2

My printer (Prusa i3 Mk3 rework clone, with a 200 x 200 MK3 ALU-heatbed, RAMPS 1.4) has a car relay for heated bed control. It works just fine using bang-bang on Marlin, but I would like to replace it with a MOSFET anyways. The specialized MOSFET heating modules are way too expensive in my country, so I was thinking about buying a MOSFET that's good enough to handle my 12 V 10 A bed with Marlin's PWM.

Would a MOSFET like the STP80NF70 be enough (0.01 Ω at vgs 10 V, 68 V 100 A) or would one of the bigger ones, e.g. in TO-3P encapsulation be needed?

It would go in a proper heatsink and ran at 12 V with an optocoupler

Trish
  • 20,169
  • 10
  • 43
  • 92
Matias Gonzalez
  • 105
  • 3
  • 1
    [Mk](https://en.wikipedia.org/wiki/Mark_(designation)) 3 of **WHAT**? Mk3 is just "Iteration 3", oyu have to say the manufacturer or base model together with it. A "Native Instruments MASCHINE Mk3" is something entirely different from a "VW Golf Mk3" or a "Prusa i3 Mk3". The only Item that has no name with it is a British heavy tank, the "Mark" series. – Trish Mar 30 '19 at 22:09
  • @Trish If there were many brands, it would be necessary. But in regards to 3D printing, I assumed most people would know (or know to google) that this applies to the RepRap project and its designs, i.e. the Mk1 is Prusa's original design, while Mk2A and Mk2B are revisions (forks) on it (all of these are etched PCB heated beds). Then the Mk3 is simply another iteration but made differently with an aluminum plate. Luckily this does not matter as I provided all the necessary info about my bed (namely, that it's a 12V 10A bed) so you don't need to know what a Mk3 heatbed is to reply. – Matias Gonzalez Mar 30 '19 at 23:16
  • 1
    If you mean the heatbed developed by Josef Prusa, the name is "[PCB Heatbed Mk##](https://reprap.org/wiki/PCB_Heatbed)". and the correct name of the 3rd iteration in aluminium is [Mk3 ALU-Heatbed](https://reprap.org/wiki/PCB_Heatbed#MK3_ALU-Heatbed_Dual_Power), **not** Mk3 aluminum or just Mk3. – Trish Mar 30 '19 at 23:34
  • @Trish and that's pedantic, but true. – Matias Gonzalez Mar 31 '19 at 02:31
  • Can't you order at those typical Chinese sites? I ordered MOSFET modules there for a few Euro's; takes some time, but they work flawlessly, they don't get hot either. – 0scar Mar 31 '19 at 07:14
  • @0scar I wish... I live in Argentina. Getting something from outside the country involves having to mess with customs and stuff like that. (usually having to go to customs and fill in paperwork) – Matias Gonzalez Mar 31 '19 at 14:16

1 Answers1

3

and welcome to the Stack Exchange 3D Printing site.

Let's look at the specs compared with the requirements.

How much power is used by the heated bed? You have specified 120W (12V, 10A). That information makes it easy! On my homebrew printer it is higher, but that seems like a good high limit for a typical i3-style design.

So, the MOSFET you propose, assuming you completely turn it on, will dissipate 0.01 ohms * 10A^2, or 1 watt. This will warm the part, and you probably should attach it to a heat sink, which could be as simple as the existing extrusions. There is a much more complete answer to this exact question on the Electrical Engineering Stack Exchange site: Dissipating 1W on a TO-220 without heatsink?

I doubt the PID frequency is higher than 40 KHz. Since the Vgs rating you referenced is specified with 10V on the gate, you can't drive that directly from an Arduino pin. You need another driver transistor (NPN or FET) to drive the power FET gate. The turn-on time for the power FET will be determined (in the simplest circuit) by the pull-up resistor to +12v connected to the output of the transistor and the gate of the FET.

cmm
  • 4,418
  • 10
  • 36
  • I would hope that is the case, but the problem is actually more complicated if you factor in PID control - the MOSFET'S capacitance MIGHT burn the Arduino pin if the frequency is too high, so current limiting resistors to the gate are used. then switching is slower and lets the MOSFET be in its intermediate state longer, which makes it heat up much more than on its saturated states. Since I don't know the PID frequency used I can't know if this is going to be a problem, and seeing all those TO3P fet's that supposedly withstand about 200 A makes me wonder if I'll be doing something wrong – Matias Gonzalez Mar 31 '19 at 21:06
  • Your question was important. I added information to the answer. – cmm Mar 31 '19 at 21:17
  • Actually you're right, I was planning on using 12V via a separate transistor... Thanks, I'll definitely do this in the near future – Matias Gonzalez Mar 31 '19 at 21:20