Jump to content
Ultimaker Community of 3D Printing Experts
DoctorBit

Heated bed temperature not reaching 100C

Recommended Posts

Hi All

I'm the happy owner of a second hand UM2 and I have a question concerning the temperature of the heating bed.

I'm setting up the bed temperature with octopi to 100C in order to print ABS and I'm having a hard time getting the abs to stick to the bed. But regardless of the ABS sticking or not I noticed that while temperature of the bed should be 100C it feels much less at the touch of the hand, like 60C.

100C should be quite hot almost burning at contact with my fingers and I'm not getting this temperature when I set it to 100C. (although octopi shows 100C at the graphical interface)

I don't have a thermometer at hand but maybe more experienced user can tell me right away if there is something wrong wiht my printer and/or how can I do to check it.

And by the way pirinting in PLA works just fine every single time.

Thanks in davance

Share this post


Link to post
Share on other sites

What country are you in? In USA you can get a very nice infrared thermometer gun that measures the glass temp accurately for about $10. I know we are spoiled here in the USA with pricing and the rest of the world things cost more. Not sure why.

There are so many things in your post I can help you with but no time right now...

If you are printing ABS and you want to get really good at it you need to cover your machine anyway so put a box over the top with an opening in the back for the feeder and tape it on there to get the air up to 30C to 40C (no hotter than 40C as the steppers might get overheated) and cover the front with some plastic wrap or something.

100C should result in actual temp of about 105C on the center of the glass and 95C towards the edges. A drop of water will not boil furiously but it won't last very long either - less than a minute. it should indeed be too hot to keep your finger on there for 10 seconds but maybe not too hot to keep your finger on the glass for 1 full second.

If your ABS prints are warping up off the bed there are many solutions that you need to do. first and foremost use the brim feature in Cura. Also bottom layer MUST be squished into the bed a bit - so maybe raise the bed slightly by turning the 3 leveling screws counter clockwise. Dont' re-level with the procedure - better to adjust adjust a little. And finally use some PVA glue. hairspray, wood glue, glue stick all have PVA glue in them - you want to water it down so it is so thin you can' t see it - that's why harispray works so well as it puts on the correct amount. Instead I use wood glue mixed with 10 parts water and paint it on with a paint brush and let dry. Also keep dust off your glass -- I typically need to clean my glass once per week with window cleaner.

Share this post


Link to post
Share on other sites

Hi Gr5! Thanks for your reply

 

What country are you in?  In USA you can get a very nice infrared thermometer gun that measures the glass temp accurately for about $10.  I know we are spoiled here in the USA with pricing and the rest of the world things cost more.  Not sure why.

 

I'm writing from Holland, I do have means for testing the temperature but not until ends of next week. I guess I will have to wait for that.

 

There are so many things in your post I can help you with but no time right now...

If you are printing ABS and you want to get really good at it you need to cover your machine anyway so put a box over the top with an opening in the back for the feeder and tape it on there to get the air up to 30C to 40C (no hotter than 40C as the steppers might get overheated) and cover the front with some plastic wrap or something.

 

Never heard of this method yet, I will have to try it :)

 

If your ABS prints are warping up off the bed there are many solutions that you need to do.  first and foremost use the brim feature in Cura.  Also bottom layer MUST be squished into the bed a bit - so maybe raise the bed slightly by turning the 3 leveling screws counter clockwise.  Dont' re-level with the procedure - better to adjust adjust a little.  And finally use some PVA glue.  hairspray, wood glue, glue stick all have PVA glue in them - you want to water it down so it is so thin you can' t see it - that's why harispray works so well as it puts on the correct amount.  Instead I use wood glue mixed with 10 parts water and paint it on with a paint brush and let dry.  Also keep dust off your glass -- I typically need to clean my glass once per week with window cleaner.

 

Thanks for all these tips :) I'm actually using gluestick but I guess that with my bed not reaching 100C is more difficult for the ABS to stick. More test to be done soon.

 

100C should result in actual temp of about 105C on the center of the glass and 95C towards the edges.  A drop of water will not boil furiously but it won't last very long either - less than a minute.  it should indeed be too hot to keep your finger on there for 10 seconds but maybe not too hot to keep your finger on the glass for 1 full second.

 

As soon as I get the thermometer I will try temps but I can tell you right now that it doesn't get near 100C at all.

So what I get from this is that my bed doesn't work properly. My next question is: How can I calibrate the temp bed? Can I? Is there a possibility that my bed is broken?

Share this post


Link to post
Share on other sites

There are 3 screws that attach the aluminum part of the bed to the steel base plate. The heads of these screws shoulld sit in holes that are "counter sunk". Sometimes UM doesn't countersink the holes enough such that the head sticks out of the hole slightly - this pushes the glass off the aluminum plate such that their is an air gap. This means the glass is MUCH cooler than the aluminum plate. This might be your problem. It's easy to countersink the holes with an inexpensive drill bit:

single+flute+counter+sink1280721358.jpg

  • Like 1

Share this post


Link to post
Share on other sites

Glass is black in the near and far infrared. That's why you get dew on glass - the glass cools below air temp. If it was transparent you would get dew on the dashboard. If it was highly reflective you would get no dew at all (you don't get dew on a sheet of aluminum foil left out).

It's true though that you want to aim the meter straight down. I did many IR tests on my bed recently and it's pretty damn accurate on glass. The edges are about 5C cooler than the center.

Share this post


Link to post
Share on other sites

If the temp is indeed wrong it's probably something that opens up as you heat it up (I assume it reads 20C at room temperature - if not then there is a bad connection). Anywhere you have a bad connection (wire to terminal block under printer, wire to terminal block at back of bed, solder connection from terminal block to bed, solder connection to little pt100) you get increased resistance which increases the temp reading which means the actual temp is lower (cooler glass).

Share this post


Link to post
Share on other sites

Also do not forget that the temp. sensor is measuring the bed not the glass, there w ill be delay in the glass reaching the indicated temp. On my printer when the temp. reading hits 60 the glass measures circa 45. Also there is a further delay between the centre of the bed hitting X and the rest of the bed hitting X - initial temp difference can hit 10; of course not an issue if the model has a small base.

Share this post


Link to post
Share on other sites

Hi All

Today I had the chance to check the temperature very quickly and the results are bed temp at arround 35C when I set it to 100C. So definitely something is not working here.

Gr5:

There are 3 screws that attach the aluminum part of the bed to the steel base plate. The heads of these screws shoulld sit in holes that are "counter sunk".

I've checked this and there is one screw that will need a "touch" to sit flush, the rest are ok. The problem unfortunately seems to be the bed not reaching the right temp.

Irobert:

Also check all the wire connections so it's not something simple as a wonky wire causing bad temperature readings.

Cable seem ok under visual inspection. I guess I have to take the hole thing apart to check if I'm getting the right signals at the main board.

Just for the record, I'm using octopi. It shows Celcius on screen and all my setup is also on celcius... Could that be that octopi is sending Kelvins instead of Celcius? Or Ultimaker receiving Kelvins instead of Celcius? (not sure how he comunication between both works).

that could be a valid reason as 100F are exactly 37,7 C

Anyone has a better bet on where the problem sits?

Share this post


Link to post
Share on other sites

Update:

After checking and in contact with UM2 Support this is what I've discovered:

First of all the tempreature sensor in the bed plate reads ok at room tempretaure but when it is heated it "jumps" at about 80C. So suddently at 80C readings jumps to about 150C wich it is completely wrong.

So UM2 thinks that the temperature is much higher than it really is and stops heating the bed with the result of the bed cooling down when it should continue to heat, up to about 100C.

Sadly my warranty did expire as it is a 2nd hand UM2 so I will now attempt to replace the TP100 sensor. As stated in other threads the replacememt from farnell is:

http://nl.farnell.com/ist-innovative-sensor-technology/p0k1-0805-2p-b/sensor-pt100-0805-class-b-0-04/dp/1266934?CMP=i-bf9f-00001000

Hope that will fix the problem as the heated bead is about 150+€ and I can't afford that.

Wish me luck

Edited by Guest

Share this post


Link to post
Share on other sites

It's unlikely the sensor. When things heat up the expand (think "move"). This movement can cause wires to no longer be in contact.

The most likely cause of failure is the terminal block on the actual heated bed - I would take it all apart and reflow the solder on all 4 of those connectors.

The next most likely problem is the wiring - you can simply loosen the 2 screws that hold the temp sensor wires (skinnier wires) and then re-insert and retighten the wires.

Third most likely problem is the terminal block underneath the printer.

Least likely problem is the temp sensor itself. You could heat it up and then when temp jumps cut power and quickly remove temp sensor wires form under printer and then measure them and measure the resistance right on the temp sensor itself. Repeat for different parts of the path. Also push and flex the bed a bit while you measure to see if it's intermittent trace (sometimes the trace has a hairline crack).

Share this post


Link to post
Share on other sites

I'm just working on it right now.

What I tried first was to resolder the terminal block on the bed and also the temps sensor. After that, the temp sensor just stopped giving any reading. Completely Dead. I did order yesterday a replacement so I removed the old one and soldered the new one. And all started working again. This part is now fixed.

Now I'm facing another problem. The readings seem to be out. Tempreature showed in the UM2 seems lower thant the real temp. The higher it goes the bigger the error.

I ordered this part: http://nl.farnell.com/...?CMP=i-bf9f-00001000 that I found somewhere else in this forum. Do you guys know if this is the right part?

Is there a way of recalibrating the sensor?

Share this post


Link to post
Share on other sites

Your link doesn't work but all PT100 sensors should have the exact same temperature versus resistance curve. Unlike thermocouples which every manufacturer has different specs for.

Calibration is trivial anyway - I always print now with bed at 60C - sensor is on the edge so the center of the glass is actually 65C (edges always cooler). This measured with an IR temp sensor which works very well on glass (glass is pretty opaque/black in IR temps).

So just figure out what temp you have to set the printer to to get 65C on the center of the glass and make that your bed temp for PLA. Repeat for any other materials.

Share this post


Link to post
Share on other sites

Your link doesn't work but all PT100 sensors should have the exact same temperature versus resistance curve.  Unlike thermocouples which every manufacturer has different specs for.

Calibration is trivial anyway - I always print now with bed at 60C - sensor is on the edge so the center of the glass is actually 65C (edges always cooler).  This measured with an IR temp sensor which works very well on glass (glass is pretty opaque/black in IR temps).

So just figure out what temp you have to set the printer to to get 65C on the center of the glass and make that your bed temp for PLA.  Repeat for any other materials.

 

Indeed! You are right. I thoug there was a mismatch of temperature but I was actually mesuring the center of the print bed (without the galss). When mesuring the temperature at the sensor the readings where completely correct.

So I have a fixed bed!

In case anybody is interested I'm explaining all about the process of fixing the heating bed of the Ultimaker 2 in my blog.

Thank you all for your help.

Edited by Guest

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×

Important Information

Terms of Use