Weird Thermal Sensor Behavior

[erichunting 0]

Encountering weird behavior with the thermal sensor. Due to a loose fit of the thermocouple in the heater block, I've been having chronic temperature control problems lately. It started with a frequent problem of the heat cutting off shortly after starting a print. Now, in addition, I see frequent 'flatlining' of the temperature where it sticks at one reading until the actual block temperature has shot far past the set temperature. Usually, I can wiggle the wire for the thermocouple or press the metal pin into the block for a tighter fit to coax it back into reading temperature normally. Unfortunately, in the process this is slowly unraveling the metal braid around that thermocouple wire. Now, this would seem to suggest a short in the thermocouple. But today, while trying to coax the thing to start working, I noticed a truly bizarre phenomenon as I was moving the head module back and forth. With the head near the minimum X/Y position (front/left corner) the sensor would flatline at any particular temperature it was at. I often start out with the head in that corner, otherwise it can't identify the center for "print at center" settings. (which ought to be called "print _from_ center", since it never actually centers the workpiece...) But moved away from that corner, it would read the rising or falling temperature normally. How is this even possible? My only thought is that, since that's the point of maximum reach of the head module cable and tube, it suggests a short along that length. Yet, when the temperature fails mid-printing I can only get it working again by finagling with the thermocouple, trying to get it seated more tightly.

Also been struggling with strange spontaneous mid-printing failures. Had a nearly perfect print run of 6 continuous hours yesterday, then, 10 minutes from finishing a tall workpiece the thing spontaneously pushes the platform up and the print head crashes into the unfinished workpiece knocking it right off! It was as though it had over-run what it thought was the Z axis limit (it was still a good 7-8cm from bottom), jumped-up, and started printing from a new level. Pre-Marlin there was a similar chronic problem at the end of printing, the machine raising instead of lowering the platform and ramming the hot print head into the piece destroying it just as it finished. I couldn't even try printing anything that was higher than a few cm. Now it's back again in this random behavior.

I seem to be having an unusual amount of trouble with my Ultimaker. Its original Atmega 1280 failed shortly after assembly, platform got marred when the screws came loose on the upper z-stop, motor driver current was tuned too low (which took months to figure out), and while most people seemed to be printing just fine out of the box I had to wait 8 months for the combination of ReplicatorG-25 and Marlin 3 to print anything that didn't look like a blob. (until ReplicatorG-25, neither Mac, Windows, or Linux versions would load any version of Marlin) With Marlin it seemed like I had finally gotten my Ultimaker for the first time, producing a couple truly excellent prints, but then this thermocouple nonsense starts and I'm dead in the water again. Sure, you have to expect bugs with beta, and I think I've been patient for what is now just about a whole year. But, really, are other users going through this? It seems like one thing is failing after another. I feel like I'm being left behind by what is increasingly looking like a lemon. If this latest problem is a failed thermocouple, I'm probably looking at yet another month sitting on my hands waiting on mail order for a replacement part--assuming I can get one. It's really getting me down.

[Daid 306]

I'm reading that you started out with an ATMega1280. Do you have one of the early workshop machines? Because those machines are quite different.

I've never seen the firmware do moves that are not in the GCode. But the software has been moving very fast for the last few months. Even the RepG25 you mention is quite old already.

As for the thermocouple, when it fails, it's almost always a short in the wires. Check if the insulation is intact, in some cases this is damaged at the hot end side.

[somerwil 0]

Hi,

I also had problems with my temperature control lately. The problems I had were a little different but it resulted in sudden stops in the print... With the printer attached to RepG26 and ticking on the printer head with my finger I saw the temperature jumping from 20 (room temp), up to 60, back to 5 degrees and back to normal temperature again in the end. During a print it will take some vibrations too which resulted in a failed print. Newer firmwares react differently on temperature changes. Older firmwares only cool down the head when a strange temperature was measured, newer firmwares stop printing at the point where the (too) high or low temp was measured.

I replaced my sensor two days ago and it shows good temperatures again. I'm able to print again!

[erichunting 0]

Yes, I had one of the early runs of the Ultimaker. (had just gone through the infamous Botmill Mendel hell some here may be familiar with and, wanting more working volume than the Makerbot, gave up on US kits...)

I haven't seen a specific Ultimaker-endorsed version of ReplicatorG 26 on the support web site. Have people generally moved beyond the Ultimaker-specific versions? I know the current version is 34, but the last time I tried using a version that wasn't Ultimaker-specific it proved futile. All flavors (Windows, Linux, Mac) seemed to be unable to talk to the machine--though I think I may have still been using the 1280 controller.

I'm not seeing any specific damage to the insulation, but most of that near the heater block is concealed by that metallic braid jacket, and clearly part of that is frayed. I'm inclined to agree with the conclusion of a short in the thermocouple. Is this a more-or-less standard form for these on the electronics market or a custom made one? Can I order this in the US from a typical electronic supplier or must I order a replacement from the Ultimaker store?

Thanks a lot for the helpful feedback, folks. I feel a bit more encouraged.

[somerwil 0]

Here's a list of Ultimaker supported software

http://software.ultimaker.com/?show=all

You can fine v26 of RepG there

[mr_seeker 0]

Yes, I had one of the early runs of the Ultimaker. (had just gone through the infamous Botmill Mendel hell some here may be familiar with and, wanting more working volume than the Makerbot, gave up on US kits...)

I haven't seen a specific Ultimaker-endorsed version of ReplicatorG 26 on the support web site. Have people generally moved beyond the Ultimaker-specific versions? I know the current version is 34, but the last time I tried using a version that wasn't Ultimaker-specific it proved futile. All flavors (Windows, Linux, Mac) seemed to be unable to talk to the machine--though I think I may have still been using the 1280 controller.

I'm not seeing any specific damage to the insulation, but most of that near the heater block is concealed by that metallic braid jacket, and clearly part of that is frayed. I'm inclined to agree with the conclusion of a short in the thermocouple. Is this a more-or-less standard form for these on the electronics market or a custom made one? Can I order this in the US from a typical electronic supplier or must I order a replacement from the Ultimaker store?

Thanks a lot for the helpful feedback, folks. I feel a bit more encouraged.

If you have bought the machine from their webstore, its an 2560. If you have got it from their workshops, contact Harma.

ReplicatorG from replicat.org is makerbot flavored... They say they "support" Ultimakers, but I wouldnt count on it (knowing that their development cycle stopped a long time ago). Thats why Ultimaker is shipping their own software.