Ultimaker original temperature problems

[JohanJonker 2]

Dear forum,

I have a Ultimaker Original that has some temperature problems

When I start printing it heats the head to the desired temperature and that look quite good. I can see the temperatre increase in steps of 1 or 2 degrees Celsius until the startinng temperature is reached.

But as soon as the printing head starts to move the temperature becomes unstable. Most of the time the temperature drops with 20 degrees and more and then sometimes increases 30 and so on. Also when the ventilation start I see effects on the temperature.

I already tried to stablize the power supply by adding a small elco of 4,7uF in parallel with the thermocouple transmitter print. Also bought a new thermocouple and a new transmitter print. Nothing helped. What can be the problem?

kind regards

Johan

[Torgeir 280]

Hi Johan,

There is two things here to be isolated, either the temp is wrong (high) -telling the system to turn power off, or there is an interrupted connection to the head reducing power to the hot end. Both reducing the temperature. I'll go for the latter, -but sure could be wrong.

IMO, your indication is OK, but the power is not!  

How many hour does you printer been working?

So, check the resistance to the heater element in the "hot end" when you're moving the "extruder head" manually around the bed. See if the resistance change during this moving action. Yes you will need an ohmmeter or any kind of "multi meter" that can read resistance to do this test. (Sure, here an ohm meter is connected to the "extruder heater" connector. as we like to see if there is any loose connection in the wiring from PCB to the "extruder heater".)

We'll need to start like this to find out if this is a power interrupt or if it is an EMC problem.

Anyway, good luck.

Torgeir.

Edited October 1, 2016 by Guest

[JohanJonker 2]

hello Torgeir

Thanks for your reaction. I wil do the measurement you suggest.

The temperature is too low. When printing ABS the filament does not fasten to the previous layers. Even when I increase the temperature from 230 to 252.

Kind regards Johan

[JohanJonker 2]

Hello Torgeir,

I disconnected the heater wires from the 1.5.7 board and connected the wires to an ohmmeter. The measured resistance is 20.7 Ohms. When I move the head manually the variation less than 0.2 Ohms and it is random. I can also switch the ultimaker on and move the head by using the display control. The result is the same.

[JohanJonker 2]

I also measured the power supply voltage on the fan on the bottom of the ultimaker.

It starts at 19.20 V.

During heating up it drops to 19.00 V. And during printing it is sometimes 18.97. There is no AC ripple on the power.

the picture above is what the temperature looks like. The temperature should me 252 but fluctuates around 240. The samples are taken about each second.

[neotko 1,417]

Have you try to run PID calibration?

[JohanJonker 2]

hello neotko,

I don't know PID calibration. How does it work and what does it do?

[JohanJonker 2]

OK I found that running the following code executes that PID calibration:  M303 E0 S200 C8

So I made a GCODE file with that content and put it on the SD card.

Then I ran it with print from SD.

But nothing happened.

I suppose that is the wrong way.

But by the way calibration is not a bad idea. What i see is that during heating the temperature is rising quit lineair and quick to 250 then decreases to 248 and then slowly increases to 252 and then printing starts.

Edited October 1, 2016 by Guest

[neotko 1,417]

Yeah that's the way but you need to use pronterface to do it and get values

Sorry busy as hell.

Connect using a usb cable to the printer using pronterface/printrun (free program) and send the gcodes.

Also you will need to write the settings to firmware after changing the pid.

Edited October 1, 2016 by Guest

[zerspaner_gerd 383]

Maybe this helps:

[sOLVED] UM2 PID Autotune failed

reprap.org/wiki/PID_Tuning

M303: Run PID tuning

[Torgeir 280]

Hi Johan,

Wow, you really made your lesson, great..

The measurement of your resistance is good, I'll think no problem here.

The temperature graph is interesting as it normally should not jump that fast..

Just come to think about some more to consider, this might also be a bad thermal connection. The thermal connection between heat block and the thermocouple (assuming you're using the Chrome-Alumel thermocouple). I'm always using the copper paste when installing the "PT100 and the heater element", this to make sure there is a good thermal connection. Another benefit here is that the temp sensor and the heat element never get burned stuck into the block anymore.

Well, maybe the PID tuning as Neotko brought up will do the trick..

Anyway, good luck.

Thanks.

Torgeir.

[JohanJonker 2]

OK I installed printrun from this site: http://koti.kapsi.fi/~kliment/printrun/

And the entered the line M303 E0 S200 C8 in the textbox on the right and pressed send.

Then the following results appeared after some minutes: >>> M303 E0 S200 C8

SENDING:M303 E0 S200 C8

PID Autotune start

bias: 76 d: 76 min: 197.63 max: 206.82

bias: 73 d: 73 min: 196.99 max: 202.76

bias: 68 d: 68 min: 197.72 max: 202.61

Ku: 35.46 Tu: 32.66

Classic PID

Kp: 21.28

Ki: 1.30

Kd: 86.86

bias: 67 d: 67 min: 197.60 max: 202.36

Ku: 35.84 Tu: 34.19

Classic PID

Kp: 21.50

Ki: 1.26

Kd: 91.90

bias: 67 d: 67 min: 197.39 max: 202.36

Ku: 34.30 Tu: 33.95

Classic PID

Kp: 20.58

Ki: 1.21

Kd: 87.33

bias: 66 d: 66 min: 197.75 max: 202.06

Ku: 39.06 Tu: 32.64

Classic PID

Kp: 23.44

Ki: 1.44

Kd: 95.61

bias: 66 d: 66 min: 197.63 max: 202.51

Ku: 34.42 Tu: 33.16

Classic PID

Kp: 20.65

Ki: 1.25

Kd: 85.61

bias: 64 d: 64 min: 198.15 max: 202.15

Ku: 40.77 Tu: 31.85

Classic PID

Kp: 24.46

Ki: 1.54

Kd: 97.38

PID Autotune finished! Put the last Kp, Ki and Kd constants from above into Configuration.h

[JohanJonker 2]

The next thing i did is manually edit the parameters on the ultimaker controller choosing control/temperature/PID-P etc.

Now its time to try the difference.

[JohanJonker 2]

The next thing i did is manually edit the parameters on the ultimaker controller  choosing control/temperature/PID-P etc.

Now its time to try the difference.

The heating up is better now. But as soon as printing starts the temperature jumps up and down.

Torgeir, its a good idea to improve the thermal connection. That will be the next thing to do.

[neotko 1,417]

Does it happen when the fans go on?

[JohanJonker 2]

Does it happen when the fans go on?

 

During heating up the temperature is stable and rising quite lineair without jumping up and down. And as soon as the fans on the extruder turn on the temperature becomes unstable.

Edited October 2, 2016 by Guest

[neotko 1,417]

Then try to use this firmware using the @amedee firmware builder

https://bultimaker.bulles.eu/experimental/

There's an option called 'Slow PWM' activate it and make a firmware with the button on the bottom.

If that doesn't work just restore the firmware.

Remember not to adjust anything that you don't use or know. Any questions here I'm.

@lepaul has a UMO and he did need to use this firmware to avoid problems on the hotend with his powerfull fans, so maybe it works for you.

[Torgeir 280]

Hmm, -no cure with the copper paste.

As you tried to decouple noise on the thermocouple amplifier with the 4.7 micro Farad capacitor, you sure knows about the EMF that can be radiated by the small electronics inside the fan..

The thermocouple amplifiers input signal is in order of a few micro volt at lo temperature levels, so any noise/EMF around here may/will jam this small signal.

I'm sure Neotko has a solution for you here, as he have plenty of experience and at last 3 UMO in various versions.

As I have an UM2 with PT100 temp sensor, such problem is not an issue here.

Good luck in your fault finding.

Best Regards

Torgeir.

[Torgeir 280]

Hi Johanes,

Come to think about those fans as they do radiate lots of pulses when we adjust the pulse width and some of them also radiate when at full speed (the max voltage).

Sometimes we'll use ferrite bed, but this is mostly for high frequencies, so at lo frequencies we simply put an electrolyte capacitor as close as possible to the fan. For those fan we're using a 100 micro Farad / 25 VDC will do the trick. (PS. remember the polarity.)

(This volatge advice is for UMO, for UM2 and like select at least 40 VDC.)

This kind of "EMI" noise from the fan use the feeder wires to the fan as an antenna!

Just my 10 cent..

Thanks.

Torgeir.

[JohanJonker 2]

I found copper paste and used it for the thermocouple.

When I heat up smoke comes from it. Maybe I used to much.

It did not change anything.

Another thing:

When i switch the UMO on. the disply says the temperature is 3 degrees for a few seconds report an MIN temp error and then says 22 degrees which seems normal.

I think I change the wires between the 1.5.7 board and the thermocouple board.

[Torgeir 280]

Hi Johan,

OK., you should use shielded for the signal wire between those board.

BTW., I've just checked the EMI from the two fans on the UM2, here there is two fan in series (12 VDC each. And 24 VDC feeding both). There was a lot of spurious (Lots of HF spikes) in the middle of the two fans.

I've also tested with a 100 micro Farad capacitor, this capacitor removed all the spikes..

If you have an oscilloscope, you'll see it all.

On an UM2, those spikes do not create a problem since the level of the signal from here is much stronger and do not need to be amplified that much.

Good luck.

Torgeir.

(Edit: Sorry for the misspelling of your name.)

Edited October 4, 2016 by Guest

[Torgeir 280]

Forgot one thing.

There will only be smoke until it dry some up, then it will be fine, just went a little.

Torgeir.

[JohanJonker 2]

I think I fixed the problem. I replaced the wiring between the head and the 1.5.7 board. I used just twisted wire and soldered it on the 1.5.7 board. I saw now only 2 or 3 degrees variation during printing.

Thanks to everyone who spend time on solving this problem. I appreciate that very much.

Edited October 5, 2016 by Guest

[Torgeir 280]

Great job Johan.

Thanks.

Torgeir.

[JohanJonker 2]

It appears not to be the total story. When I tested the new wiring I did not guide the new wires in parallel with the other wires to the print head. In fact the distance between the new cable and the other was quite large. When I made a nice cable bundle again. The problem was back.

I also found that when I put my hands around the printer head the problem was decreased. So it is a EMC problem. So the next thing I will do is to use shielded wire.