External auqarium airpump for hotend cooling?

Hi,

the loudest component on my printer is still the hotend fan and when I recently looked into the berd air for model cooling, I thought that the same could be done for hotend cooling as well.

When I compare a 30x30x10 fan with a pump like this they have pretty much the same air throughput (60cfm vs 1,6lpm).

I also checked for adaptors for this on thingiverse and there some implementations of airpumps but no straight hotend collar that connects to a small airhose.
So have you seen a machine that uses an airpump for hotend cooling and what do you people think about this?

Cheers

Edited 1 time(s). Last edit at 03/23/2019 09:05PM by sungod3k.

If you want quiet, and lots of air, check this.

If you can't find a CPAP machine to take apart, you can make your own blower using a BLDC and printed impeller.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

You're right, noise reduction is not necessarily a strength of these pumps. I read differing results about these pumps. Although I can imagine that aquarium pumps are somewhat optimized for noise.

My primary concern is getting fresh air into a heated chamber and a pump like that seems like a cheap and practical way.
Hooking up a 10$ pump and some tubing with a simple pwm controller and a connector like this [a360.co] should be easy.

What type of filament do you print in a heated chamber and run a print cooling fan? I print ABS in a 50C chamber and there's no issue with the hot-end fan keeping the hot-end cool enough at that temperature. I don't use a print cooling fan for ABS at all.

If you bring outside air into the chamber, it will cool the chamber unless you pipe the air you brought in back out of the chamber.

You can get really quiet hot-end fans... look for Noctua...

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Mainly PLA. I have a cold garage, so I only want 30-40C inside the chamber. Which now that you say it like this makes external cooling unnecessary because its still below the glass transition temp of pla. Although the material could still expand at those temps and increase the resistance in the cold zone. ABS might be more forgiving there.

Yes the cold air would cool the chamber, but the bed heater should be able to cover the loss. Although its worth to thinking about getting the warmed up air out again.

I once tried a noctua fan. Although I didnt find any 30x30 ones. I use the duet smart effector and I have a torus shaped 360deg model cooler. A 40x40 fan didnt fit in there without a long funnel that reduced the airpressure too much.

My latest thinking is to combine model and hotendcooling into one external pump. I just havent decided how important controlling the airflow for model cooling is. Its nice to give an extra push for bridging and turn of modelcooling during the first layers but its probably not strictly necessary. Then one could just get two lines into the chamber fed constantly from outside.

On the rare occasions I print PLA, I set the chamber to 35C and just use the regular hot-end fan. I haven't had any problems with jamming, but I run a 3:1 geared extruder (BMG) with a decent sized motor (46 mm, IRIC) so it has no trouble pushing the filament even though the current is set pretty low to keep the motor running cool. If you're using an ungeared extruder and/or a pancake motor, and having problems with jamming, you might consider a change there.

The problem with using one air source is you'll need some way to divert some of the air from hot-end cooling to print cooling. If you don't use a diverter and just switch it on after the first pew print layers, if the print is large the heat may have a chance to creep up the hotend and cause a jam. If you leave it on all the time you may have problems keeping the print stuck to the bed.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

I have the bondtec QR so that should not be an issue.

Ill test bed adhesion with increased bed temp for the first layer instead of reduced fan.

Having one source and a Y split hose would a neat setup though

I use this pump to cool a single hotend
[www.amazon.de]
And this
[www.amazon.de]
For two hotends and part cooling at the same time
The second one is a bit loud for a living room, but if you place it a bit hidden it is silent enough.
I much prefer compressed air over fans for reliability and efficiency. I was able to design a super effective extruder that includes the cooling for the merlin.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

That looks good and meets the specs that I was thinking about.

Do you want to share a pic of you cooling setup?

You can see it in action here
[www.youtube.com]
The part cooler has changed in the mean time and is fully integrated into a printhead cover now that also manages all cabeling.
Compressed air makes it easy to point the part cooler to a very small area.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

This shows the inenr setup. The Merlin is screwed in via the M10 thread, you can see the air channel that goes through the base of the extruder and delivers the air to the hotend. The air exits at the very end of the PEEK part of the hotend, cooling exactly the most needed part. This enables the PEEK/PTFE hotend to print 260° easily.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Interesting so youre going to the model cooling first and the "rest" of the air is used for hotend cooling?

No, the heater block is in the way for this and has to be covered in silicone for this to work. The model is cooled by a seperate stream and has a solenoid so you can turn it off. I print ABS most of the time so i seldomly use model cooling it.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

I see, yeah im still debating if control over the model cooling air stream is necessary, but having split air stream that can be closed by a small controllable valve is imaginable.

With a solenoid you only have off and on, but for me that is sufficient.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Check it out. I just saw toms video about the mosquito hotend. Ive seen that before on the bondtec site and kind of dismissed it, but here they explain the concept. At 4:50 they also mention that the heatbreak is now so thin that it appearantly doesnt require a heatsink fan.

Im not 100% convinced because the heatsink is so small, but its an interesting concept. If this would work as advertised and you pay the high price, one could go for controlled modelcooling and split of a path for hotendcooling.

The Merlin is running in my oldest printer for >7 years now without any cooling and it is printing ABS. The M10 thread acts as cooling fins and the movement of the head is enough. I don't advertise this since it is not realy a robust solution, but it can work.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Yeah i watched some more vids about the mosquito and it delivers with a small fan by default.

What motivated you to build that setup? I dont see a heated chamber, or did you just notice that you dont need a fan?

The Merlin itself i developed out of neccesity, at the time there simply was no reliable hotend to be had.
My current printers are just one point in many years of RepRap evolution. In the video the housing was not applied yet, one of the reasons why i use cubic frames is the easy way to enclose it.

The attached image shows it how it looks printing right now.
One drawer has the RaspberryPi Screen behind it, the other is for tools etc.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

A system such as the Mosquito may well work very well for most prints but may have problems if heat is finding other routes past the heatbreak. If the print and the filament need frequent retraction this may take more heat past the heatbreak area. Another source of excess heating will occur in heated enclosures and this will be aggravated if the filament has already been warmed during its journey through the heated enclosure to the hotend.

A good point made by the video of the mosquito is that the heatsink size can be minimised with a good heatbreak - no oversized finned heatsinks with big fans needed and even a fairly small supply of air from an aquarium pump may be all that is required. Fans mounted to the hotend have the drawback in heated enclosures that they are drawing in hot air so not great for cooling.

On the cooling of the print I try to make a distinction between bulk cooling of the print which is needed for plastics such as PLA but causes problems with others (ABS) and the cooling of the print immediately adjacent to the nozzle. Plastic near the nozzle will loose the finer features in a fairly short time if the temperature of the plastic remains high for too long. Cooling the print near the nozzle, "near field cooling" will help to prevent fine detail loss with all plastics. I have been cooling with air from an aquarium pump for a while now with good results [www.youtube.com]

Mike

What materials did you print with that near field cooling?

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Quote
leadinglights
A system such as the Mosquito may well work very well for most prints but may have problems if heat is finding other routes past the heatbreak. If the print and the filament need frequent retraction this may take more heat past the heatbreak area. Another source of excess heating will occur in heated enclosures and this will be aggravated if the filament has already been warmed during its journey through the heated enclosure to the hotend.
Mike

I wondered about that as well. it is possible that in a heated chamber pla will start expanding a bit and therefore have more resistance in the coldzone of the hotend. altough with a proper extruder that effect could be mitigated.
i ordered a precision scale so i can actually measure commanded vs actual extrusion, so it will be very interesting that see if actual extrusion gets worse in a heated chamber.

As for bulkcooling, Im not sure if its good for pla. im using a non magnetic flexplate as the bed and i notice that even medium sized pla parts shrink enough that the plate curls up a bit. having 40-50C in a heated chamber could minimize that.
Also the air from the near field cooling probably creates a little cold bubble around the model anyway.

Your setup looks good though. Is the airstream from the pump for nearfield cooling controllable or constant?

Quote
Srek
What materials did you print with that near field cooling?

Those were all PLA.There was additional cooling from heatsink fan directed over the model.

Mike

Quote
sungod3k
............ Is the airstream from the pump for nearfield cooling controllable or constant?

I have not previously used any control but I am going to use switchable control on a "Berd" air-ring type cooler and a turbulent fan bulk air cooler on the printer I am building now.

Mike

Aha, what convinced you to go from no control to switchable?

Quote
sungod3k
Aha, what convinced you to go from no control to switchable?

Mostly bridges I guess - although simple idleness may account for me not having any control before. The latest printer has lots more bells and whistles so it seemed a good idea to put in what had been left out before.

Mike

Has someone ever tried using a bigger blower fan to pump air through a hose?

Common knowledge says that centrifugal fans have more pressure than axial fan, but how much more?

its a bit hard to find pressure data for the cheap blower fans. but here [www.alibaba.com] it says ca 350 pa
the berd air pump however says 80kpa thats one order of magnitude more...

The problem with fans of all kinds is that even at low pressure they start creating turbulences instead of airflow, basically the fan just turns in chaotic air and you get basically no pressure anymore. With radial fans you get a little bit more pressure than with axial, but far less than with any kind of pump. The order of magnitude you noticed is quite realistic.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Quote
Srek
The order of magnitude you noticed is quite realistic.

good to know.

Just remember that pressure is not equal to flow.

A fan might not be able to deal with a lot of back pressure, but in free air, it can generate a lot of air flow, which is what we need for cooling. A pump can generate a lot of pressure, but not always a lot of flow.

---

MBot3D Printer
MakerBot clone Kit from Amazon
Added heated bed.

Leadscrew self-built printer (in progress)
Duet Wifi, Precision Piezo parts