Heated Bed and RAMPS 1.2

I followed the RAMPS wiki and hooked up my Prusa PCB heated bed according to instructions "powering the heated bed directly from a separate PSU, grounding on the RAMPS". With the bed temperature set to 80C, the power MOSFET (D8) on the RAMPS board is already too hot to touch. I don't think I can go up any higher without putting a heatsink and a fan. I would like to reach 120 C.

Does anyone have their RAMPS set up this way with a heated bed running at 120C? I'm thinking that I have to use a relay.

Edited 1 time(s). Last edit at 06/16/2011 11:39AM by brnrd.

Nooooo. Don't do that. Plug one wire from the bed to (-) on D8. Plug the other wire directly to (+)12VDC on the power connector to your power supply.

That's what I did. But I don't think this works for ABS (120C) unless there's additional cooling on the power MOSFET.

Use a MOSFET with a lower RDSon @ VGS = 5V or fit a heatsink.

Or add a level changer to increase the gate drive from 5V to 12V. You can see how I did that here: [hydraraptor.blogspot.com] with Q1 increasing the gate drive to Q5 from 3.3V to 12V to get lower on resistance. The extra components only cost pennies.

A relay won't last very long if you look at the spec for how many times they will operate under load and assume it will toggle every few seconds when controlling a bed.

---

[www.hydraraptor.blogspot.com]

... i'm operating my hot end with a separate temperature controller ... and the heated bed with a SolidStateRelay, so it's switching the 230VAC powerline on the heater at 12Volts with only some milliamperes through the MOSFET ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

According to the specs for the power MOSFET on RAMPS 1.2 (stp55nf06l), the Static drain-source on resistance is typically 0.016 ohm and maximum 0.020 ohm with Vgs of 5V. Assuming that the resistance of the PCB heater is 1 ohms, the power being dissipated by the MOSFET should be less than 2.8 Watts with a 12V power supply. That doesn't seem like much so putting a heat sink in combination with the fan in the RAMPS case should be enough.

Would a heatsink like this AAVID THERMALLOY with a thermal resistance of 24 C/W be sufficient?

The other concern that I have is the amount of current that will be going through the ground leg of the power supply terminal of the RAMPS board which should be around 12 A. I think this might be too much in addition to the current drawn by the extruder heater. So, I plan to connect the ground leg of the 12V supply to the heated bed to the 3 ground terminals on the board next to the power connector. Is this good enough?

Edited 1 time(s). Last edit at 06/16/2011 07:51AM by brnrd.

brnrd Wrote:

> Would a heatsink like this AAVID THERMALLOY with a
> thermal resistance of 24 C/W be sufficient?

It should be with a fan, but I would go for something a bit bigger to be on the safe side as without a fan it would get to about 90C. Also bear in mind what temperature RDS is quoted at. It is often 25C, which implies an infinite heatsink and it gets higher as junction temperature increases.

>
> The other concern that I have is the amount of
> current that will be going through the ground leg
> of the power supply terminal of the RAMPS board
> which should be around 12 A. I think this might be
> too much in addition to the current drawn by the
> extruder heater. So, I plan to connect the ground
> leg of the 12V supply to the heated bed to the 3
> ground terminals on the board next to the power
> connector. Is this good enough?

I am not familiar with the board layout, but you should be able to reinforce the ground track with a thick copper wire on the back and then it is only the connector rating you have to worry about.

---

[www.hydraraptor.blogspot.com]

Thanks! I'll look for a bigger heatsink that can fit.

There's a thick and short trace for ground going from the power connections to the D8 powerMOSFET. You can see it in the lower left corner in this screen capture of the board. The power plug terminals are in the corner and just above it, there are 3 terminals that also connect to ground that I intend to connect to.

To be safe, I have reinforced the ground trace by soldering a thick wire between the MOSFET and the 3 terminals.

Edited 1 time(s). Last edit at 06/16/2011 11:41AM by brnrd.

I mounted a heatsink on the D8 power MOSFET as planned and reinforced the ground connection between the 3 gnd pins and the MOSFET. It's now working and getting up to 120C and the heatsink and the MOSFET is cool to the touch. My Mendel is now back up with RAMPS electronics! First print is about to start in progress.

Thanks everyone for the help.

Edited 1 time(s). Last edit at 06/16/2011 08:58PM by brnrd.

brnrd Wrote:
-------------------------------------------------------
> According to the specs for the power MOSFET on
> RAMPS 1.2 (stp55nf06l), the Static drain-source on
> resistance is typically 0.016 ohm and maximum
> 0.020 ohm with Vgs of 5V. Assuming that the
> resistance of the PCB heater is 1 ohms, the power
> being dissipated by the MOSFET should be less than
> 2.8 Watts with a 12V power supply.

The TO-220 package can dissipate up to 2 watts (175c die-50c ambient=125c dT / 66c/w = ~2w) with no heatsink. Above that and you will definitely need either a better mosfet or a heatsink. This mosfet can carry up to sqrt(2W/20mR)=10A with no heatsink.

I use IRL3803 which has only 9mR@4.5v so can handle 15A with no heatsink.

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

I don't know how others are able to run their RAMPS with the Prusa heated bed without using a heatsink on the power MOSFET. Before I switched to RAMPS, I was using the same part number MOSFET and even with a heatsink, it got hot enough to melt the solder.

Are you sure you followed MY setup? One wire from the bed goes to the GND on D8, the other wire goes to +12vdc on the RAMPS power connector. Check the resistance on the bed. Maybe you have a short.

That was the instruction on the wiki. Yes. No short for sure or it would have blown the power supply fuse or the board would have gotten real hot even with a heatsink. The heatsink and MOSFET feels that it doesn't go above 30C when the bed is at 120C. Without the heatsink, I couldn't touch the MOSFET for more than a split second or I would probably burn my hand.

Also, I don't see how you can run 5 (RAMPS+extruder) + 12 A (bed) = 17 A through one ground pin in the power connector on the RAMPS board.

jcabrer Wrote:
-------------------------------------------------------
> Are you sure you followed MY setup? One wire from
> the bed goes to the GND on D8, the other wire goes
> to +12vdc on the RAMPS power connector. Check the
> resistance on the bed. Maybe you have a short.

There's no short, his math is good, I checked it myself. simple result of P=I^2.R, Rds(on)@Vgs=5v, Tj(max) and Tj-amb shows that that particular mosfet isn't capable of supplying more than 10A without exceeding its maximum die temperature unless heatsinked. (I always assume that ambient is 50c because it gets that hot in several parts of the world)

The bed resistance does need to be around an ohm to put out the power that it does from 12v. P=V^2/R, so 12v into 1 ohm will give 144 watts.

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

My RAMPS board uses STP55NF06L - MOSFETs, and they get the job done.

That's what I have in my RAMPS too. You don't have a heatsink on it? How hot does it get when you're bed is at 120C?

brnrd Wrote:
-------------------------------------------------------
> That's what I have in my RAMPS too. You don't have
> a heatsink on it? How hot does it get when you're
> bed is at 120C?

better question, how hot do they get when the bed is at ~65c on its way to 120c? At that point they've probably been on for enough time to reach equilibrium without any breaks for temperature control...

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

I have to say I don't touch it. For all I know its really hot but its been fine since April.

Thanks. That's good to know. I can probably do with a smaller heatsink. But since the resistance of the MOSFET goes up with temperature, it's probably best to keep it cool to put more power to the bed.