Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Liquid Cooling?
#1
Just throwing it out there, as I have seen many attempts, successes and failures, at cooling the extruder. Has anybody tried placing a waterblock heat sink on it? Obviously managing the tubes would be a hassle, but with a little trial and error, using an Arduino to monitor the outgoing fluid temp, you might be able to get very precise temps in very specific locations.
Reply
#2
It would be too complicated and expensive, so change the extruder at this price.

Alternatively to improve cooling of the extruder, adding the kapton around heater head (Piece brass color), add a heater around the tube hotend (Upper part, at the motor)
Reply
#3
Actually, changing out the extruder is just as complicated and just as expensive - if not more expensive. A new extruder means a totally new carriage in most cases.

Making a water cooling block for the thermal tube isn't that complicated if you have access to a small mill. In my view, the advantages of a water cooled thermal tube are two fold. Not only do you get better cooling, but you also totally eliminate fan induced air currents. You can get small radiators which take a 120mm fan for about $20. Small brushless 12v water pumps with reservoir for about the same. Add a few feet of 1/8" x 1/4" Silicone flex tubing, and that's about all you need once you have the water block. You could 3D print a mounting fixture to attach the fan/radiator to the back of the machine.
The silicone tubing is highly flexible, and can be easily bundled with your current wiring harness with little to no noticeable increase in stiffness. No temperature control is really needed. With the radiator and 120mm fan your water temperature will always be very close to the ambient room temp.

One big factor to getting better thermal tube cooling is to improve the thermal connection between the thermal tube and the aluminum block. The stock DaVinci has a loose fit between those two parts. Adding a little heatsink compound at this point significantly improves the cooling.

Insulating the nozzle block (brass on the stock machine) by wrapping it with something is only necessary if you are using a layer cooling fan. You are probably only using this if you are printing PLA. However, with improved thermal tube cooling, PLA printing becomes much easier.
Reply
#4
That's an interesting idea... something like this:

http://www.frozencpu.com/products/25143/...tl=g30c293

..could be very easily hacked to attach to the aluminum block in place of the tiny stock heatsink, I expect.

Now, regarding cooling in general, I expect we would need to determine what temperature the alumimum block should be at for the various extruder temperature settings and/or plastic types. I'm doing an experiment right now, for instance, with controlling the speed of my extruder cooling fan to see if I can reliably print ABS with the modified fan shroud; it is surprising that even at 50% duty cycle the extruder can barely make 230C, I think because the fan is blowing air rather efficiently across that tiny little aluminum heatsinc now and that on its own can cool the head too much. So far no jams or clicking though (although airflow is a big problem as it's keeping the ABS from sticking to the bed). Long story short, I expect this setup may need a lot of tuning to do right (resevoir size, thermostat to control resevoir temperature, etc...
Reply
#5
I don't believe that the absolute temperature of the aluminum mounting bar for the thermal tube is that critical. Temperature stability is probably more important. You want the temperature of the incoming filament to be pretty constant so that hotend is always raising it by a fixed amount. Even so, different flow-rates (when printing slower or faster) create varying heatloads. That's the job of the PID controller to compensate.
I can easily get 250C + with my cold end cooling shroud because I am getting little or no airflow around the hotend heater block (I have the fan pulling air out vs. blowing it in) That's also with silicone heatsink compound on every mating surface and all the air leaks taped over. I typically print ABS at 230C, so I think I have sufficient headroom.
The purpose of the necked-down heat break section on the thermal tube is to keep the upward heat loss to a minimum. However, it can't be eliminated. If you can't maintain temperature after eliminating airflow around the heater block, or by insulating it, then you need a more powerful heater cartridge.
One nice thing about water cooling is that it completely eliminates air currents inside the printer.

I think the goal is to keep the filament as cool as possible until the last possible moment - to sharpen then thermal transition zone - in order to promote smooth flow regardless of filament type. You can try to improve the insulation between the hot and cold areas, but you'll always have a heat leak in the thermal break area and within the filament itself. That leaves applying better cooling to the cold end, and probably a bit stronger heating to the hot end at the same time.
Reply
#6
I'm still waiting on a few parts, but here is a mockup of the new water-cooled extruder thermal tube mounting bar. I have two versions. The final parts will be milled out of aluminum. The first uses the stock hotend and thermal tube.









The second one is M6 threaded for the use of a PTFE lined thermal barrier tube from RobotDigg. I'm also using their Mk8 nozzle, which has a very nice finish, and extremely small land area at the tip which I think enhances performance on overhangs because it virtually eliminates flow attachment (Coanda effect). These sharp tips might not last as long, but at $2.50 each, who cares? I'm also using their heater block, 40 watt heater cartridge and thermistor. All these parts are dirt cheap - just a buck or two each.





I've checked routing of the silicone tubing and I think I've got it figured out. This stuff is very flexible and lightweight.
For cooling the water, I'm looking at several options. The cheapest is going to be a small brushless pump and 120mm radiator w/fan. Here are ebay links to the radiator I have (still waiting on the radiator).

If you wanted to go crazy with active cooling, this would be cool (ha!)

I just so happen to have one of these also - it's a spare chiller for my laser cutter. Full PID temp control from 2C to 45C with pump all in one unit. Probably overkill :lol:[Image: DavHotWater1.jpg][Image: DavHotWater2.jpg][Image: DavHotWater3.jpg][Image: DavHotWater4.jpg][Image: Mk8_Water1.jpg][Image: Mk8_Water2.jpg]
Reply
#7
This is pretty cool, although Ive personally never seen water cooling on a moving part.
Any concern around the tubing developing a weak spot from bending repeatedly?
Are you going to cover all the frame locations that would let a water mishap spray onto the board?
Im a big fan of over doing it, but I smell burning electronics when I read this post, but thats mostly coming from hearing about friends water cooling and frying their PC.
I do like the design, just dont agree with the statement that replacing the extruder completely is more work/cost. I guess if you already have the supplies on hand it would be more cost effectively than a hundred dollar e3d. Good luck with this!
Reply
#8
The silicone tuning is not going to fail. This is exactly the same tubing that I use in my laser cutter. It travels much longer distances and at much higher rates.
Unlike inside a PC, there is really no exposed electronics in the build chamber anyway.
Aside from machining the water cooling block, this is a simple project. There is nothing magic about the E3D. The parts that make it work well can be bought separately and assembled much less expensively.
Cost savings is not my main objective here. An E3D, nor any other air-cooled extruder is going to work for what i have planned anyway. I will also be water cooling the steppers and insulating the build chamber to reach my ultimate goal.
Reply
#9
I used to use medical grade silicone tubing for DIY paintball grenades, and could only re-use the tubing 3 or 4 times before it would develop a weak spot and a subsequent leak. The weak spots were much worse if I stored them somewhere hot like a car trunk, or were exposed to sunlight. Granted, I was blowing the tubing up with paint and that wouldnt happen with low pressure water. Looking inside the machine, from the front, I can see the motherboard in several locations, this was what I was referring to in regards to a water mishap.

Anyway, Im really curious to see how this turns out and wish you the best of luck. I have no idea what you are trying to print with this setup, but I cant wait to see!
Reply
#10
The objective is a heated build chamber. The controller board in the rear can be isolated with insulation, but the steppers can not. They will also need to be water cooled to prevent overheating. The build chamber will be heated to about 50C when needed (large ABS parts) and held there with a PID controller. Same technique used by Stratasys to alleviate ABS warping stresses.
Water cooling of 3D printers is not done often, but it has been tried and proven by several others. Here is one example.
Although I am not using an E3D hotend, there is no feature of it that can't be duplicated with readily available off-the-shelf parts. My water block will have different mechanical dimensions in order to be a direct replacement for the original DaVinci extruder mounting bar. The thermal barrier tube with PTFE liner is $3 part and already in-hand. The bushingless idler wheel of the stock extruder that can wear out prematurely (as pointed out by Keith) has already been addressed. New drive wheel of much better quality (another $3 part) is also enroute.

Original idler wheel:


Now with 2x 8x3x3 ball bearings from the tail rotor axle of a Trex 450. These originally ran at 8000 RPM under heavy load. I think they'll hold up fine for an idler wheel :lol:

[Image: nobearings.jpg][Image: withbearings.jpg]
Reply
#11
Where did you get those beehive-looking heatsinks?
Reply
#12
Quote:Where did you get those beehive-looking heatsinks?

Made on my lathe.
Reply
#13
Damn, was afraid you'd say that lol

There are a couple places that make stackable heatsink washers... might try those
Reply
#14
Quick update...

I'm still waiting on a few parts (Chinese New Year really slows things down). The aluminum water blocks are partially machined, but it's been difficult to get time on the mill to finish them.

In the meantime, I needed a way to mount the cooling pump and radiator on the back of the DaVinci. Decided to design a printed part for the job, so here it is:









I'm sure this part could be used by PC overclockers too...

Update: Link to part on Thingiverse. I just got around to uploading it.[Image: Rad-PumpMount1.jpg][Image: Rad-PumpMount4.jpg][Image: Rad-PumpMount3.jpg]
Reply
#15
Greg -

Slightly off topic, but you'd know the answer - What is the diameter of the hole in the OEM aluminum block?
Reply
#16
I'm guessing you're thinking of installing off-the-shelf hotend parts in the original aluminum block...
The hole in the original block is 8mm. I Iooked at the possibility of putting in an M6 Helicoil, but the 8mm hole is too big.
(M6 is the size of a standard thermal barrier tube).
You'd need a bushing that was 8mm OD, by 5mm ID and then run an M6 tap through it.

Not hard to make if you have a small lathe. Drill with #8 drill, tap, then cut to diameter and length. Try to make it a nice tight press fit.
I doubt you'd be able to find such a part off the shelf.
Reply
#17
I know machinists. Thanks!
Reply
#18
If you get some made, please post pictures. I'd think you could sell'em for a few bucks Wink
Reply
#19
it is something like this ?
http://www.memfast.co.uk/shop/vclose1.as...2250005134
Reply
#20
luc,

yes, but I think the problem is - you need a roughly 7mm clearance hole for an 8mm tap. We've got an 8mm clearance hole - that leaves no material for threads. You'd need a Helicoil with a 9mm OD (or larger) and 6mm ID. I haven't really looked for that. Maybe there is such a thing?
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)