperage (should be around 4A in mine) which is what the heater is advertised for (48W). I guess I may have to "push" it a bit with a 24V supply and relay, which may give me the added benefit of faster heating (it took an hour to reach 116.7)

I'll take some measurements on mine after work.

Solved & I write it for the case others will search it up..
It was combination of:
1. PSU was too old & didn't deliver the expected [A], I've changed to new PSU + took all 12[V] (yellow) wires and attached them together to increase the power consumption.
2. Wires weren't designed to pass more than 5 [A] switched the 5[A] so I've switched 2 wires to 10 [A] wires (they are thicker)
3. Added a fan directed towards the MOSFETs.
Currently my printer is up to 110 [C] and stable at that temp without dropping down regardless of other issues.