So. I’m guessing the chamber/exhaust temp sensor is on the back wall, where it’s getting influenced by the host/mcu temps. The chamber is DEFINITELY not 41+C, but the back wall is definitely warmer. Seems like that sensor needs to be relocated.
My Zero is in a 29 C room. The Zero hasn’t printed in a few hours. The printer’s door has been open all day. I placed a thermometer in the center of the Zero’s chamber and it read 29 C. Mainsail is reporting the chamber temperature (Exhaust Fan) as 31.3 C. That’s a bit higher than the actual chamber temperature but definitely close enough for regulating the chamber temperature. If you want a more accurate control of chamber temperature, maybe set the exhaust fan activation temperature 2 C higher, but there’s a 2 C hysteresis between turning the chamber fan on and off.
Was your Zero powered on for any length of time prior to making the temp measurements? @msheldon “idle” temps for all 3 circuit boards seem high to me. I just can’t come up with a logical reason all 3 would be high.
I was trying to convey the temperature data from my Zero for comparison but it wasn’t very useful without also including the other temperatures relative to the theory that the chamber temperature sensor reading is artificially high as a result of its proximity to the Zero’s electronics. Here is my Zero’s long term idle temperatures in a 28 C environment.
It had been an hour since the previous print, which was a tiny part, about 15 min print time. The temp eventually came down to within 2C of ambient, but it took a while.
I found where the temp probe is, and, IMO, not ideal. It’s on the back wall just above the filter. Prob OK-ish for when you have the exhaust opening blocked for high-temp filaments, not a great place when it’s open (PLA/PETG, etc) It’s also below the bed surface for the first 10-20mm of a print. So when the lower air vent is open, that spot is REALLY not getting good airflow. I’d think it would be better higher in the chamber, and on a side, not the back wall.
So, yesterday, because the print was small, and the bed finished right in front of that temp sensor, the residual heat from the bed was probably keeping the sensor high.
I was expecting the chamber temperature to be a useful metric for avoiding heat creep clogs and for consistent printing of warp prone filaments. I’m not finding the current arrangement very good for either. As @msheldon noted I’m also seeing chamber temps generally increasing with on time and very slow to respond to actual changes in chamber air temperature such as when I open the door, lid or run the exhaust fan.
I’m guessing a good place for the sensor might be near the air intake of the heat break fan. Probably simpler though to mount it near the chamber or auxiliary fans where one might run either fan gently to better sample air temperature rather than the temperature of whatever the sensor is mounted on.
When measuring air temperature, understand that whatever method you use you are NOT measuring the air temperature you are measuring the temperature of a sensor that is being heated and cooled by the air you are interested in the temperature of.
Sovol’s choice of sensor has a lot of thermal inertia. It is always going to be slow to respond to air temperature changes. Mounting it to the metal “wall” that is highly influenced by the heat generated by the power supply and mainboard was also a poor choice. Surface conduction is likely to skew the temperature of the sense element away from actual air temperature. Having the sensor at a level such that it as below, then above the build plate also is an interesting choice.
Easy improvement - Mount a 4010 turbo fan to the back wall near the floor of the chamber (so the source air is always from below the plate) with the outlet pointing at the temperature sensor. Getting the air moving will improve the sensor accuracy. Installing a duct to concentrate the air flow would further improve accuracy.
Better improvement - The chamber temperature sensor is handled by Klippper like a generic heater thermistor. Mount a glass bead (low mass) thermistor above the gantry away from the walls. Use materials with low thermal conductivity to fabricate a mount. (Optional) mount a fan blowing on the thermistor. Unplug the chamber temp wires and hook up new sensor. Double check printer.cfg to make sure the correct type thermistor is defined.
After watching it’s behaviour, I think the placement is worse than it looks.
At the beginning, while everything is heating up, and for the first 10-20mm or so of the print, the sensor is below the bed. This is just useless. There is no point at which anyone should care about the temperature below the bed.
For the next 20-40 mm, the sensor is alongside or just above the bed. Now it’s being very much influenced by the heat radiating off the bed, and is definitely going to be reading higher than it should. (observation agrees with this)
It’s only after printing 60-80 mm of the print that the sensor is in a fairly reasonable position to measure “chamber temperature”.
A circulation fan like you describe would definitely help this. I don’t think it would even need to run that fast, 20% might be plenty. Attach to “fan3?” A Bento Box filter? Might be nice to have charcoal filtration even when chamber temp is below target.
But ideally, like you mentioned, the sensor needs to be relocated too. I’d probably look for somewhere just above or below the y-axis rails. Close to where the print is, without being too close to the bed.
