Bed Temp not accurate

[adamfilip]

Im running into a issue where i set my bed temp to 80 and the display says 80. but when i measure the bed temp with two different thermometers/readers it says around 55, i have to compensate by setting it to 110 to hit close to 80

I have marlin set to use

#define TEMP_SENSOR_BED 1
// 1 is 100k thermistor - best choice for EPCOS 100k (4.7k pullup)

I assume this is what came with my printer

[Mjolinor]

Measure the resistance with a meter if you are unsure, it could be anything. They are very inaccurate usually because they are cheap, the thermistors and the measuring hardware / software. Most people just live with it and compensate as you do.

[Roxy]

Most people just live with it and compensate as you do.

Yes, but that is a very big 'compensation'. I could believe being off 5 degrees F at 100 F. But 110 to get to 80 seems wrong.

Adam, I believe that 100 K ohm number is at room temperature. You should be able to unplug the wires from the controller board and measure what it says. But the thing is, those parts are so cheap, you may as well order a few off of eBay so you can be certain you know the correct thermistor number to use.

I've broken off the leads on my thermistors on occasions. What I started doing was just ordering a few different ones off of eBay (and then clearly marking what Type # they are and put them away until I need them.) At least in my experience, it really doesn't matter for the stuff I do what type of thermistor I have in either the nozzle or the bed. I just adjust the Configuration.h file to be correct and I can use my printer again.

The one exception is I just broke the thermistor wires on my FolgerTech i3-2020 printer. The hole in the nozzle was bigger than any of the thermistors I had sitting around. I ended up wedging a little piece of stranded wire into the hole to force it to be in good contact with the nozzle. I probably should go order the right thermistor, but that seems to be working just fine.

[uncle_bob]

Hi

Measure the temperature right at the thermistor and I'd bet you get 80C +/- 5C. Measure it on the top of the glass and you will *always* get something colder. You have heat loss (thermal resistance) through the heater PCB and through the glass. A difference of 20 to 30 degrees is pretty typical. It's an issue with all similar heated beds so it's already "in the numbers" that you see posted. If you build a bed that actually *does* monitor surface temperature, you have to re-calibrate all the posted numbers so they match your (unusual) bed. Yes, I have a printer like that and going back and forth is a bit of a pain.

Bob

[kd7eir]

When my bed temp is set for 80°C, the surface temperature through the 5/16" borosilicate glass is 57.3 °C

I've never cared about the temperature of the surface - I know that setting my bed for 80° works for my needs.

[uncle_bob]

As long as everybody runs a heater pcb with the heat traces on the back side *and* the thermistor is on the back side *and* there is insulation below the board *and* the board is about 1.5 mm thick *and* there is an air gap between the board and glass *and* the glass is about 2 mm thick *and* the room is between 20 and 30C ....

Setting the bed temp to 80C will get you the same surface temp as everybody else gets. This all dates back to the original i3 design. It is common to about 90% of the printers out there. Unless you start looking at information from a MPSM or something else with an aluminum bed, your numbers will be the same as the other guy's numbers.

[longjohn119]

I used 3D Benchy to calibrate my bed temps for PLA. Basically you look at the lower part of the bow and it should a nice clean slope. If the bed temps are too high then it will warp or bow a little there. I came up with for first layer/rest of 65/60 for most PLA and 70/65 for eSun PLA+ although your's may be higher. I have a Pegasus 10 full metal with bed extenders which I modified to about 11 mm gap below the PCB and have 10 mm of cork (2 5mm pieces) trimmed so it lies flat, topped with a 1 mm layer of ceramic 'felt' insulation and the PCB heater sits on that and the glass and clips pushes it down firmly (But not too firmly) I also ditched the mechanical relay for an SSR with PID control which all together seemed to give me more consistent temperatures especially around the edges. I also have auto bed leveling with an 8mm inductive probe which works with just the PCB traces and hardware store glass but would be pushing it for 3 mm boro glass. Although those probes don't sense copper very well the PCB heater itself is a big inductor (technically a flat wirewound resistor with inherent inductance like all wirewound resistors) so I get about a 5mm sensing range rather than the 8 mm advertised. It works every time flawlessly and I haven't had to do any tweaking since I got it set up properly back in August.

[tsteever]

I used 3D Benchy to calibrate my bed temps for PLA. Basically you look at the lower part of the bow and it should a nice clean slope. If the bed temps are too high then it will warp or bow a little there. I came up with for first layer/rest of 65/60 for most PLA and 70/65 for eSun PLA+ although your's may be higher. I have a Pegasus 10 full metal with bed extenders which I modified to about 11 mm gap below the PCB and have 10 mm of cork (2 5mm pieces) trimmed so it lies flat, topped with a 1 mm layer of ceramic 'felt' insulation and the PCB heater sits on that and the glass and clips pushes it down firmly (But not too firmly) I also ditched the mechanical relay for an SSR with PID control which all together seemed to give me more consistent temperatures especially around the edges. I also have auto bed leveling with an 8mm inductive probe which works with just the PCB traces and hardware store glass but would be pushing it for 3 mm boro glass. Although those probes don't sense copper very well the PCB heater itself is a big inductor (technically a flat wirewound resistor with inherent inductance like all wirewound resistors) so I get about a 5mm sensing range rather than the 8 mm advertised. It works every time flawlessly and I haven't had to do any tweaking since I got it set up properly back in August.

Can you describe your probe setup a bit more in-depth? It may be best if it were it's own thread. A probe intrigues me but I thought you had to have a metal bed to make them work. My servo mounted probe is getting inconsistent.

[uncle_bob]

Can you describe your probe setup a bit more in-depth? It may be best if it were it's own thread. A probe intrigues me but I thought you had to have a metal bed to make them work. My servo mounted probe is getting inconsistent.

Hi

Servo mounted probes tend to be temperature sensitive. You need to be careful about their position relative to a heated bed before the probe process starts. They also can have issues with mechanical rigidity if you are not careful about how you design and build your setup.

The BLTouch probe is a mechanical probe that does not have any strange requirements in terms of bed surface.

Bob

[Roxy]

I know it is frustrating... But it really is helpful to turn on your bed and nozzle and let them stabilize for 5 or 10 minutes prior to starting a print. When I decide I want to print... I want to print. But if it is something I care about, waiting 5 or 10 minutes really helps!!!!