RepRap Builder

May 2014

RepRap Heating Issue Continued


May 2014

The RepRap was finished now in so far (image 1) that it was possible to get it working when the heated bed was covered with an insulating oven glove in order to prevent heat escaping into the environment. The oven glove had to stay on the bed until the set temperature of 110C was reached and than -when the RepRap started printing- the oven glove had to be grabbed away rapidly. Printing was performed well with a heated bed that dropped its temperature to about 103C but the problem appeared now after printing when it was tried to remove the printed object. It was really impossible to remove it from the bed!
This called for a more drastic approach.

reprap complete
Image 1: RepRap Completed ( FRS 2014)

Image 1 shows the RepRap in the completed state. The RAMPS controller system unit (black top cover with ventilator shown at the center right in the image) has been mounted on a transparant polycarbonate support.

For the now chosen more drastic approach it was decided to review the heated bed setup totally by changing the heating power to a 24V at 20A system. Apparently more RepRap builders had suffered from heating problems with the bed because recently Mk2b heating beds had become available that would permit operation at either 12V or 24V. The 24V was obtained by doubling the etched circuit (that was in the Mk2a version at one side of the board) to both sides of the Mk2b circuit board.

It was quite sure that changing to 24V would resolve the heating problem but because the heated bed assembly had to be taken apart now, it was decided to implement a few more improvements:
  • Replacing the two multiplex support boards by one 3.2 mm thick aluminium plate (image 2 and 3). This was considered to be sufficiently stiff for carrying the heated bed, it had less weight than the current construction and it would enable to gain some more height for objects to be printed.
  • The earlier mentioned Mk2b 12/24V heated bed (image 6 and 7).
  • Replacing the glass plate (mirror) covered with Kapton tape print bed by a plastic coated aluminium print bed (image 8), providing better heat conduction.
  • Placing a 4 mm thick cork plate below the Mk2b heated bed to prevent heat to escape from the bottom of the bed.
  • Adding a solid state 12V/24V relay, capable of switching 40A (image 9).
  • Constructing a separate power supply capable of delivering 24V at 47A (image 10, 11 and 12).

The images below show the different components to be added or changed:

aluplate carrier aluplate mounted
Image 2 and 3: Aluminium heat bed support board (Source: Supplier, China)

Image 2 shows the pre-drilled and threaded aluminium support board and image 4 shows the support board mounted in a RepRap in a four point adjustment position (this is an example picture and not our project).

The actual aluminium heat bed support that arrived from a supplier in China had to be adapted because it had the following shortcomings:
  •  the support board lacked a fifth hole for mounting the heated bed in a three point position enabling easier leveling (image 4, green marked);
  • the support board had three sets of holes for mounting the board on the horizontal rod bearings, wheres the RepRap has four sets of sliding bearings (image 4, red marked;
  • the distance between each set of holes for mounting the sliding bearings on the board appeared to be 28.5 mm, whereas the printed parts have a distance between the holes of 30.0 mm (image 5).
When mounting the support board in the RepRap it appeared that the horizontal rods had to placed over the threaded rods instead of initially hanging underneath the threaded rods ( see image 1 for the smooth rods under the threaded rods). This meant that the idea of gaining more vertical space for printing had to be abandonned, though a little vertical space was indeed gained.

Image 4: Aluminium support board with extra holes needed ( FRS 2014)

Image 5: Aluminium support board with wrong distance for mounting sliding bearing holder ( FRS 2014)

The red and green marked holes
in image 4 had to be drilled as an extra job. The too narrow distances between the four sets of two holes for attaching the sliding bearing holders (image 5) were left as they were, because the size of each of the holes was 0.5 mm larger than required and this just permitted (with some forcing) to mount the bearing holders.
Apart from these little problems, which had to be resolved, the aluminium support board was considered a useful improvement of the RepRap.

Mk2b a    Mk2b b

Image 6 and 7: Mk2b Heated Bed (left) and Connecting Schema (right) (Source: Supplier China)

print bed
Image 8: Aluminium Print Bed; insert: bottom side (Source: Supplier. Germany)

Image 8 (large picture) shows the aluminium print bed with the plastic coated top surface. The cutout in the print bed at the left of the large picture is for leaving sufficient space for connecting the wires to the heated bed. The smaller insert shows the bottom side of the print bed with the milled groove ending in a well shaped space for guiding the thermistor cables and for placing the thermistor bedded in thermoconducting paste.

Image 9: Solid State Relay

The solid state relay in image 9 has been chosen to protect the power switching FET on the RAMPS controller board to be blown because of the higher voltage and amps that will be drawn by the heated bed. The RAMPS will switch the relay with its usual 12V output on the RAMPS D8 pins. The secundary on the relay deals with the 24V at max. 47A from the additional power supply of images 10, 11 and 12 and connects this power to the heated bed.


Images 10, 11 and 12: HP Server Power Supply: left: front, center: back, right: label ( FRS 2014)

The 24V power supply will be constructed fromn two 12V HP server power supplies (images 10, 11 and 102. The HP power supplies are very reliable and they are sold at prices ranging from € 12 to € 20 each. Instructions how to convert these power supplies and how to combine them in series for obtaining 24V can be found here and here.

The choice for these powerful power supplies has been made because they will also be used in a different project for feeding a high voltage power supply. More details about this projet can be found here.

When all new parts had been collected it was decided to replace them with the old parts in the RepRap piece by piece and to test the RepRap after each change. This would permit to observe if these part by part upgrading had the desired effect.

The first part that was exchanged was the wooden bed by replacing it with the new aluminium support board. Then it was decided to retain the old heated bed but to exchange the glass plate on top of it by the new aluminium print bed and also by now installing a cork plate underneath the heated bed for better insulation against bottom heat loss.

After re-levelling of the print bed the RepRap was tested for heating up the bed to a set point of 110C. To my surprise it appeared that the exchange of parts so far did not make a significant difference. After a heating period of 20 minutes the bed had heated up to a deltaT of 70C, i.e. from ambient 20C to a temperature of 90C, and after 15 minutes more the temperature reached a temperature of 102C and remained at that temperature with a variation of 0.5C. The bed output was all the time at 100%, i.e. the PID algorithm did not cut in and full power was on the heater bed permanently, or with other words, the very slow speed of heating up (the slope angle of the temperature curve) did not require the PID algorithm to start working. Apparently the heating power of the bed heater was insufficient to compensate for heat loss at the surface of the aluminium print bed, similar to the glass plate! Moreover, the cork insulation underneath the heater plate did not (sufficiently) reduce heat loss from the bottom of the heater bed. This was tested by measuring the temperature of the aluminium support board, which is located 24 mm below the heater plate (20 mm air gap and 4 mm thick cork plate). At 100C of the aluminium print plate, the aluminium support board was found to have a temperature of 44C.

A quick conclusion that can be drawn from this experiment is that with the same heater, an aluminium coated glass plate of 2 mm thick and a surface of 400 cm2 has the same heat loss as an aluminium print bed of 3.5 mm thick and the same surface!

Next steps will be to add more insulation under the heater, to exchange the Mk2a heater for a MK2b heater and to operate the RepRap at 24V.


Last Updated on: Mon Nov 10 22:19:03 2014