Measure difference between requested and actual temp of hotend and bed, the MacGyver method

[Indeed]

For those of us who don't have an IR thermometer or a thermocouple, or don't trust the one we have, an ordinary oven (grill, meat, frying) thermometer with a wired probe can be used to get an idea of the difference between set temp and actual temp. The oven thermometer will probably be off by a couple of degrees, but not much more than that. It should have a wired probe, or it may be difficult to see the temp display due to how the probe will be placed.


HOTEND
I have cross-checked the results of this method against a fairly accurate thermocouple, temporarily placed within the hot end of the heatbreak.
They did differ, but no more than a couple of degrees (three at most) at temperatures 20, 40, 60 and so on, all the way up to 260 celsius.
Maybe my ten dollar oven thermometer is exceptional, maybe it's average. In any case, the results from a such a thermometer should be good enough to reveal any huge temperature offset.

1. Cold pull the filament to remove all plastic from the hotend.
   Open the extruder lever, then start the hotend when it's cold and watch the temp rise.
   When it has reached maybe 80 celsius, grab the filament and do a short and firm pull, but not firm enough to break the filament.
   If it didn't come loose, pull again at 100, and again at 120 and so on until it comes loose.
   If you are lucky you'll be able to clear out all filament from nozzle and heatbreak.
   After the pull you can disable the hotend heater if you like.
   
   
2. Remove the nozzle.
   
   
3. Push the tip of the thermometer probe into the heater block where the nozzle was, until the probe makes contact with the heatbreak.
   
   If the probe is pointy, make sure the tip is centered and inside the heatbreak filament hole.
   Arrange for the probe to be in good contact with the heatbreak while making measurements.
   Maybe put a folded towel, or a sponge, on the bed and then lower the gantry until the probe is held firmly against the heatbreak.
   
   Make sure the the Z motors are off (gcode M84) and lower the gantry by hand.
   If you use the Z motors you may overshoot and bend the temp probe or damage the heatbreak.
   
   To get more accurate results you can also wrap a small isolating patch around the probe where it enters the nozzle hole.
   Perhaps a small piece of rockwool, glass wool, some kapton tape, or maybe even a small piece of ordinary cloth.
   The keyword here is "small", because if you use cloth or similar and you are going far above 200 celsius the cloth may catch fire. Be aware.
   
   Just in case, have pliers and a cup of water nearby and don't leave the hotend unattended.
   Be prepared to quickly remove the probe and then use the pliers to throw the burning cloth in the cup...
   
   
4. Now you can set different hotend temps and compare them to thermometer readings.
   When I did this it took quite a while for my oven thermometer to settle, many minutes.
   I waited for at least 10 minutes, and sometimes more, after the hotend temp sensor had settled to make sure the oven thermometer got enough time to catch up.

BED

I have cross-checked the results of this method against a fairly accurate thermocouple, placed just beside the oven thermometer probe.
They did differ, but no more than two degrees at temperatures 20, 40, 60 and so on, all the way up to 110 celsius.

Place the probe on the bed, with the tip near the bed center, under a small patch of isolating material.
Don't cover up the whole bed. Use something big enough to cover a couple of centimeters around the tip of the probe, but not much more.
Just make sure the sides of the probe are also covered enough to keep air from circulating.
Maybe use a sock folded twice or more, or a piece of cork with a groove for the probe.

Because we don't want to completely remove the usual heat transport from the bed to the environment.
Many of us have the bed temp sensor below the bed and if there's a thick towel over the entire bed would we wouldn't see the temp difference that usually exists between the bed sensor and the top of the bed.

Now you can set different bed temps and compare them to thermometer readings.
The oven thermometer is quite slow (at least the one I have), and so is the bed.
So wait for at least 10 minutes after the bed has reached the set temp and the average bed heater load has stabilized.
In Mainsail you can see the average load when you hover over the bed "State" reading.
Or just wait until the bed temp has stabilized and then for another 20 minutes before you take a reading.



PEACE OF MIND

I used these methods to measure my own hotend and bed.

Now I know that my hotend is always a few degrees hotter than the printer think it is, with a max error of 4 degrees at 200 celsius.
The error is less than that below and above 200 celsius.

I also know that the bed temp is around 58 when set to 60, almost spot on!
At higher temps the error gets worse, at 100 I get 94. Not perfect, but good enough.

Nice to know, if nothing else.
I could try and compensate, but I won't bother. It's good enough for me.


Note: If you are less than 100 years old and have no idea what a "MacGyver method" is, just search for "macgyver household items" on google.

 

[Yonkiman]

Great tips! I've done similar testing. Waiting 20 minutes (or longer, at least until the measured temp completely settles) is key.

I found that for a V6 style nozzle, I was able to loosen the nozzle, wedge the thermocouple tip between the bottom of the heater block and the top lip of the nozzle (the top of the hex head) and lightly tighten to get a decent mechanical connection (without crushing the thermocouple). I was also generous with the thermal grease to maximize the heat transfer to the thermocouple.

For the bed, I applied thermal grease to the thermocouple and the section of the bed where the probe was. I put a small section of rubber over the thermocouple, and then a large socket (about 200g) on top of the rubber to hold it all in place and keep the tip of the thermocouple pressed into the print surface.

Like you, I found that the hotend temp was fairly close. However my bed temp was about 10° off at 60° (Klipper said 60° but it was really around 50° - I have to set it to 70° to get 60° at the top of the bed), and that at higher temps there was around a 20° difference (I'll set it to 110° to get 90° actual). My experience with bed adhesion seems to back up these results.