Prusa i3 XL not able to reach 110C@24V

Dear Community!

Few days ago I was investigating together with you the issue where my Prusa i3 XL (large heatbed 200x300mm) was melting the prints directly on the heatbed (link). With your help I figured out that I soldered the heatbed incorrectly, so one part of the heatbed had 140C, while the second had only 40C.

My heatbed is rated for 140W. It has 3 soldering pads with numbers 1, 2 and 3. When using 24V input I should solder only pads 2 (VDC) and 3 (GND). For 12V input pad 1 is VDC, while 2 and 3 are both GND. I have corrected the soldering, so now my heatbed is connected to pads 2 and 3.

The problem that I'm facing is that the heatbed is not able to reach 110C.

The resistance between pads 2 and 3 on the heatbed is 6 Ohms.
The current that is being drawn from the PSU is 4,5-4,73A.
PSU is composed of two ATX power supplies working at 24V/17A.

At these values, it seems that the bed is running 108-115W at maximum, while it's rated for 140W.
The printer is working in a cold room at 0-5C, but it's located in a closed box.
The heatbed is driven by IRFZ-44N FET transistor (datasheet) capable of running 55A. It reaches 30C when the heatbed is pumped the most.

Please, oh please, help me figure out why this printer cannot reach 110C.
When the bed was connected incorrectly it was soldered to pads 1 and 2. The resitance at those points is 3Ohms, so the bed was running at 205W. With these values I was able to reach 140C in 15 mins. Of course I don't need this temperature, but it seems that the PSU is able to supply enough current to drive the bed to this point. My last idea, and the craziest is to connect 24V PSU just like 12V power supply, but the resistance there is 1,8Ohms meaning that at constant drive the power would reach about 370W, but this will for sure damage the bed. In order to protect against that I could upgrade the Marlin firmware in the ATMega board to version 1.1, where I could control heatbed PID, making something like PWM control for the bed. If I would set it up for half impulse infill it would reduce the power supplied to the heatbed by 0.5. I don't know if my assumptions are correct, but I'm thinking hard about it. Let me know what do you think.

I was using this printer for 2 years with successful PLA prints, but the printer was running on stock RAMPS 1.4 FET transistor with 12V/11A supply. The heatbed was barely reaching 65C.

Edited 2 time(s). Last edit at 02/17/2018 02:36PM by kjakobczyk.

Have you verified with your multimeter that the bed heater is actually getting 24V?

140W on a 200x300mm bed is about 0.23W per square cm. That's rather low. You really want at least 0.3W per square cm and preferably 0.4W. Also, the resistance of a PCB bed heater increases with temperature, so you need a higher power when cold to ensure that you still have enough power when it is hot.

My advice is to give up on PCB heaters and use a piece of aluminium tooling plate with a silicone heater on the underside. But if you want to persevere with that bed heater, you could buy a 24V PSU and turn it up to 28V.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

On my CoreXY I use a 600W DCDC-boost converter for the bed heater and cranked it up to 28V. It only cost ~20$.
The downside is, you can only run the bed in bang-bang mode, but I still achieve +/-0.2° temp error.

I was forced to do that, because Duet boards are only rated for 25V max. So following DC42's advice might kill the controller.

Edited 1 time(s). Last edit at 02/18/2018 03:15AM by o_lampe.

Thanks for the replies Guys!

@dc42: indeed in my whole research I forgot to verify the voltage that goes to the bed. I guess I assumed that since the voltage on the PSU is 24V (measured) and the only element between PSU and heatbed is FET transistor I can expect the voltage to be around 24V (the only voltage drop will occur between Drain and Source). Anyway the bed is rated for 140W, so I can't bump the W/cm^2 too much. Having 0,3W/cm^2 should be reachable, because the bed would run at 180W (it was already working at 205W and the only thing that happened was the white paint become more yellow), but going for 0,4W/cm^2 would certainly damage the bed in bang-bang mode. Quite interresting is the fact about the aluminium bed you mentioned - can you point me to some shops where I could buy good aluminium heatbed? It may be the case that I'll damage curent PCB heatbed, so in that case I'll go for aluminium one.

@ o_lampe: I've already bumped up the voltage from 12 to 24V and this didn't bring any change. For 12V the bed resistance is 1,8Ohm and it was barely reaching 65C, for 24V the bed has 6Ohms resistance and still the temperature cannot go much higher and it takes forever to reach that temp. Maybe a better idea would be the use voltage splitter to lower the voltage from 24 to 18V and to solder the heatbed for 12VDC input? But where I will get a voltage splitter for 400W? Tough issue... As for the rest of the electronics in my printer, the ATMega and steppers run from 12V, so there's no limitaiton on that side. I won't burn the electronics by messing with heatbed voltage because it's totally separate output.

There can be a huge voltage drop between FET and bed, when using thin or too long wires. Same for the wires from PSU to the controller.
The resistances you measured indicate me, that it's more like a 100W bed. ( even less )
A powerful DCDC buck converter is also easy to find. If that's what you meant with voltage splitter. Just make sure, the output voltage is adjustable...