3D Printer Heated Bed Insulation Mod

Improve your 3D printer's heated bed for better thermal stability and reach higher (ABS print) temperatures

Similar projects worth following
We wanted to print with ABS with our 3D printer but couldn't maintain a stable temperature over 95° C. After a little research we found SpaceLoft, a material from Aerogel Technologies, that got us to a stable 110° C. Since it is both light and has very low thermal conductance, we were to able to insulate the bed without affecting gantry move performance. The project and instructions can found on Instructables in the links.

Notes on the Instructable Guide:

Step 1: We found that we couldn't reach 110° C or even maintain 95° C with any sort of stability with our cardboard and aluminized mylar backed foam homemade thermal isolation. Without any thermal isolation (as built) the print bed could only reach 70° C so the first hack was a big improvement. But we were looking for even higher temperatures, so we went looking for a better light weight insulator. We found a reference to SpaceLoft and went from there.

Step 2: We were in the US when I ordered the material and guesstimated the clearance. I was a bit off, since there was less clearance under the bed cantilever.

Step 3: We had ordered special high heat tolerant tape (3M) and getting the old isolation off was tough. Cleaned the adhesive residue with a little acetone. That worked well as the heated bed heater was based on Kapton tape.

Step 4: We used the old isolation as a template for the new SpaceLoft + Kapton Film isolation layer. Luckily we had a pair of decent heavy duty scissors and brand new X-acto blade. The Aerogel Blanket clogged the permanent ink marker dry, so we ended up using a whiteboard marker instead. We noticed that gloves were a must when handling the insulation.

Step 5: Oops the material was too thick for the cantilever. Go back one space...

Step 4: After marking the Kapton film, we cut out it's silhouette in the plastic and excavated about 3mm of the insulation to make room. The sharp X-acto knife did the trick.

Step 5: Rinse and repeat, now the wiring was binding between the insulation and the cantilever.

Step 4: A bit more Kapton cutting and material excavated.

Step 5: Success!

Step 6: The dust is fairly nasty as it immediately drys out your skin when touched. We guessed it's going to play hob with extruded plastic sticking to the bed. So we decided to seal the ends with Kapton tape.

Step 7: Tested again using the same enclosure and we got our target temperature. I'll post our first (successful) ABS prints in the near future.

  • 1 × Aerogel Technologies SpaceLoft 30cm X 30cm x 10mm Aerogel Technologies insulation blanket featuring low weight and low thermal conductivity
  • 1 × Heat Tolerant (~50° C) 30cm x 30cm x 1mm Film Can be Kapton or other film that can tolerate a constant 50° C heat
  • 4 × Heat Tolerant (~120° C +) 35cm Tape Can be Kapton or other film that can tolerate heated bed temperature

  • Example of First ABS Prints

    Bill Gertz09/09/2015 at 11:34 0 comments

    Added example of first prints. Previously we couldn't get the filament to stick to the bed properly. After the mod, using 110° C for the first few layers and 100° C for the rest, we're having no problems.

View project log

View all instructions

Enjoy this project?



Bill Gertz wrote 09/15/2015 at 13:19 point

Hey SUF, 

It took a bit of work as there are no schematics for the Felix Printer Controlboard (their name). After a little reverse engineering, they've implemented a straight RAMPS style direct drive (MCU ATMEGA2560 Output/PWM Pin connected biased to ground on the MOSFET gate - take a look at the RAMPS 1.4 Schematic Heaters & Fans below). 

Nonetheless this printer is set for dead time rather than PID. This means the switching rate is far lower than for PWM. (1/5 Hz and more like 1/10 -see holding 105° C) Recent bed output graphs bear this out. Am I wrong in saying that I shouldn't be looking at much power loss due to switching that you've outlined above?

Rather given the large surface area of 30 x 27 cm (.081 m^2) and the anemic bed heater at 144W  (1 Ohm at 12 V), the heater seems underpowered for stable 110° C. Working through the specific heat calculations and heat loss (convection and radiative) 144W looks low.


  Are you sure? yes | no

Bill Gertz wrote 09/09/2015 at 12:30 point

Let me open up the enclosure and cross reference the case marking. Sitting in the middle of a print and don't want to bork it.

  Are you sure? yes | no

SUF wrote 09/09/2015 at 12:25 point


Same kind of problem, different kind of solution.

I had exactly the same problem, finally it turned out that most of the 3D Printers has horrible MOSFET driving for the heated bed, what eats up significant part of the power for the heating. Changing the driver solved the problem. May I ask what kind of driving electronics your printer has?


  Are you sure? yes | no

Bill Gertz wrote 09/10/2015 at 23:41 point


The vendor documented it in his support site thread. The MOSFET is an NXP PSMN1R3-30YL (see: for datasheet). I'll verify that and try to grok the specs tomorrow.

  Are you sure? yes | no

Bill Gertz wrote 09/13/2015 at 00:31 point


Verified the part and seems that specs make the MOSFET suited to task. Checked the part at full on and we are getting a little heating. Am I missing something?

  Are you sure? yes | no

SUF wrote 09/13/2015 at 20:24 point


The problem is usually not the MOSFET itself, if it is suited for the task, mostly yes. The problem is around driving the MOSFET.

The MOSFET as a device is a Field Effect Transistor. It means it need an electric field (Voltage) to keep it in conducting state (in opposite of the BJT what needs current for this). This means micro Amps, even pico Amps.

The designers of this kind of switching circuitry keep only this figure in mind. A typical MCU can drive ~20mA per pin, what is way enough for this task.

What thay mostly don't consider, is the input capacitance of the device. When you switch on/off a MOSFET, you should charge/discharge this capacitance. If you want do this fast, the MOSFET gate can consume even several Ampers, what the MCU can't provide, the result is slow on/off switching. 

When a MOSFET is between the on/off state, it dissipates, because it has some resistance, and large amount of current is flowing through.

If we talking about the heating. If you just switch on the MOSFET, and keep it switched on for a long period, this behavior is not a problem. But if the MOSFET is driven from a PID controller, what is switch it on/off a several hundred or even thousands of time in a second, the MOSFET will be kept in this (linear) state, and dissipate, limit the current through the heating.

The solution of this problem, is either not using a PID style code for driving the MOSFET and just switching the heating on/off a few times in a minute, or using a proper MOSFET driving circuitry with few Ampers gate driving capability.

Sorry for keeping this so long/technical, I just wanted to explain the situation, what I faced trying to properly drive my heated bed.


  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates