Thermistor reading low
-
So I am setting up my Duet2 Wifi (1.04) with v2.02 firmware and have all the sensors working except for the hotend. I'm using a basic 3950 thermistor which was working fine before upgrading. I cannot get the Duet to use it to read the correct temperatures once the hotend starts heating up. For example the temperature will show 180°C, but measuring the hotend shows ~100°. I have been trying all kinds of values to see if I can get it to read any better at the hotter temperatures, but have failed thus far. I have tried at least one other temperature sensor with the same result.
Here is my current config.g setting:
M305 P1 T100000 B3950 R4700Just wondering if I am missing anything. Is there any way to get the raw values from the Duet to see what actual values it is seeing?
-
the generic B3950 +/- 1% thermistors are somewhat problematic.
the B3950 refers to the range of 25 to 50. They behave quite differently at temperatures around 200.
i had the same problem and i just replaced it with a pt1000 and its been very accurate.If you actually calculate the values for a B3950 from the datasheet (if you are lucky that you get one)
like https://www.tinytronics.nl/shop/index.php?route=product/attachmanager/getfile&product_attach_file_id=552
and use the 220C value you will actually get a Beta value of 4171.
And if you put in 25C, 50C and 220C you get β: 4666 K C : 1.072208e-7 -
I have tested the thermistor against the heat bed with fairly accurate readings up to ~65°C. Obviously, I'm trying to go much beyond that with the hotend, but its reading less than half.
Veti: Did you never get your thermistor to work before going to the PT100?
-
as i wrote from 25 to 50 the B3950 value will be accurate, as that is what you bought. 65is is still close to that range.
As i could not get a temperature table for my thermistor i just went with the pt1000.
did you try the values i posted?
-
@veti said in Thermistor reading low:
the generic B3950 +/- 1% thermistors are somewhat problematic.
the B3950 refers to the range of 25 to 50. They behave quite differently at temperatures around 200.
i had the same problem and i just replaced it with a pt1000 and its been very accurate.If you actually calculate the values for a B3950 from the datasheet (if you are lucky that you get one)
like https://www.tinytronics.nl/shop/index.php?route=product/attachmanager/getfile&product_attach_file_id=552
and use the 220C value you will actually get a Beta value of 4171.
And if you put in 25C, 50C and 220C you get β: 4666 K C : 1.072208e-7If you do have a temperature table then I suggest you use points such as 25C, 180C and 240C. That will give you greater accuracy at typical printing temperatures.
-
@Veti I did try your values, but the results were the same.
@dc42 Last night, to the best of my ability, I mapped out the resistance based on detected voltage and calculating for the resistance at that voltage using a multimeter and an external temperature gun. Though I wasn't able to get the sensor up past 220 when doing my tests so I used the values of 25, 190 & 210 to get:
M305 P1 T108000 B5615.05 R4700
Now I'm measuring within 5 degrees of what I would expect.
However, now I am running into the issue that the heater is not heating up fast enough. Less than 1.7C per second. I can't get it to finish a tuning cycle when setting it to 240°C. Oh bother...
-
is that really worth the hassle? just get a thermistor where you can get the temperature table.
like an ATC Semitec 104GT-2