ABS melting on the heated bed at 110 degree

Hello Dear Community,

I'm new to the topic, but was struggling hard past times to get my ABS printout correct without any result, so I decided to ask for professional help. I've built Prusa i3 with larger heatbed (200x300), 24V power supply for the bed, 12V for the rest, hardened 4mm glass on the bed. I have no problem in reaching 110 degree on the heated bed even if the printer starts in an environment with 1-2 degrees Celsius, but when it comes to printout about 7-10 mm of layers counting from the bed gets melted and deforms heavily.

This is the printout of 30x30x30 calibration cube I've tried several days ago:


And the bottom looks as follows:


Awful, right?

Please have a look at the degraded bottom of the part - can you tell me what I'm doing wrong? I'm printting at 240 for the nozzle, 107-110 for the bed, 10mm/s for the shole model, retraction enabled, 0,2mm layer high.
Please also look at the top of the part - can you see the last layer is not completely solid? Can you point me why it's happening?

I'm printing with Cura 14. I've been able to print on PLA with success.

Best regards,
Chris.

Edited 1 time(s). Last edit at 02/12/2018 05:18AM by kjakobczyk.

Independently verify with a thermocouple that your bed is actually at tempature you expect.

For that to happen it looks like it got way hotter.

I suspect incorrect thermistor setting or the thermistor has moved away from the bed giving lower reading

Im with Dust. The bed temperature is obviously wrong. Don't forget its the bed heater that runs at 24V not the thermistor. Also you have to solder the proper connections on the bed for 24V not just plug a 12v bed into 24V.
How many top layers are you printing? Looks like one so its effectively bridging and you can get some "droop" between supports. I usually run 2 or 3 top layers so as to build support for the final layer.

Thank you for the answers guys!

@Dust:
I'll verify the temperature with external termometer tomorrow. I initially thought this may be the cause, but didn't check it out since there were some confusing information over the internet about bed leveling that may lead to the same effects.

@MCcarman:
The bed is connected to 12V leads unfortunatelly and this was my intended action. Initially the bed was running at 12V with stock transistor on the Ramps 1.4 board, but the heated bed could not reach more that 65 degree what was barely enough for PLA printout. Since I wanted to move to ABS I had to figure out how to heat the bed to 110 degree. I started by changing the FET transistor on the Ramps, but it didn't change a thing, except the temperature on the FET. The stock one had 90 degree when working, the new one dropped to 50-60 degree. Then figured out that heated bed has too much resistance (about 6 ohms when connected properly to the soldering points), so bed instead of 140W was reaching something around 100 and I thought this may be the reason, so I went for 24V/17A power supply (two ATX joined together - bare with me). Having so much of power, FET that was capable of running huge current I was sure that the bed will reach target temperature. Wrong! It was still stuck at 60 degree, so I hooked up my multimeter and figured out that the resistance on the 24V bed circuit is also incorrect compared to what documentation says, so I have found the path on the bed that has correct resistance, soldered the points, run the power supply and reached 110 degree. Also, my thermistor is located near the power soldering points, on the edge of the bed and not in the middle of it and that may cause the difference, but could it be that between center of the bed and the edge there is 10-15 degrees difference?

As for the quantity of top layers for the printout I don't have the idea where to check it in Cura. From my observations from PLA prints I think it's 2 layers with 10% infill.

Edited 1 time(s). Last edit at 02/13/2018 02:05AM by kjakobczyk.

Maybe half of your heat bed heater coil is broken/burnt? That would explain why it gets too hot, but the thermistor doesn't read it. It would also explain why the coil resistance differs from the manual.
For a test you could heat up the bed to 80-90°C and add some water droplets to different places of the bed. ( You DO have Borosilicate glass, right? )

@o_lampe:
I'll do the test as you suggest, this is easy. I have however notice that the bed heats up pretty evenly without much measurements, so I think it's not the case that coils are broken. Such metling occurs in every part of the bed, I've tried dozens of prints all around the bed with similar results. As for the borosilicate glass you told me something interesting: initially I had normal glass on the bed, 3mm thick and this ended up in two pieces of glass crushing at 85 degree. So I went for hardened glass, 4mm thick. I don't know if hardened and Borosilicate are the same things, but this glass is able to take thermal shock easily. The only thing that was hard for me was to properly level the bed at 110 degree, because both the nozzle and bed expanded due to temperature and first prints was like the nozzle was too close to the bed, they were actually touching each other. Now the bed is being properly levelled I guess, but still there's significant difference between the thickness of the glass at 5 degree and 110 (or whatever the temperature truly is).

Hello again guys,

I've got the updates. Following your advices I've made an experiment: I hooked up external temperature sensor with kapton tape directly to the bed and I placed 8 drops of water in the corners of the bed and in the middle. I let the heatbed to heat up, it took about 15 minutes from 5 degree Celsius and the results suprised me a lot! Take look at the photos:

In the above photo you can see two drops of water on the bed heated up to 110 degree according to YARRH software. It seems that the bed is not heating up evenly (thanks to @o_lampe for this idea with water)

Next, you can see more droplets, but most of them vanished already due to heat and some others keep staying at the bed. You can also see red line which shows a gentle mist on the glass where the temperature if falling down.

And finally the root cause of the issue, bed is heating to 137 degree looking at the bottom of the photo, and 40 on the top side. Instead of that YARRH is showing 110 on the thermistor.

I guess I need to start with dismounting the bed and soldering it properly. Next I'll measure the temperatures again to see on external temperature sensor if I'm able to reach 110 degree.

Edited 2 time(s). Last edit at 02/13/2018 03:23PM by kjakobczyk.

Using 24V for the 12V trace leads to 4 times the power. Maybe too much for the trace and it burnt.
I've raised the voltage of my 12V PSU to 14V and now I have enough power to heat up to 110°C. With sufficient powersupply and thick wires, making sure they have good contact with the terminals on the controller.

OK, but do you think that if the paths on the bed would be damaged I could visualy see it? I was inspecting the bed several times and didnt notice any problem. It of course may happen between inspections, but this I will verify today, because I need to dismount the bed and desolder from 12V, soldering it back to 24V.

I remember that with stock FET on RAMPS and 12V/400W ATX power supply the FET heated up dramatically, up to 100 degree, the voltage on the ATX dropped to 11,7V and the temperature on the bed could not go more than 65C. Also keep in mind that this is an XL (200x300) bed rated for 140W, so at 12V there will be 11,7A. Although the ATX 12V lines were rated for 12A I was rather reaching the PSU limit, that's why I decided to go for 24V. How did you bump up the voltage from 12 to 14V? Do you have regulated PSU or you went for 200W power booster?

Edited 2 time(s). Last edit at 02/14/2018 03:28AM by kjakobczyk.

I have Meanwell 12V/350W PSU with a potmeter to adjust voltage. A booster would only help with a stronger (ATX)-PSU. It has to deliver the losses of the booster, too.

I'm afraid, the stock FET got so hot, because the Atmega onboard 5V regulator doesn't deliver 5V. Then the FET doesn't fully switch open and get's hottttt.

ATMega is powered directly from the PSU and to be honest I've blown up 3 boards in the past and all of them (when they were working) caused the FET to heat dramatically. I'm using Marlin v1.0 firmware, so it's not capable of doing any PWM on the FET, just regular on-off cycle, so it would be strange that it's not fully opened. The regulator in the ATMega board works from much lower voltage than I was supplying even when the PSU was 12V/400W (giving 11,7V at heatbed operation).