No Klipper problem: Maximum CAN bus toolheads temperatures regarding heated chambers Question? Poll?

Basic Information:

Printer Model:
MCU / Printerboard:
klippy.log

maz0r has a good overview concerning CAN bus toolheads https://github.com/maz0r/klipper_canbus/blob/main/index.md
This question, poll, or whatever you want to name my post is pushed by https://klipper.discourse.group/t/closed-printer-not-extruding/.
Sineos and Arakon started my worries. Closed printer not extruding - #2 by Sineos, Closed printer not extruding - #3 by Arakon
Stupid me, I never thought about this problem. I found two places in the www, where people asked this question. No good answers at all. What is max operating temperature? · Issue #16 · bigtreetech/EBB · GitHub, Reddit - Dive into anything

CAN bus toolheads don’t use mil specs (125°C) regarding the placed ICs.

example

All CAN bus toolheads µCs from maz0r’s listing use “Temperature range = -40 to 85°C (105°C junction)” and similar temperature ranges regarding the other ICs.

My question, poll is: Would you share your experiences regarding CAN bus toolheads used in a heated chamber?

What is your maximum heated chamber temperature? Any problems regarding temperature?

Thanks a lot, hcet14

I’ve wondered about this myself and, as I have been experimenting with a heated chamber, I’ve made a couple of observations and a bit of experimenting. The maximum temperature I’m running at is 60C, with 50C being the most used temperature. The toolhead controller is on top of its associated NEMA 17 stepper running 0.6A (0.2A hold current) with 6mm standoffs. For additional testing, I tried a 40mm fan above the toolhead controller 10mm away (high enough to clear the TMC2209 heatsink). There is no cover on the stepper or the toolhead controller.

At 20C (room temperature) the reported temperature of the toolhead controller is 45C to 50C. With the fan running, the temperature drops to around 35C - this is measured after running for about an hour to get a saturated temperature.

When the chamber temperature is 60C with the same conditions, the reported temperature is 80C to 85C. At 50C ambient the reported temperature is 80C to 82C. Using the 40mm 5V fan the reported temperature drops to around 75C in both conditions.

Regardless of the ambient temperature and whether or not the fan is present, I haven’t had any problems.

When you talk about temperature ratings for chips, you should note that it’s not a change in silicon but a change in testing. A chip rated for industrial/military temperatures is the same one that is sold as a commercial temperature grade but tested at higher standards and maybe placed in a different package.

Based on what I’ve seen if the toolhead controller is in the open air, you shouldn’t have any problems. I can’t comment on what happens if it is underneath a cover (to make things look nice without a bunch of wires going every which way).

1 Like

I have recently had problems related to temperature and a closed chamber.

It began to print well, but not when it had not finished the first layer stopped extruding, that with the closed Chambers.

However, the same gcode with open chamber prints well. It must be said that these problems I was having in for 1 month, when it is hotter where I live, is to facilitate that it reaches 35 degrees …

After many tests, the one that at the moment is not working well for me, I know that it is not ideal, is to lower the temperature of the bed about 25 degrees7, with this I get that the hotend does not heat up so much in

Was it really a temperature issue with the CAN bus toolhead or with the filament/hotend?

I thought you were with Canbus Toolhead… But after many tests it seems a heatcreep problem, being summer here, 35 degrees of ambient temperature

Therefore I was asking.
Here the OP is asking for/considering safe environment temps for the CANbus toolheads to still reliably work.
But those enclosure temperatures might already be too much for particular filaments or even hotends causing maybe similar issues. This has to be excluded here.
Especially when printing PLA in an enclosed build volume especially in the summer when the room temperature is already quite high it can lead to heat creep. That is not the toolheads or feeders fault.

I haven’t seen any evidence of heat creep with what I’m doing.

My hotend is a Micro Swiss NG which has some additional internal “diverters” added to prevent extruder cooling air from moving down the shroud to the heat block and nozzle:

This was originally done to eliminate cooling air from passing down and providing unwanted cooling for ABS/PETg materials.

Thank you very much for sharing the results of your interesting experiments. That’s exactly what I expected, when opening this thread.

How warm or hot gets your NEMA 17? Could you share a picture how the toolhead is mounted on the NEMA? I guess it looks a bit like this

I know, the tests are probably standardised in some JEDEC standard. That’s the reason why those ICs with higher temerature range are more expensive.
And that’s why I started the question/poll, we use 85°C.

Who wants it nice? The print head with direct extruder (plus everything else you need) shouldn’t be a heavy weight :wink:

Thanks again.

Sorry about that - here’s a picture of the EBB42 mounted on the NEMA 17:

I haven’t measured my stepper motor temperature as stepper motors are somewhat unusual as their maximum temperature isn’t often specified, but their ambient temperatures and RISE in temperature are often specified.

The typical rise in temperature specified is 80C with a maximum ambient temperature of 50C. LDO has a “high temperature” motor that runs at 180C ambient plus rise temperature.

In a 3D printer application, I don’t think you really need to worry about the steppers overheating.

Be careful: This rating is the stepper’s core / winding temperature targeted at the electrical insulation properties. You can expect that the casing temperature is at least 30 °C to 50°C below its core temperature. The max allowable rise temperature is even lower.

For example NEMA Insulation Class H:

  • Max insulation temperature: 180 °C
  • Max rise temperature (at temperature equilibrium): 125 °C
  • Max operating temperature: Max rise + 40 °C = 165 °C

Edit:
Your example above with a max rise temperature of 80 °C would be a Class B motor with a max insulation temperature of 130 °C

Thank you for the corrected information.

I’m surprised at the difference between the core and the casing temperature - that seems high based on my experience with silicon chips; it’s usually a 2C to 5C (the difference junction temperature and case temperature). I would think that when the motors are saturated (ie running for an hour or so) the core and case temperatures would be a lot closer.

Regardless, when I get some time I’ll measure the stepper case temperatures in the chamber when they’ve been running for a while.

The heat source are the windings. There are a lot and they are tightly packed and then you have air gaps between rotor and stator etc.

You can compare it to your hotend: The inner heating cartridge temperature is much higher than the temperature you measure at the temp probe’s location. And it will stay higher even if you reached the PID equilibrium.

Edit:

2 Likes

Sorry for getting back so late. Thank you @mykepredko & @Sineos for your replies. Talking about mykepredko’s extruder stepper.

Doesn’t make real sense. I assume it’s a ~1.7" (~43mm) stepper. Next assumption, it’s a 1.0 S.F. motor according to

for a Class B insulated, 1.0 S.F. motor, if we add the NEMA allowable rise of 80°C to the reference ambient temperature (~50°C), results in the motor having an operating temperature of (80+50)=130°C. Are you with me?

I think

ebb42_on_top_of_stepper

this is not a good place for the EBB42. It’s sitting right above the hot stepper.

I’m anxious to your measurements.

Thanks a lot.

1 Like

[quote=“hcet14, post:13, topic:9940, full:true”]
Sorry for getting back so late. Thank you @mykepredko & @Sineos for your replies. Talking about mykepredko’s extruder stepper.

Where is a better place to put it? As I indicated before, I did put a 5V 40mm fan above it to cool the MCU and TMC2209 on the EBB42 with something of an improvement.

This is not a facetious question; I’d love to find a better place for the EBB42 so that I don’t have the rat’s nest of wires but, at the same time, it does need some airflow.

NEMA standard typically assumes 40 °C ambient temp:

Also see:

1 Like

Of course. Talking about mykepredko’s extruder stepper

Oooups, I wanted to write "if we add the NEMA allowable rise of 80°C to the ambient temperature (~50°C)!

That’s why I choose 50°C. Wrong?

I don’t think it’s wrong - I’m just not sure if it’s optimal.

As I understand it, the reason for having a heated chamber is to reduce the internal stresses of the part during printing. I haven’t seen an authoritative reference indicating what is the best chamber temperature for different materials.

On the point of best chamber temperature, I’m trying to figure out how to get an even temperature throughout the chamber (as well as how to prove that I have an even temperature). Right now, I’m blowing heated air onto the (stationary) printing bed with an exhaust at the top, which seems to be how most chambers are configured, but when I measure the temperature on the Plexiglas walls, I see some pretty significant temperature variations - I would think that if I have an even temperature inside the chamber, the six walls of the chamber would all be at the same temperature.

I would choose the coolest place.


Sorry, I don’t really know how that thing looks like.

I’m not sure what you are referring to:

  • The maximum operating temperature of a stepper shall not exceed 40 °C + temperature rise
  • IMO the NEMA standard is unclear (or I missed it), if this limit is targeted at the inner coils or the housing → I would assume it is targeted at the coil, since the temperature rise is a function of the coil resistance (Chapter 4.3.3.3.1)
  • Now given that the housing is at least 30 °C degree cooler than the coils, this would mean that with such a motor the chamber temperature should not exceed 90 °C (40 + 80 - 30)

That’s the heat break/heat sink cooling fan. No space for the PCB there.