Cooling the Hot End

Going back to my Spare Parts Printer, I was really surprised at the anemic fans put on a Creality CR-3/CR-10 hot end (I was using a genuine cooling shroud originally) and I wanted to really understand how necessary a high flow fan is in regards to keeping the overall heat sink temperature down (with a focus on the top of the heat sink).

So, I created a simple mount for the Creality CR-3/CR-10 hot end with a small (30mm) fan to test its operation. I put the fan on a PWM control rather than on the basic 24V always on so I could control it on or off. The mount is printed in ABS and has a 40mm blower installed because of the other side of the hot end was open to minimize printing time and I expected that in a printer, this side would have a nozzle cooling fan/blower.

I should point out that I’ve been amazed at people who have designed 3D printed hot end holders and used materials like ABS and PETG. It honestly didn’t seem reasonable to me as I would expect things to get hot enough to melt the mount.

The CR-3/CR-10 has a Mk 8 nozzle, a 40W cartridge (with 100k NTC thermistor) and I took out the interior PTFE tube and the M10 PTFE tube lock. I did leave on the silicone rubber insulator with the idea that if heat was going to escape, I wanted it to go up towards the heat sink, rather than out into the open air. This is really the worst case scenario as I continually read that the filament inside the PTFE tube inside the heat sink absorbs heat within the heat sink and passes it to the heater block/nozzle.

Too measure the temperature, I used my 64x64 sensor thermal imaging camera and tried to get the target point (the cross hairs in the middle of the image) on the heat sink itself.

With everything at ambient, the thermal view looks like:


No surprises, everything is at roughly the same temperature.

I next set the heater to 100C and waited for things to warm up.

After one minute:


After five minutes:


After seven minutes:


In this series of images, you can very clearly see the heat sink temperature rising up into the heat sink (as would be expected).

Next, I turned on the fan while keeping the nozzle temperature at 100C.

After 2 minutes:


After 7 minutes:


After 20 minutes (I got a call that interrupted me):


Note that the heat sink temperature drops over time while the maximum temperature (which is noted with the square box) remains constant.

I then ran the nozzle temperature up to 150C and checked the heat sink temperature.

After 5 minutes:


and after 10 minutes:


There is a bit of a rise in the heat sink temperature but its much lower than the worst case where the fan was off.

Next, was at 200C.

After 2 Minutes:


After 5 Minutes:


You can see that there’s definite heat creep up the heat sink. When I put my finger to the heat sink itself, it felt cool to the touch, it seems like the heat was coming from the central pillar of the heatsink and does not extend out to the fins.

Finally, I put the temperature up to 250C.

At 5 minutes:


I ended the experiment here with the nozzle heater set to 0C and the fan left running.

After a couple of minutes:


A few things I would do if I were to repeat this experiment:

  1. Mount the thermal camera on a tripod so I am always looking at the same spot on the heat sink.
  2. Use a thermal touch sensor to validate what my uncalibrated fingers are telling me about the temperature of the fins.
  3. Repeat the experiment with one of my 3D printers with a commercial heat sink fan/duct work.

Regardless, I think this was a useful experiment and I may want to repeat it with filament running through the hot end.

My conclusions from this experiment are:

  1. A fan is necessary to keep the hot end’s heat sink at a reasonable temperature.
    1.1. I should probably move the fan downwards by 5mm or so - most of the air coming from the top of the fan’s annulus is blowing against the top of the heat sink and not through the fins lower down on the body of the heat sink - I suspect that much of the airflow is wasted and could provide more efficient cooling.
  2. The heat sink fan can be very modest.
    2.1. I should point out that the heat sink had 5mm of free space around it with nothing impeding or directing the airflow.
  3. The heat sink temperature measurement/image from my thermal camera seems to be dominated by the central pillar of the heat sink.
    3.1. I’m noting this because it’s important to keep the temperature of the filament in the PTFE tube in the heat sink at a reasonable temperature so the hot end doesn’t get clogged.
  4. It’s very reasonable to design a hot end holder out of ABS/PETG and maybe even PLA as long as there is a fan ensuring that it keeps cool.
  5. After printing, keep the heat sink fan running until the hot end is under 50C. This is to minimize problems due to heat creep.
    5.1. Looking at the heat progression with the fan off, I would think that there could be an issue here if power was shut off with a hot end at 200C or more - but with making sure the temperature goes below the melting point of any filament, that shouldn’t be an issue.

Comments from anybody?

I wonder how your experiment would fair with different heat breaks.

Originally i thought heat sinks like the v6 were optimal for heat dissipitation, but when compared to say a mosquito heat break, or copperhead, im more inclined to think the heat break having a thin termination between the heater and hearsink plays a bigger role in controling heat creep.

With your inital findings being as they are, do you think your going to explore this rabbit hole?

Right now, I’m really not looking to go any further research/experimentation unless there’s a compelling reason to. The question I had really was understanding the hot end heat profile (to validate what I have been reading about) along with its cooling requirements. I think that’s pretty much answered.

If I do anything, I’ll modify my hot end mount to move the fan down 5mm or so to ensure as much air from it flows over the heat sink’s fins - I might go back and repeat the experiment to see if that changes anything.

As for heat creep, I’ve never really seen it and with the results above, I’m not sure I will. After I post this, I’m going to add the conclusion “After printing, keep the heat sink fan running until the hot end is under 50C”. When I have had problems with clogging or uneven flow, I’ve always found that replacing the Nozzle fixes the problem - I’ve had one case with my M200 where the liner of the heat break disintegrated but that was after more than five years of use and I have no reason to suspect that was due to heat creep.

I would only consider looking at different heat breaks if I thought that I would get a significant improvement in print quality and I think there’s more that I can do with calibrating my printers and slicer(s) than fooling around with heat breaks.

1 Like

That was well done! To answer your question about the temperature of the actual radiator, I attached a temperature probe to the front of my Ender radiator, about halfway up (the prob itself is about 2cm long) and did a print at around 250c. I wished that I wrote it all down as I forget now, but I believe that it maxed out at about 32c shortly after starting the print and did not rise further even after an hour of printing.

I also tried it with a print at 210c, and it was lower, in the twenties. I also tried attaching the probe behind the radiator, in case the fan blowing on the probe affected the results, but almost exactly the same.results.

At the time I had just replaced the stock fan that was making noises with an equivalent sized double bearing fan that seemed to be a little faster, and was thinking of modifying it and slowing it down (because of the noise) but I did not know what was happing inside the hot end, my worry was that maybe the radiator stayed cool because it was not very good at pulling away the heat.

It seems like you have proved that wrong and the hot end fan can, in fact, be quite modest and does not have to be a monster to keep us from suffering heat creep.

Thank you for sharing your experiment!

1 Like

You’re welcome - thank you for the kind words and sharing the results of your own measurements.

I thought about putting a thermistor on the heat sink but since I had the temperature probe, I thought it would be more interesting to make a mount that exposed the hot end assembly so that I could look at the temperature distribution on the heat sink.

If I were to do this again (and I probably will), along with moving the fan down a few mm, I would probably put it behind the heat sink so that the maximum fin area would come into contact with the air from the fan.

1 Like