Thermocouple failure?

[3poro 9]

Today, in the middle of a fairly typical print with my UM2, I saw smoke rising from the nozzle and material was not extruded. I switched off power and examined the print head. I noticed the thermocouple being somewhat loose - I could actually pull it out from the nozzle unit without using any force:

 

 

What's the cure? Stuffing the TC back? New TC? New nozzle unit? Something else?

 

[illuminarti 18]

The thermocouple is supposed to be in a metal sleeve that fits into the block. Did it come out of that? I presume that the thermocouple is failing somehow, and reporting incorrectly? What temperature does it report?

You probably need a new thermocouple, at the very least. How does the rest of the head seem?

 

[3poro 9]

Yes, the thermocouple seems to have come out from its metal sleeve. I see the rim of the metal sleeve in the block, but that one doesn't come out - at least not without using some amount of force.

I now switched on my printer to check the temperature the TC reports - it looks quite normal, although a bit low. When I keep the tip of the TC inside my palm, I get 28-29C - while a body thermometer shows around 32C. Well, I understand the UM TC is probably calibrated for temperatures much higher than that.

The printing head looks pretty normal.

Do you know if the TC is some standard component which could be ordered from different sources - or is it something UM-specific? Maybe the metal sleeve is tailored for UM2 or something...

 

[gr5 2,167]

I'm certain UM will send you one for free if you open a ticket at support.ultimaker.com.

I believe it is a PT100 which is a platinum part that is 100 ohms at 25C. This is a very common part and comes in all kinds of shapes and sizes and they all have identical temperature versus resitance curves so you can replace it with any of those.

you can't get the metal sleeve out without removing the long screw in the top of the heater block. But to get to that you probably need to take most of the head apart - it's not so hard though. Remove the fans (4 screws) and then twist the 4 long thumb screws and take them all the way out. At this point you can remove 2 screws to seperate the rear fan (hold it together while you remove as there is a spring which is quite obvious). Now you should be able to remove the long screw (need the right size hex driver) and then the heater and thermoresistor slide out.

 

[gr5 2,167]

I suspect nothing is wrong with your temp probe other than it slipping out. Being partly outside the heater block it was reporting a lower temp and so the heater compensated and overheated your nozzle.

 

[3poro 9]

gr5 - thank you for the instructions, I really appreciate. The sleeve is quite tightly stuck to the heater block, here's a photo:

 

Do you happen to know the dimensions of the PT100, I could probably drill the sensor out from its compartment? I would be ready to order a new sensor online somewhere if I knew what to order. With Ultimaker's response times (besides some helping hand from Sander), I might end up waiting for their response for some unspecified time.

 

[stevegt 1]

The UM2 github repository shows that thermocouple as Ultimaker part number 1185-B2P-A, PT100 B sensor, 3 mm diameter. https://github.com/Ultimaker/Ultimaker2

Some quick googling found this one -- not an exact match, but the diameter is right:

http://www.tme.eu/en/details/pt106054/resistive-temperature-sensors/

I don't see any epoxy step in the assembly manual -- I wonder if the thermocouple sleeve is fusing to the block in some sort of thermochemical process?

It just occurred to me... the STEP file is in github, and Shapeways can print in brass. You'd want to tweak the STEP before exporting as STL (make drill holes smaller and make sure the external threads have the right OD for use with a hand die). Then finish the printed part by hand, drilling and threading the details. So, if you're in a hurry, you could get a raw brass block in hand from Shapeways, in a couple of weeks, for about $25 USD (not including shipping). :smile:

Steve

 

[stevegt 1]

On the other hand, if your existing PT-100 is still working electrically, you might just get some high-temperature adhesive and re-install it in the old sleeve that's stuck in the existing block. It sounds like that's what failed in the first place anyway. Could be the vendor for that PT-100 cartridge used the wrong epoxy. The 1185 data sheet wants a 0-400C temperature range.

Steve

 

[anon4321 16]

I suspect that the constant heating and cooling of the block causes the sensor to stick in the block,

Get some high temp gloves and pliers and heat the block back up to temp and see if the sleeve loosens.

 

[3poro 9]

 

 

 

The UM2 github repository shows that thermocouple as Ultimaker part number 1185-B2P-A, PT100 B sensor, 3 mm diameter. https://github.com/Ultimaker/Ultimaker2

 

Thank you! So far the only real help I received from Ultimaker was the link to a document where I could see the dimensions of the Pt100. That helped me to DRILL the sleeve out.

 

I don't see any epoxy step in the assembly manual -- I wonder if the thermocouple sleeve is fusing to the block in some sort of thermochemical process?

 

I guess the reason for the sleeve getting stuck is simpler than that. It looks like it's very tight fit. When the long brass screw locks the sensor to its place, it leaves scratches on its surface - those scratches are coarse enough to lock the sleeve in place permanently. At least in my case the threads on the long brass screw were worn out from the tip, so there must have been some pressure applied on it.

 

It just occurred to me... the STEP file is in github, and Shapeways can print in brass. You'd want to tweak the STEP before exporting as STL (make drill holes smaller and make sure the external threads have the right OD for use with a hand die). Then finish the printed part by hand, drilling and threading the details. So, if you're in a hurry, you could get a raw brass block in hand from Shapeways, in a couple of weeks, for about $25 USD (not including shipping). :smile:

 

Well... If "being in a hurry" prompts getting the part custom made by Shapeways rather than relaying on Ultimaker's support, I'm really worried. Actually, ehem - I am really worried :S

 

[3poro 9]

On the other hand, if your existing PT-100 is still working electrically, you might just get some high-temperature adhesive and re-install it in the old sleeve that's stuck in the existing block. It sounds like that's what failed in the first place anyway. Could be the vendor for that PT-100 cartridge used the wrong epoxy. The 1185 data sheet wants a 0-400C temperature range.

 

What I didn't document on this thread - but I did describe it to Ultimaker through the ticket system - were the next phases of my adventure. I managed to drill the old Pt100 sleeve out and I found from a local brick and mortar store a Pt100 which was smaller in size - so I used high quality thermal paste to fit it in. It worked, but not for very long:

 

I don't know what exactly went wrong, but this time the sensor didn't cause anything to start smoking. However, when trying to disassemble the head, I encountered another problem:

 

There seems to be a small leak of PLA through the threads fixing the hollow bolt to the nozzle unit. Strange as such - I'm 100% sure I secured the bolt pretty well. Now, in any case, it is pretty well secured :S

So - I should heat up the nozzle unit in order to melt the PLA - which would be needed to get the hollow bolt off. Heating the nozzle is not that easy with a broken Pt100 inside it.

What adds to the challenge is the fact that the long brass bolt now is _very_ tightly in its place due to partly worn out threads - and its Allen head is not in great shape.

I guess I have an idea how to get the thing into pieces, but it would be easier to get some spare parts. I have offered to pay for them, but so far no luck.

 

[gr5 2,167]

Heating to 100C might be enough so you could use boiling water to loosen the "isolator nut" (that's what UM calls it). Although I recommend 180C. Once the white teflon part is out everything else (the metal parts) can handle 300C no problem.

Regarding the thin screw with the allen hole on top - you can just use vice grips as damaging the upper threads shouldn't be a problem.

Regarding paying UM for parts - I don't think they have a way of accepting payment other than for the stuff in their store. So they usually send you free parts even long after warranty is over and even if you admitted that it was your own fault that it broke. I think they are interviewing for more tech support positions but that takes time also.

If you need a spare part and you are in the USA you can get special treatment - please update your location settings in your profile and if you are in the USA also send me a PM and I can hook you up with free replacement parts (nozzle and temp probe in this case).

 

[gr5 2,167]

By the way - when putting the head back together make sure the distance between isolator and isolator nut is as small as possible (about 1mm). This reduces the spring force on the isolator and reduces the friction and reduces some underextrusion issues.

 

[gr5 2,167]

Once the white teflon part is out everything else (the metal parts) can handle 300C no problem.

 

Which allows other options like flame, heat gun, iron, steam, torch, soldering iron, angry bees...

 

[stevegt 1]

Skimming through temperature.cpp, if I understand it correctly, you'll only get that error if the thermocouple is either shorted or open. How and where is the new thermocouple joined to the old cable harness wires?

My biggest concerns for spares so far are that hot end, and the motherboard -- I think I can find or fabricate most other things that are likely to break or wear out. Now that I realize that I can fabricate the block pretty easily, that makes me feel better about any eventual rebuild of the hot end I may need to do.

Steve

 

[3poro 9]

Skimming through temperature.cpp, if I understand it correctly, you'll only get that error if the thermocouple is either shorted or open. How and where is the new thermocouple joined to the old cable harness wires?

 

The Pt100 is open. My guess is - its temperature range was not as broad as promised and it gave up. The cabling is OK.

Thank you to gr5 also for the hints regarding disassembling the head; I was actually planning to use my soldering iron for heating up the nozzle unit in order to get the hollow "crown" bolt out. After that it will be a lot easier for me to get the long brass bolt out.

I'm located within the EU. You are probably right about Ultimaker's limitations in receiving payments - I have faced the same rigidity with filament orders. The situation just really leaves me speechless - I have ordered stuff from Ultimaker worth of a bit over 5000 USD. The result - misinformation, seriously delayed deliveries, confusions - is something I have never experienced with other sellers. Only in 2014 and only through Amazon I have placed 41 orders, so I think I have some experience.

 

[3poro 9]

Regarding paying UM for parts - I don't think they have a way of accepting payment other than for the stuff in their store. So they usually send you free parts even long after warranty is over and even if you admitted that it was your own fault that it broke. I think they are interviewing for more tech support positions but that takes time also.

 

Yesterday I received the rest of the parts I was in need of - specifically, I needed a new PTFE part as the original one had gotten slightly malformed. I believe the malformation was because of overheating of the nozzle unit during the meltdown.

I also ordered some spares for the spare parts in an attempt to avoid such a lengthy downtime with my printer. It turned out to be possible to pay for the parts, although not online. IMHO the PTFE part should have been covered by the warranty, but I don't mind too much - the prices of the spares were mostly within some reason.

[3poro 9]

My Pt100 gave up again over a month ago. I've been too busy with other projects to start investigating the problem further, but now I took the time for it. I opened the failed Pt100 (my 2nd failed sensor) and I'm not sure if it is actually suitable for temperatures occasionally exceeding 260C. The problem I had was with the solder which seemed to have melted. What kind of solder is used for the bond between the actual Pt100 sensor and the wires? Also, there is some insulation tape around the Pt100 sensor, inside the metallic (outer) sleeve of the temperature sensor head - what material is it?

[3poro 9]

OK, I figure the detailed specifications for the Ultimaker's original temperature sensor are not available. I don't think it's good for 260C and the reference to 400C is a joke of some sort. I made my own temperature sensor out of these parts:

 

• http://www.tme.eu/en/details/pt106054/resistive-temperature-sensors/
  
• http://octopart.com/mm01006-multicore-1013894
  
• Kapton tape (as insulator for electrical connections)
  
• cord from broken UM2 temperature sensor
  

 

After installing I don't have to rely on adhesion characteristics of the Kapton tape, so the solder is setting the maximum temperature - at 301C. Already by now this self-made sensor has lasted longer than the two Ultimaker original sensors.

 

[gr5 2,167]

Excellent! Well it's 93% lead so it isn't RoHS compliant and in the USA must be considered toxic lead-containing waste. I assume this is true in the European Union also. But this is a small price to pay for 301C melt point.

 

[bez 9]

I'm glad I found this topic. I just accidentally murdered my PT100 B, as it had seized into the block even after removing the fixing grub screw, I may have broken a connection as its still stuck in there now lol I just need to figure the best way of removing it in one piece ........ I suppose its heating the block up then worst case drilling it out :(

Is it as straight forward as just soldering the new sensor onto the existing wire and reinserting it into the block?

 

[3poro 9]

Is it as straight forward as just soldering the new sensor onto the existing wire and reinserting it into the block?

 

Yes. Also, if you use the sensor I listed (http://www.tme.eu/en/details/pt106054/resistive-temperature-sensors/), it will not get stuck to the nozzle block the same way as the original one.

 

[bez 9]

That's good to know, I have ordered both the items you listed so now all I have to do is wait.

I think it should make a very welcome upgrade to the stock part at a fraction of the price of a direct replacement

From what I could tell when i dissected the sensor I finally extracted from mine it looks like the material used to pack the stock sensor melted then glued its self to the brass, making it impossible to remove without damaging the wires in the process.

While I have the hot end apart I'm tempted to upgrade the cooling fans ......... bloody tinker lust lol

 

[wwiktorass 0]

Hello.

 

I have problem in my printer Ultimaker 2: "ERROR - STOPPED Temp sensor contact support@ultimaker.com". I changed temp sensor PT100 to new but printer still showing problem: "ERROR - STOPPED Temp sensor contact support@ultimaker.com" . How it remove?

 

Thanks,

Victor