Welcome! Log In Create A New Profile

Advanced

My 6 input (5+1) mixing hot end

Posted by deckingman 
My 6 input (5+1) mixing hot end
February 24, 2020 04:52PM
I've made a bit of progress designing his beast if anyone is interested. That is to say, I now have a prototype that I can start testing (almost). Details on my blog here [somei3deas.wordpress.com]


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
February 24, 2020 05:38PM
Quote

I don’t have a thermal imaging camera and my pension won’t stretch enough for me to buy one, so I’m trying to beg, borrow, or steal one.

You can buy NoIR cameras for the Raspberry Pi... basically an ordinary camera without the IR filter. I don't know how good they are at thermal imaging, but maybe worth a try? Maybe someone else here knows?

Some random thoughts...
  • I wonder about water-cooling... you could move all the heat-breaks closer together and get more uniform heating/cooling effects.
  • Why not put all the filament feeds in a circle, which would mean that they're all essentially the same (same angle, same distance), and closer together? But then you're getting towards the Diamond design.
  • If your 30W heater isn't good enough, (or even if it is) try insulating the outside of the heater block.
  • Tap some helical threads into the holes in your labyrinth to improve colour mixing.
  • If there is some significant amount of filament in your labyrinth, it not only impacts the amount of purging at colour changes, but also how far the colour changes need to be ahead of the nozzle position.
  • Wish I was a pensioner with a milling machine so I could play with ideas like this. smiling smiley
Re: My 6 input (5+1) mixing hot end
February 25, 2020 04:38AM
Quote
frankvdh
Quote

I don’t have a thermal imaging camera and my pension won’t stretch enough for me to buy one, so I’m trying to beg, borrow, or steal one.

You can buy NoIR cameras for the Raspberry Pi... basically an ordinary camera without the IR filter. I don't know how good they are at thermal imaging, but maybe worth a try? Maybe someone else here knows?

Some random thoughts...
  • I wonder about water-cooling... you could move all the heat-breaks closer together and get more uniform heating/cooling effects.
  • Why not put all the filament feeds in a circle, which would mean that they're all essentially the same (same angle, same distance), and closer together? But then you're getting towards the Diamond design.
  • If your 30W heater isn't good enough, (or even if it is) try insulating the outside of the heater block.
  • Tap some helical threads into the holes in your labyrinth to improve colour mixing.
  • If there is some significant amount of filament in your labyrinth, it not only impacts the amount of purging at colour changes, but also how far the colour changes need to be ahead of the nozzle position.
  • Wish I was a pensioner with a milling machine so I could play with ideas like this. smiling smiley

Thanks for the tip about the IR camera (I do have an RPi). I've been told that to get accurate thermal images, it's best to paint the object mat black - (barbecue or stove paint).

ref the random thoughts - I've had most of them too. Water cooling - I'd need to fit jackets around the heat sinks so I doubt very much if I could get them much closer. Besides, there is only about 3mm clearance between the hexagonal part of the heat sinks so I can't them much closer. Filament feed in a circle - tried that to start with and it doesn't work - well it sort of does and doesn't. It makes the hot end narrower in one direction but wider in the other. And it gets complicated when 5 of the inputs end up at one point but the 6th one has to end up elsewhere. Insulation is on the list of things to do - for reasons of part cooling fans blowing over the nozzle as much as anything else. Helical threaded holes - good idea - I like that - easy to do too!. Yes, keeping the volume of molten plastic to a minimum while still mixing it all together is one of the design challenges.

Finally, don't wish your life away - being of pensionable age has it's downsides too (as my recent surgery will testify). smiling smiley

Edit. Men have a gland that women don't have. The name of the gland is an 8 letter word beginning with P. For some reason, that word is banned so when I wrote "recent [nameOfGland] surgery" I had to take out the [nameOfGland]. I can't think why the use of that word could in any way be thought of as offensive.

Edited 2 time(s). Last edit at 02/25/2020 04:45AM by deckingman.


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
February 25, 2020 08:34AM
Some random thoughts.

I don't think that a NoIR camera will do thermal imaging, at least not in the under 200°C range as it is not sensitive to the longer-range IR at these temperatures - I have never tried it so I may be wrong though.

I use a RayTek Autopro non-contact IR thermometer which has a sensitive spot size of half an inch at 8 inches. It seems to be accurate on 8 pin DIP packages so should be able to get a useful measurement from your heatsinks. I think these are affordable but it is some years since I got it.

Watercooling does not need to take up much space. The water-cooled hotends I use have drilled channels and keep the temperature of the filament immediately before the thermal break to 18°C above ambient worst case (slow feed, lots of retraction). In the photo below the block is 18mm by 13mm by 30mm. Probably worth investigating for your next iteration.



You can get away with spelling Prоstate using the lower case o from the Cyrillic alphabet.

Mike
Re: My 6 input (5+1) mixing hot end
February 25, 2020 12:42PM
Quote
deckingman
[Water cooling - I'd need to fit jackets around the heat sinks so I doubt very much if I could get them much closer. Besides, there is only about 3mm clearance between the hexagonal part of the heat sinks so I can't them much closer.

I was thinking to have one water jacket around all the heatsinks. And that you should be able to reduce the heatsinks or even eliminate them entirely with watercooling.
Re: My 6 input (5+1) mixing hot end
February 25, 2020 01:04PM
Thanks for the ideas guys. I'll see how the air cooling works out because those small fans should be almost (to my ageing ears) silent if not efficient. And it's simple....


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
February 28, 2020 05:40AM
The next step of my journey along this particular road is now live on my blog for anyone who is interested [somei3deas.wordpress.com]


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
March 09, 2020 06:19PM
Part 3 is now live on my blog if anyone is interested [somei3deas.wordpress.com]


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
March 10, 2020 07:43AM
Thank you deckingman, your blog is always an interesting read with both data and discussion. I will continue to watch with interest and it may spur me into starting a blog myself on my humble efforts.

Mike.
Re: My 6 input (5+1) mixing hot end
March 12, 2020 04:29PM
Quote
leadinglights
Thank you deckingman, your blog is always an interesting read with both data and discussion. I will continue to watch with interest and it may spur me into starting a blog myself on my humble efforts.

Mike.

Well some say that my posts are too verbose but then again, there seem to be quite a few pedants about who pick up on the tiniest (missing) detail. Thanks for the kind words though.

I sometimes wonder why I bother because it's surprising just how much time it can take. But it's a useful way of forcing me to keep written records, and on more than I occasion I've referred back to my own blog to check something that I had forgotten (that happens more and more the older I get). smiling smiley

Part 4 is out now by the way [somei3deas.wordpress.com]


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
October 29, 2020 06:24AM
I'm now up to part 12 of this particular journey. But I've moved over to posting further updates on my YouTube channel instead of my blog.
Link here if anyone is interested [www.youtube.com]
(Parts 1 to 11 are still on my blog [somei3deas.wordpress.com] ).


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
October 29, 2020 08:05AM
Thanks, deckingman, it looks like I will be following you on YouTube from now on. I always find your progress inspiring and it is nice to see how well your printer is coming on - but also good to see that I am not the only entertainment that the gods of adversity have.

A couple of thoughts from your video: I have started to move to Torx stainless steel screws as I am also getting too many rounded out hex screws, along with rounded out hex drivers in the smaller sizes. I see that you use cable ties to hold your silicone tubing as I used to untill I found that I had to cut off the cable tie every time I released the pipe. I have started using Mikalor spring band hose clips (shown below) for this and they have proven invaluable in every way.



A final thought is that water cooling seems to need much less of a water jacket that intuition tells us. Although my tests have only just begun, little more than a drilled gallery is needed for a coolant path in the cold end.

I look forward to seeing how the mixing works in real life.

Mike
Re: My 6 input (5+1) mixing hot end
October 29, 2020 08:59AM
Thanks Mike.

I've used those sort of clips plenty of times in the past - they are very common in automotive applications. I didn't have any the right size to hand - hence the cable ties. But by way of an update, where the flexible silicone tubing slides over the less flexible but "non-barbed" PTFE tubing, cables ties don't work well. They don't form a perfect circle when tightened. Having mopped up the resultant mess and dried out the hot end, I've replaced that silicone tubing with some smaller bore stuff which I just about managed to stretch over the PTFE tubing. It'll do for now until I get hold of some proper clips........

Ref the "water tank" - good to know - if ever I make another one, I'll try that simpler approach. Those 3mm slots took me forever to machine as I kept breaking cutters. In the end I milled them out by using 0.5 mm depth of cut, so to get the 12mm depth took 24 passes ........per slot............and there are 7 of them. sad smiley

Ref the screws - I personally don't like stainless steel because they break too easily and the heads are too soft. I used to use them by the thousand when I built decks for a living but changed to exterior grade coated steel. The stainless steel ones would go in OK but would almost always sheer off if I needed to get one out. Torx drive heads are good though.


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
October 29, 2020 08:27PM
Quote
leadinglights
I see that you use cable ties to hold your silicone tubing as I used to untill I found that I had to cut off the cable tie every time I released the pipe.

I have some releasable cable ties smiling smiley
Re: My 6 input (5+1) mixing hot end
November 19, 2020 11:51AM
looking over your youtube 12 video....

bad stuff first, good stuff last. *meh*

THE BAD STUFF
1) to many stacks of metal
2) 1 or 2 heater elements for everything not per filament
3) water cooling for everything not per filament
4) filaments are at different spacings to single nozzle end vs same spacing to nozzle end (60 degree offset around a circle to each other)
5) cold water mix with hot water.
6) hot water goes everywhere vs cold water going to per filament tube and hot water exiting away per filament tube.
7) you are getting further away from "direct extruding" and getting more and more with the "bonwen tube"

THE GOOD STUFF
1) pipe inside pipe.
1.1) water jacket = hydraulic cylinder modified
--swap out the rod for a tube (aka filament tube)
--instead of pumping oil into either of the two hose connections, one connection cooled water goes in, and other connection hot water goes out.
1.2) google "RC DIY hydraulic cylinder" RC = remote control, example RC car, RC plane, etc... GOAL is the end or ends of the RC hydraulic cylinder, this should help ya figure out a way to "cap" both ends for a pipe inside a pipe without welding. outside pipe = water jacket, inside pipe = filament tube.
1.3) if at a local hardware store, ONE 4" pipe, ONE 2" pipe, TWO 4"x2" bushing reducer, TWO 3/4" barbed male thread nipples.
--extra step = little sanding on inside bushings "due to pvc/abs plumbing fittings"
--glue bushing on one end, and then glue bushing on other end. ((dry fit a few times to ensure 4" and 2" pipes get correct length.))
--nipple holes = drill press + V Block "2x4 with a notch cut into it to hold pipe" + pipe inside pipe, to drill 3/4" hole on both ends).
--tab/thread/glue 3/4" male to barb nipples into pipe.
--you now have a modified hydraulic cylinder / water jacket. "filament" slides through center pipe, and water comes into one side/end, and hot water goes out other side/end.
--you have a pipe inside of a pipe. OK size is extreme but standing in local hardware store, should help you visualize it. (see 1.2 RC = remote control details)

2A) individual water pump per water jacket (pipe inside pipe), per filament.
2A.1) running single large pump, pushing water to all filaments = heat transfer / creeping
2A.2) running single large pump, and using flow valves, something like in radiant heating, to control water per loop / zone *meh* more cost more frustrations.
2A.3) running a single small pump, per water jacket, per filament = you can ensure water per filament is getting exact same amount of water at all times, and can turn/on off water, by simple turn on/off water pump.

2cool smiley water circuit (suggested 1 of the 3) to reduce heat creeping, and not sending heat were it is not wanted.
2B.1) water pump outlet -> hose -> water jacket (pipe inside pipe) inlet -> water jacket outlet -> hose -> radiator -> hose -> reservoir -> pump inlet
OR
2B.2) water pump outlet -> hose -> water jacket (pipe inside pipe) inlet -> water jacket outlet -> hose -> manifold (connect 6 hoses together) -> radiator -> hose -> reservoir -> manifold -> to each 6 water pump inlets.
OR
2B.3) water pump outlet -> hose -> water jacket (pipe inside pipe) inlet -> water jacket outlet -> hose -> radiator -> hose -> manifold (connect 6 hoses together) -> reservoir -> manifold -> to each 6 water pump inlets.

2C) water hose harnesses.
2C.1) *meh* it has been a while. google "min bend radius hose", I want to say outside diameter of hose x 3 or 5 = min bend a hose can bend before you risk a kink. example 1" outside diameter hose = 3" or 5" bend diameter before risk of hose kinking. ((this is assuming you can not find manufacture specs of min bend radius for hose you are using))
2C.2) J hooks alternate back and forth spaced every so far. can allow hose to rest and wiggle and move. but still support it from falling down, or hold it vertically without fallout of place.
2C.3) local hardware store in the electrical and plumbing sections. they have "hose/pipe" bending/kinking protectors, along with nail/screw protection, and something to nail/screw to hold the thing in place on a 2x4. they normally are cheap plastic things, other words easy 3D printer easy things to print.
2C.4) I no long remember what it is called. take say a 2" PVC pipe. and cut 1/2 way to 3/4 way through pipe towards bend you want to make. the cuts will let the pipe give a little and bend in given direction without cracking/braking. reproduce same concept on some hose harnesses with 3D printer object, to help hold hose in position, but let hose flex back and forth in given directions some.
2C.5) *CRINGE* waiting and watching for some "vibration rubbing" to happen, and have one of your water hoses get a cut into it and leak all over..... local hardware store, window/door, or HVAC (heating/cooling), should have some "thin foam" for insulating, say 1/4" or so thick most likely have a glue backing. wrap around your hose once or twice before using a clamp or like. so your clamps due not vibrate into hose and cut it. TALKING physically clamping hose to a structural support not clamp hose to barb fitting.

3) heating elements
3.1) google "heating element types" then click up on "images" lots of heating elements. example: [www.eloybecker.com]
--be nice to find a "heating element" that fits snug and slips over end of "filament tube". I would not care about heating anything else but the individual "filament tube" less metal it heats up less heat creep.
3.2) I really never got into 3D printing until now. but just looking through various 3D builds. *meh* I simply do not like, the fact heating element is placed so far away and so much metal is being used. it simple does not make sense to me. it is like making a fire pit, lighting up a fire and letting the wood burn down to coal, then setting up a metal grate up away from the coals to cook some food. vs a standard oven top, turn on heating element set down pan, cook some food, no spacing nearly direct heating element to food. yet 3D printers seem like heating with coal.
3.3) the filament tube sticking out of the water jacket (pipe inside pipe), finding a heating element that wraps around the tube very end I would like to see. something like a "washer" sliding over a bolt in this filament tube. even a U or J shape heating element I would prefer, over a straight heating element that runs perpendicular to filament tube and heats everything else up vs directing heat to the filament.

4) filaments being same space to single nozzle end.
4.1) the viscosity of the melted liquid filament is going to cause trouble trying to get certain filament nearer or further to nozzle end to flow the same way. ya bunch of math, bunch of code, bunch of headaches when slight change of temperature, and/or humidity happens and changes all the settings and need to redo it.
4.2) 6 filaments = 60 degree offset around a circle, equally spaced to the single nozzle center. clean easier program code, and only changes that will be needed is different filament types.
4.3) take step further same angle to nozzle (referring to vertical axis) like done with the "diamond hot end"

5) KISS (keep it simply stupid) paths from filament hard, to filament liquid, to central nozzle.
5.1) without going in to a CAD program. purely guessing on overall size of disc.
5.2) disc = 2" thick metal, 8" diameter.
5.3) drill press + V block = drill 6 holes (60 degrees off set). and drill from the circumference (outside) to the center of disc. (V block = 2x4 wood with a V notch cut into it. helps hold pipe and like in place).
5.4) tap/thread holes on the outside. to fit say a 1/4" to 1/2" long bolt. ((remove bolt to clean clog, put bolt in to seal up))
5.5) drill press, lay disc flat, drill 6 individual holes say 1/2" form outside diameter, 1 hole per 1 of the previous holes drilled. ((these holes attach filament tube ends, need to be away from edge so not to clip bolts used in 5.4).......... really tempted to drill all the way for these 6 holes, so a bolt could be removed and access filament tube straight up.
5.6) flip disc over. drill in center for central nozzle.
5.7) tap/thread, screw in short nozzle.
5.8) I am not much of a fan, were you take 2 discs and bolt them together, and between the 2 discs you have routed out a path for fluid to flow. the 2 discs are harder to seal, more so when cleaning out clogs, you end up creating a bigger mess having to razor blade excess goop on both discs, and when both discs are not cleaned up good even when cleaned up good if there is no gasket (because of high heat) liquid crud will end up flowing out between the 2 discs, and or filament most likely end up burned black between the 2 discs were they did not fully squash together. was hoping with the 6 bolts ((5.4)) smaller areas that need to be seal up, and a little (insert some sealant) toss on threads or little bit at top of bolt head to act like a gasket for around bolt head, would obtain a higher sealing rate long term with abuse.
-- examples of above messes, is gaskets for various car engine parts, 2 stroke and 4 stoke engine parts (weed eaters/trimmers, chain saw engines, etc..), various transmissions, gear hubs, etc... these gasket surfaces or large size surface areas are to easy to mess up and cause leaks vs getting a good seal. and when assuming multiple times needing to get into an area. example cleaning clogged up spot. removing 1 bolt for 1 trouble filament, vs 4 to 8 bolts for a single filament. *meh*

6) water lines and extra.
6.1) if going by a redesign of the "diamond hot end" for 6 filaments and water jacket (pipe inside pipe)
-- cold water = comes in from the top, goes down through water jacket, and out the bottom, then out and up.
-- would consider keeping hot water hose on the outside, vs trying to run it up through center. trying to remove heat, vs taking heat from heated hose and having the blow fan take the hot heat and blow it right back down.
6.2) water cooling for desktop computers = water pumps, hose, radiator, reservoirs, fittings.
--not sure on total max temperature rated for though. only issue would most likely be, between outlet on water jacket (pipe inside pipe), up to radiator. then radiator itself.
--automotive, to tracker hydraulic hoses, to mowers, you should be able to find a small diameter cooling hoses or hydraulic hoses for rated temperature if issue comes up. but the issue more likely come if rated temperature they can handle is TO high and not enough.
--most of the desktop computer stuff has PWM, so getting an Arduino, raspberry pi, RC (remote control) circuit board, or desktop computer fan or water cooling circuit buttons and sensors setup should be possible.
6.3) EXTRA - might suggest heat sensor per filament tube. say between bottom of water jacket, and heating element. to turn on/off water pump/s, and/or turn on/off heating element, so as to obtain and maintain certain temperatures.
6.4) EXTRA - goal = each filament, filament extruder motor, filament tube, water jacket (pipe inside pipe), water pump for given water jacket, heat sensor, heating element. keeping them all together as a single unit per filament. and only combing the HOT END, right at the last moment, were having multiple heat elements, and high heat going towards central place / nozzle.
-- if it really needs it, having central nozzle area on top, having a CPU heat sink / water cooling jacket, with temperature sensor, with heating element. goal small heat sources, keeping smallest amount of metal warm, and cooling all other areas to a certain control temperature that is not really on room temperature to determine how nozzle behaves.
--err correction. to central area above nozzle. dedicated water pump, radiator, reservoir, heating element acts more like an actual water heater vs directly heating up central area near nozzle. more deals with heating up give area near to ensure all 6 path ways plus nozzle area are at a constant even temperature. more of a 7 way radiant heating system. hhmmmsss...

Edited 1 time(s). Last edit at 11/19/2020 11:56AM by boggen79.
Re: My 6 input (5+1) mixing hot end
November 20, 2020 05:02AM
Thanks for the critique. I haven't read every word because life is short, but I think I understand the gist of it. I did make note of the first sentence of you point 3.2. When you have a little more experience I think you will begin to understand that there are certain practicalities that have to be taken into account and which make most of your suggestions unviable.

Aside from the significant cost and complexity of using 6 individual water pumps feeding 6 individual cooling "tanks" or water jackets, and using 6 individual heaters, there is the matter of physical size. Taking your assumption that six two inch water jackets with their associated pipework could be fitted onto an 8 inch diameter plate, and assuming that belt attachment points could be fitted inside that same envelope, that would mean that the hot end would be 200mm in X and Y. Given that many consumer printers only have 200mm or so of travel, that would mean one would not actually be able to print anything. My printer is quite large and I currently have about 370mm travel in X and Y but your proposal would mean I would lose about 150 mm of travel which would reduce the size of anything I could print from about 370mm square to about 220 mm square.

Another issue with mixing hot ends is that all filaments are loaded at all times. Which means that they are all held at print temperature, even those that are not currently in use. This is particularly problematic for filaments such as PLA which will hydrolyse over time. So it is necessary to purge out that filament when changing colour (and of course anything in the mixing chamber has to be purged too). In my latest design, the length of the filament path from each inlet to the central point is about 30mm so that's how much filament needs to be purged when changing from one colour to another. Actually it's a little more because one needs to purge roughly 3 times the volume inside the mixing chamber for the new colour to be "pure" and the mixing chamber has a filament path length of around 15 mm, so I have to purge about 45mm. But that aside, a quick calculation reveals that having the filament inlets spaced around a 200mm circle (8inches) and assuming an angle of 45 degrees, means that the path length from the inlet to the central point would be some 141mm in length so that amount of filament would need to be purged every time the colour is changed. That is both very wasteful and time consuming.

There are other reasons why the volume of molten plastic needs to kept at a minimum. One is to do with retraction - as a rule of thumb, the bigger the melt volume, the more retraction is needed. But too much retraction can pull molten filament back up into or past the heat break, where it will cool and solidify, causing a blockage. The other factor is pressure. As filament gets heated, it expands and if the volume is fixed (which it is) then the pressure will build up and filament will "ooze" out the nozzle. When printing, the highly viscous, molten filament acts like a sponge as new filament is forced into the melt chamber. At the start of a print move, this leads to a delay between pushing filament in at one end and it coming out of the nozzle at the other end which leads to under extrusion. Conversely, at the end of print move, the pressure built up inside the melt chamber continues to cause filament to be pushed out of the nozzle even though the extruder is slowing down, resulting in over extrusion. The higher the volume of molten plastic, the greater this effect. Pressure advance can compensate to a degree, but only within certain limits.

So, for those practical reasons, the hot end needs to be as compact as possible. If you scroll up a few posts, you'll see that the topic of having the inputs arranged in a circle leads to a large and less compact design - hence the reason why they are arranged as they are.

Taking your "bad stuff" points...........

"1) to many stacks of metal" -
The mixing chamber is a complex 3 dimensional labyrinth. The only way I can make that is to mill slots in flat plates with inter connecting holes. It cannot be machined out of one solid part. For sure, one could 3D print that part in metal but that's a very expensive process and I need to develop this prototype first. Using a series of plates, means that I can modify the chamber reasonably quickly and cost effectively. Remember, this is very much a prototype design.

"2) 1 or 2 heater elements for everything not per filament". -
Given that the mixing chamber needs to be heated and the filament needs to be heated before the mixing chamber (inside any combining block) then having individual heater prior to that doesn't solve anything. (But it was an idea that I explored some time back).

"3) water cooling for everything not per filament" -
As discussed above, it's impossible to do without increasing the physical size which in turn causes other problems and makes such a design impractical..

"4) filaments are at different spacings to single nozzle end vs same spacing to nozzle end (60 degree offset around a circle to each other)" -
Due to the compound angles used, the difference in length of the filament path is about 10% between the longest path and the shortest of the "combining block" (about 31mm vs 34mm) or about 6% of the total filament path length. Although not ideal, I don't see this as being problematic.

"5) cold water mix with hot water" -
I don't understand this. The water flows from the pump through the cooling block and back through a cooling radiator to the pump reservoir in a continuous loop. How does that make cold water mix with hot water?

"6) hot water goes everywhere vs cold water going to per filament tube and hot water exiting away per filament tube." -
As discussed above, using 6 separate water jackets makes the hot end physically too large to be practical due to the severe restrictions on axis travel, excessively long filament paths, and excessive volume of molten plastic.

"7) you are getting further away from "direct extruding" and getting more and more with the "bonwen tube" -
Again, this is related to physical size (to say nothing of the mass) of accommodating 6 extruders onto a hot end. Flexible shaft extruders such as the Zesty Nimble might work but these have there own limitations.

The other issue with your proposal is the moving mass involved. Whilst I am not a proponent of the belief that reducing mass is always better, having too high a moving mass does mean building an extremely rigid machine, with large wide belts and pulleys, and using large powerful motors in order to provide sufficient torque to accelerate that mass at the desired speed. One would need a machine more akin to an industrial CNC machine than a 3D printer to move such a hot end as you propose.

But of course, all of this is just my opinion, albeit based on many years of using mixing hot ends. You could build a hot end such as you describe and prove me wrong smiling smiley.


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
November 20, 2020 09:26PM
sorry for being critical was not meant to be mean, honestly wanting to help. I do apologize it was not meant to be harmful. but to look at it and give feed back in positive behavior. your blog was awesome! and help me better understand thank you for taking the extra time with detail smiling smiley i do truly understand blogs, videos, just plain documenting is job and in of itself *been there done that*

1) looking at your #12 video [www.youtube.com] 8:30 to 10 minutes. it truly seem like you were mixing cold water and hot water. you have a lot going on.

2) mass as far as weight and momentum and 3D printer head moving all around i never really gave much though to it. you have it rather overkill setup (ya i am jealous!). but mass as in "physical amount of material" that needs to be heated up / cooled down. (cooling down a 10 foot x 10 foot room vs 4 foot by 4 foot room type of mass / volume) err ya should of said volume. *doh* my bad.
2.1) any way to laser zap / microwave the bloody filaments right on the spot to melt them? *sigh* wishful thinking.

3) with your last explanation, you sound like you are wanting to run all 6 tubes right on the "heating elements". to reduce both amount of material (volume) and pathing filament needs to take. less heating elements the better. and more in contact you can get the 6 filaments next to the 2 heater elements the better.

4) 8" ya over kill. looking at... [www.bilby3d.com.au] to get sizes.

quick search what came up. did not take much into account.

filament tube
4mm OD, 3mm ID pipe = [www.amazon.com]

water jacket tube
10mm OD, 9mm ID pipe = [www.amazon.com]

above was what i had more in mind.

5) *ughs* this is my version of chicken scratches with a pen/pencil and what ever near me to draw on.


looking at your version if i am taking things correctly, your version is "much more compact"

6) random thoughts to get stuff to mix.
6.1) looking at your recent blog. can you reduce filament pipe / path ways after turning to liquid. from say 3mm down to 1mm. and then when filament mixes/combines, bring area back up to 2mm to 3mm in size? (this is more of goal of feeding in smaller faster streams of a given filament at a time, so perhaps it would result in smaller "stripped tooth paste".
6.2) and use the extruder motors to act like driving the hard filament down like a push rod / piston?
6.3) going as far as push filament down, then retract filament up a little bit, then push filament back down. (create more of a vibration for liquid filament to mix better?

7) there are "mixers" but i am going to take "filament liquid" and more so the viscosity of the filament liquid as a bigger issue.
7.1) if you can bring the viscosity up so the filament is more liquid, vs a paste or syrup, but rather more water flowing, it would be easier to mix. but if the 2 filaments are at different viscosities it would be much harder to mix. since all you have to work with is "path shapes, and path sizes, heat, and extruder motor.
7.2) most mixers require a impeller propeller / some sort of chopping wheel to blend things together. adding an additional "motor with most like a long shaft to run all the way down to nozzle, would be out of question...
7.3) then ensuring a good even "heat distribution for the filaments" would be first. if one filament is going fast and steady while another filament is slow but still steady, the high faster filament being used up, is more likely to be "colder" vs the slow and steady filament that has a longer time to warm up evenly through it and become hotter.
Re: My 6 input (5+1) mixing hot end
November 21, 2020 10:18AM
@boggen79. I think you are under the impression that water cooling has to remove a lot of heat energy from the cold end of the filament. The heat-break section of the filament feed tube is deliberately made as nonconductive as practical. This is commonly done by making it from a material with poor conductivity (stainless steel) and having a small cross-section. During normal filament feeding operations, very little heat will get to the cold end as the fresh filament being fed down the tube helps to cool it. The main problems are when filament feeding is stopped for some time or when there are repeated retractions of filament between short periods of plastic deposition.

Cooling "jackets" have been as simple as a single capillary tube soldered at right angles to the filament tube and these have been satisfactory. The arrangement used by deckingman is very similar to that used in cylinder blocks of i.c. engines - where the cooling is much more demanding.

I have recently done some tests on how much cooling is needed and you can see this on [reprap.org] I have given some figures on the temperatures which indicate that even the simple approach to cooling is more than adequate. Complicated individual jackets or even complicated multi-jackets are not needed. BTW, for scale in the first figure of the quoted URL, the bore of the barb pipe connectors is 2mm.

Just a reply to your earlier point about heaters. Many on this forum have experimented with coaxial heaters but the gain given is vanishingly small. A heater element in a block gives a thermal mass which, along with PID heating, gives very good control of temperature. The main problem, also not solved by coaxial heaters, is that there can be a temperature drop between the block and the tip of the nozzle.

Mike
Re: My 6 input (5+1) mixing hot end
November 22, 2020 04:29AM
@boggen79. Don't apologise - you haven't upset me. Criticism is a good thing and everyone is entitled to express an opinion. In turn, I hope my reply didn't cause you any offence - I just happen to disagree with some of your ideas. For what it's worth, I happen to think that the entire process of making something using what is effectively just a robot controlled hot melt glue gun is somewhat archaic.

Taking your random thoughts.....

6.1. The filament tubes through the "combining block" and the mixing chamber are actually 1.8mm diameter (the filament is 1.75mm dia). The problem with making them any smaller is that they would be too restrictive when printing just a single colour.
6.2. and 6.3 . Every downward movement of the filament results in extrusion at the nozzle tip. For sure it might be possible to "oscillate" the filament in such way that the forward motion is greater than the retraction motion, to simulate a continuous feed rate. The filament inside the melt chamber would act like a sponge and effectively damp out the pulses such the the flow through the nozzle is constant. But that would have to be done in firmware and if the oscillations were too fast, it's likely that they would exceed the step pulse frequency limit. It may or may not be effective, especially when one or more filaments are being fed at single digit percentages of the whole. In any case, this old retired mechanical engineer doesn't have the skills required to hack the firmware.

7.1. Viscosity is a measure of a fluid's resistance to flow so I think you mean bring the viscosity down, not up. For sure using higher temperatures does reduce the viscosity but from a chemistry and fire hazzard point of view, there are numerous reasons why that isn't a viable option.
7.2. There are one or two "paddle" type mixing hot ends that I am aware of, but none that are commercially available. The biggest problem is sealing the shaft, given that the "paddle" is inside a chamber full of hot molten plastic. But one or two people seem to have overcome that. However, things have gone very quiet and I haven't heard any updates recently. I don't know, but I suspect that such designs might suffer from not being able to stop the flow of filament for non-print moves. With a conventional hot end, this is simply achieved by partially retracting the filament which causes a negative (or zero) pressure at the nozzle tip. I suspect that retracting the filament as it enters a "paddle type" mixing chamber might not result in a reduction of pressure at the nozzle tip. Perhaps also such designs might have a relatively high volume of molten material which is not a good thing for the reasons that I stated earlier. Maybe that's why nobody has as yet made such a design available to purchase - just guessing here.
7.3. That's why I use aluminium - it's a very good thermal conductor. I did do some temperature profile testing and the results showed that at a point some 30mm away from the heating element, the temperature was only in the order of 1 or 2 deg C below what it was right next to the heater when running at 200 deg C. Having said that, filament itself is a relatively poor conductor so it does take some time for the heat to find it's way through to the core. But because of the length of the filament paths, each filament has plenty of time to reach the same temperature, regardless of their individual flow rates.


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
November 22, 2020 04:46AM
Quote
leadinglights

..............................
The arrangement used by deckingman is very similar to that used in cylinder blocks of i.c. engines - where the cooling is much more demanding.
.................................

Haha, that made me chuckle. You are exactly right. Being a (retired) automotive engineer by training, it's exactly like an engine block.

For info, I've taken on board your suggestions. I'll stick with this cooling chamber for now, just because I've already made it and it seems to work. But I'm working on another design which will be a two in, one out and all metal. The intention is to use it for "exotic" filaments but also for high flow rates with large nozzles. Think super volcano but cut in half to make two shorter melt chambers, then position them side by side and feed filament through both at the same time. The advantage is that not only does one have a large surface area but also, each individual filament passes through at half the rate of the whole, so spends twice as long in the melt zone (one does need two extruders - but I have six on the machine in any case). But anyway, the cooling block for this is simply two vertical holes (in and out) linked with a single cross drilled "blind" hole capped at the end. None of that milling slots and making a lid stuff......


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
November 23, 2020 12:11AM
how much control due you have over the "feed rate" per filament. (critical)
how much control due you have over the "x,y,z" (critical)

standard printer (to print crud on paper). has black, yellow, blue, red, etc... ink cartridge. and they print "dots" out.

example: for one row....
one blue, one yellow, another yellow, another black, empty white space, empty white space, etc...
next row
one red, one black, empty white space, one yellow, etc...
next row
one yellow, one black, empty white space, black, blue, etc...

3D printer....
tooth paste strips instead of "dots of ink" the tooth paste strips are the dots. of ink.
can you control the dots per layer to line up with dots above on next layer to help control wanted color?

a blending of tooth paste strips in nozzle ok to some degree. but can you control so level below, current layer, and next layer above in 3D print layers, help align your wanted color? with some degree of blending in the nozzle?

side note: could care less internal structure mess is printed in what color, if anything used to clear out one or more colors out of nozzle, and tuned for color that is printed on outside that is showed. even it means making an "internal mess" of crud.
ya doing it cheap jerry rigging things way. *shrugs* but start one spot, then fine tune it down.

===============
if your going for a paddle to mix.
1--whisk for beating eggs, would be nice but that small "meh" trying to get something that small to last. S shape chopper blade perpendicular to nozzle center.
2--straight rectangular cube shape is out = more likely to cause color shift in a greater area
3--a gear with teeth are out, even multi gear setup as beaters, would "pump" liquid, or more than likely just spin and do nothing to mixing.
4--any sort of fan blade, propeller, impeller shapes, with more likely "pump" liquid.
5--a flat piece of sheet metal, paper thin, with holes in it, might cause a backup of liquid going back out the bowden tubes.
6--a flat piece of sheet metal, paper thin, and cut out 4 to 8 blades. but all blades are still flat to one another might be something. but still might send fluid back up bowden tubes, or more thank likely have some "squirts" of extra filament coming out of nozzle in one spot and less in other spots.
7--something between wood screw and bolt threads, but you may end up "pumping liquid" out of the nozzle at a faster speed. (creating a screw pump), though adjusting feed rate of extruder motors perhaps?
8--scroll pump = look at refrigerator, freezers, (a/c) air conditioner pumps. this would really be a "pump", and the blade would not be thin piece of paper sheet metal at end of a shaft, but more like a piece of pipe, with some cuts in it, and then the now "tabs" at end of pipe slightly bent in towers center of pipe. and this pipe just feed down nozzle part way maybe?
9--upside down U shape beater or even right side up U shape. as long as drive shaft comes down into the center i suppose.
10-- hhmmsss -------|-|-|-| multiple blades (paper thin sheet metal rectangle shaped). but spaced at different lengths along the drive shaft, blades not stacked directly on top of each other (that would create a screw like thread), but a distance between the blades. and rammed down the nozzle. say a desktop computer motherboard stand offs. male thread one end, female other end. stack them. with something between them.
11--ugly critter to above 10, solder / weld, little wires to drive shafts to act as chopping blades at different lengths along the drive shaft.
12--umbrella (refer to 9), but an umbrella without the fabric, just points of wire sticking out of drive shaft.
13--directly bend the heck out of the drive shaft to form...(S on top of S) see below. putting a wave form in the end of the drive shaft.
|
|
S
S
14--going further on 13), put long drive shaft in bench vice, grab end of drive shaft with pliers, and just the heck out of the drive shaft, twisting it like a pretzel. warping the drive shaft to make S shapes in it. more like taking a shot glass of whiskey, and adding little bit of something else, and taking a tooth pick to stir it up.
15--take above back on 14), route out something with groves for a WWWW, and then a mating other side. and then put drive shaft between the two pieces. toss in bench vise and crank down to "physically press" the drive shaft into a shape. (more controlled and repeatable results) hhmss jaws on a pipe / plumbing wrench. hhmm vice grips some of them have teeth going correct way.
16--a spring attach to drive shaft, i doubt would get anywhere, just clog up around spring, and liquid filament shoot directly through. well take that back a "slinky" toy, that fully collapse (smallest bolt threads) = no good. but a slinky fully stretched out. forming more like a Wood screw thread possibly. but something that small long term, even short term i just do not see happening.
17--drive shaft with end bent L shape. or most likely better at an angle. i do not know maybe 35 degrees from drive shaft, to inside diameter of nozzle or hole going down to end of nozzle. assuming hole is larger than end of nozzle. and some space for drive shaft to "wiggle".
side notes: trying to stay away form "flat blades" like a ceiling fan might have. and more of a round rod, to reduce any type of suction/pumping action. causing liquid filament come out more in one spot and less in others.
18--stacked washers, inside diameter =1 mm, next washer inside diameter 2mm, next washer inside diameter 1mm, repeat a few times. thickness of larger inside diameter washer may cause difference. more of trying to play with orfice sizing (hole sizing) of causing the liquid to speed up, slow down, speed up, slow down. (slam on gas, slam on brake, bloody down town stop lights!), but in reality, liquid would speed up in 1mm inside diameter, then slow down as it spreads out in larger 2mm diameter. i was thinking if tooth paste stripping spreads out "thinner", then shoots down to next then spreads out again. more of a mixing. more likely be a 0-ring (metal), to washer (metal disc with hole in center of it), to 0-ring, repeat. path or rather layers may need to be too long to make worth while.
19--5 gallon bucket paint mix. toss into a drill, has a long shaft, and blades are more vertical with shaft, vs ceiling fan blades that are more horizontal. *meh* i have seen something like that a few times. in a "Dremel bits" not really a router bit, that is to horizontal cuts, but there was some more vertical cut to it like a drill bit. but had more of a 2 to 4 blades to it.
20--dremel bit, little metal wires on it. more for say brick, metal grinding and clean spots up. (wire brush that spins), less wires = good.
21--wrap metal wire around a shaft tightly to create a coil. before removing wire / coil form shaft. stretch wire out, forming more of a pig tail curly cue. *meh* would most likely need an actual "spring making machine" example: google how springs are made, and check youtube for some video.
22--you have a "cheese grater" in kitchen. run a block of cheese over grater to get tiny strips of cheese to fall down. but i want to say there is a "pipe" version for like a dremal / router. its like someone took an awl and punch it into side, then before removing awl bent awl 90 degrees to make teeth.
23--blender in kitchen,
24--electrical "reed switch" that goes on/off when there is a magnetic force. no clue if you can make one to strong enough to flip back and forth. or a few of them.
25--brushless motor (a/c or d/c) good example would be cheep water pumps for aquariums, small fountains, etc... can ya toss magnet in there and just let it spin with some little blades on it?
Re: My 6 input (5+1) mixing hot end
November 24, 2020 03:46AM
I thought that I had made it clear that I have discounted using any sort of "paddle mixer" because of the high volume of molten plastic, the inability to reduce nozzle pressure for non-print moves and the difficulty of sealing the shaft. If I was to choose that route (which I'm not), I'd likely use some sort of high Curie temperature rare earth magnets for the rotor part and excite these using external electrical coils.


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
November 24, 2020 04:01PM
instead of moving paddles / mixers. "STATIC non moving mixers"

A = hydraulic cartridge valve = google it and click images.
--it allows a "large" path between, to liquid filament, to nozzle. and the only time there is small sized holes would be right at the cartridge valve. were the liquid filament would enter small holes into the center channel of the "3D printer filament cartridge nozzle"
--higher cost O-rings to deal with temperature.
--was hoping each level to act more like a "shower head" and stream out different color filaments per level, to allow for better mixing.

B = worst type, layer upon layer of channel, and then holes.
--fighting even distribution of filament, fighting layers ballooning in between layers, fighting filament going between other holes.
--no easy way to clean. except toss in oven, and let it melt off the troubles a few times. and that is if you can even remove all the plates. and needing to remove entire 3D printer head.

C = washer with large hole (0-ring more or less)
C1 = single smaller hole stacked to larger holed washer. (see what happens)
C2 = multiple smaller holes stacked to larger holed washer.
C3 = kitchen faucet "screen" built up crud and had to clean it today, nice little washer O-ring, with a screen. stack some screens. no need to use screens to "filter" the filament liquid. but act as static mixer no moving parts. if the liquid filament is moving. then put stuff in its path to go around to mix the liquid filament. a few screens stacked, so screen / mesh is not to small in hole size. more multi stacked = more mixing. spacers / washers used ensure screen/mesh does not lay on top of each other and blocks/clogs up certain parts.

C3 info real world.
hunted in back of bathroom drawer. found a old stick of "Aquafresh" tooth paste, RED, WHITE, BLUE. side by side strips as it came out of the tube.
i still had tool to undo faucet sink strainer / screen. from cleaning it earlier.
the Aquafresh tooth paste stuck end right on screen and squeezed.
there was a tiny bit of "blending" / "mixing". it was more the paste came out in long spaghettis like noodles. but when i turned it horizontally the noodles collapse / fell off into sink. but broken off part had some blending, as the string noodles did not fall back together again perfectly.
no old window screen, set down a washer, fold window screen over, another washer, fold window screen over. have same size washers, just no window screen *frown*
took bathroom sink faucet apart to get to its screen, little finer mesh.
larger hole kitchen sink mesh washer, bathroom sink small mesh = spaghettis noodles to more of a "smash" and "smear" of the 3 colors of spaghettis noodles. causing things to spread out before building up enough to go through mesh. some blending / mixing.
bathroom sink smaller mesh by itself = smaller spaghettis noodles but no real blending/mixing pretty much same thing with larger mesh kitchen sink screen.
i do not have any thinner washers. everything I have is for 1/2" bolts or bigger. and the thinner washers the inside diameter to small for the tooth paste. *frown*
screens = one is flat, other is cone shaped. all i am doing is getting tooth paste every were. *frown* not able to lay the screens more near or on top of each other.



*meh* remember something, never used a static mixer before. so no clue on anything. siphon mixing yes. but static mixer no. there is no motors, no sensors, no electronics, just a piece of material setting in a tube not moving at all.
google "mixing valve pond ozone static mixer" clicked up on images.


anonymous image hosting

UGLY BORING STUFF, where, why, how, stripped tooth paste filament. my attempt to comprehend as i think it through.
primarily going to be talking about pressure loss / resistance. those that are dealing sizing pipe for water, air, pulps, other liquids, including waste water more along the lines.

hot glue gun (quick compare)
--you have a heater element, a heat sensor, a glue stick, and some sort of handle grip that helps grip the glue stick to push it into the hot end of the glue gun.
--glue stick = filament, handle gripper = stepper motor, hot end of glue gun = heating element/sensor of 3D printer head.
--you pull on handle gripper squirt your glue, release, and the glue keeps coming out.
--you use the glue gun, keep it on, and just warming up the gun, glue comes out.
--put a big old glob of glue onto something, then it cools down, then it peels away and possibly cracks. (shrinkage).
--filament acts same way as above.

-- if you had six hands, and 6 glue guns. how due you get the glues to mix?

===========
stepper motor
--stepper motor each step it takes as it turns, there is a low and high areas were it produces the most pressure on the filament.
--micro stepping can also cause additional issues between the micro steps in how much pressure it places on the filament. some times it is more some times it is less.

--if the filament is being pushed or rather "pulsed" down into the hot end. it is going to cause the liquid filament to pulse as well. some times the filament will have good pressure pushing the liquid filament out to the nozzle. while other times, it is like steeping on a garden hose / crinking the garden hose and the water just goes down to a trickle.
result = liquid filament fighting over the other pressures of other filaments. which ever has the "higher pressure" pushing to the nozzle WINS and exits towards the nozzle. leaving the other filaments watching waiting for there turn at the stop lights.

extra = filament rolls, hang up some, or a little extra tug need to spin the roll to get some more filament. little less pressure going to pushing filament towards the hot end. who knows when or how much this will be. so lets just over size the stepper motor with enough power.

===========
hot end
--one filament is speeding away 100 miles per hour, and going out the nozzle, the color is needed and going quick.
--one filament is in there own world slow move 5 miles per hour, going out the nozzle. just enough to give a little tinting.

--the fast filament, takes say 2mm to go from hard filament to a liquid filament and then another 1mm to come to a constant temperature.
--the slow filament, takes say 1mm to go form hard filament to liquid filament and also constant temperature.

--the fast filament = requires more pressure to push it along. (it is harder to push filament that extra 2mm, gotta have that engine power to spin those tires!)
--the slow filament = requires less pressure to push it along. (barely tapping the gas, and woops me moving slowly)

result = fast filament = fighting over the other filaments going to nozzle is going to have less pressure to fight against them, (due to having to fight that extra 2mm)
result = slow filament = has more pressure and pushes away other filaments. (never had to fight those extra 2mm)

============
path ways, path shapes, path lengths, path connections, etc...
--long paths = require more pressure to push liquid through them.
--shorter paths = require less pressure to push liquid through them.
--20 foot garden hose with sprinkler, and the sprinkler shoots 25 feet or more, VS 100 foot of garden hose with sprinkler that shoots a 18 feet or less.

--larger diameter / larger area liquid flows through = less pressure to push the liquid through.
--smaller diameter / smaller area liquid flows through = more pressure to push the liquid through.
--cheapest 20 feet garden hose, ID (inside diameter) of hose 3/8" hooked up to sprinkler, sprinkler shoots out 20 feet.
--expensive version 20 feet garden hose, ID (inside diameter of hose 3/4" hooked up to sprinkler. sprinkler sheets out 25 feet.

google "orifice size for valve"
--you have a garden hose, and you take a thin piece of plastic and drill 1/8" hole in the center. and put the "washer and plastic piece into the female end of garden hose. and tighten to faucet. turn on garden hose. and you get a ""trickle of water"" coming out.
--you have a garden hose, and you take a thin piece of plastic and drill 1/4" hole in the center. and put the "washer and plastic piece into the female end of garden hose. and tighten to faucet. turn on garden hose. and you get a ""steady small stream"" of water coming out.
--low water shower heads = small washer with a little hole in it, higher water shower head = no washer with tiny hole in it. or washer with big hole in it.

fluid flowing through angles.
--J = nice curved bottom of J, = less pressure to push fluid through a J curve
--L = 90 degree sharp curve, = more pressure to push fluid through a L / 90 degree curve

============
=============
FIGHTING TORNEMENT OF FILAMENTS
A color -- new large spool of filament, longer path from stepper motor to hot end, short path from liquid state to nozzle.
B color -- 3/4 of large spool of filament, longer path from stepper motor to hot end. long path from liquid state to nozzle.
C color -- 1/2 of smaller spool of filament, short path from stepper motor to hot end, medium path from liquid state to nozzle.
D color -- new large spool of filament, different brand and is softer, short path from stepper motor to hot end, long path from liquid state to nozzle.
E color -- older medium spool of filament. different brand and is harder, long path from stepper motor to hot end, medium path from liquid state to nozzle.
F color -- 1/4 of medium spool of filament damp, different brand, short path from stepper motor to hot end, short path from liquid state to nozzle.

the fight from spool to stepper motor.
E,F,C,B,D,A
the fight from stepper motor to hot end.
C,F,D,A,B,E
the fight from liquid state to nozzle.
F,A,C,E,B,D

Wait, Wait, hold me beer! everything changes the next second, and the next second it again changes, then second after that... ... ..(UGHS) constantly changing of which color wins the fight exiting the nozzle constantly changing never stops changing.

HEY HEY, watch 2 colors actually in an even fight they both went out of the nozzle together, first time in the last minute!

===============
so what is going on?
as A color enters the center pipe (tournament battle arena) with goal for end of the nozzle.
B color has extra pressure (stronger) kicks A to the side and heads on out.
A color thinking he still has pressure, but sudden upset D color comes in and blows A backwards from pressure difference.
A color still thinking he has a shoot, been building up some pressure from moving so slow, and comes head to head with E color, and the fight it out as they go out the nozzle.
C color looking around thinking it is finally safe, big wimp, not much pressure left fighting its was to the arena, looks left, looks right, looks up and down, seems good. gets part way out, and gets chopped in half as B slams his way through with his high pressure.

what ever filament has the highest pressure (entering the center pipe), wins and goes out first, the lowest filament with least pressure goes out last, that or until enough pressure builds up to push others out of the way.

OH SNAP! a breeze just went through the room. B is constantly loosing now, from being to cold and needing more pressure to push through the cold!

===================
OLD DAYS AND COMPARE!
can we set things up in such a way, we can accurately predict and calculate pressure of filaments as they enter the central pipe leading to the nozzle?

2D paper ink jet printer
--in a standard printer, printing on paper 2D, we can measure the in level, and make some calculations to rough predict how much pressure is at the nozzle.
--as the ink level goes down, we know there will be less pressure, and make adjustments.
--the ink is for most part air proof cartridge, with just a tiny air hole that is control electronically to open/close, keeping moisture out and ink is in a control environment.

2D paper tape / ribbon printer
--2 spools, good ribbon wound up on one, and as it got used, wound up on other spool.
--ribbon would dry out, and other times be to damp.
--some times it would stick, other times it would not.
--some times it would flip flap and another color would print out.
--some times color from above would bleed down to below color.
--reminds me of current 3D printers and filaments.

Edited 1 time(s). Last edit at 11/24/2020 04:12PM by boggen79.
Re: My 6 input (5+1) mixing hot end
November 24, 2020 05:25PM
i grabbed a tooth pick, 20 awg magnet wire (wire to make coils for motors / electric magnets).

made 5 to 7 turns one direction, then continued to wrap it back down the tooth pick 5 to 7 turns, then wrapped it back in the original direction for 5 to 7 turns, then back one more time in other direction 5 to 7 turns.

IT LOOKS FREAKING UGLY!!! it looks like a picked up out some trash heap *frowns*

removed tooth pick, and it makes it look even worse that not even a dog wood sniff.

turned on kitchen faucet to a trickle, needed to get a flash light so I could see things better.

beside my shaky hand and trying to line it up correctly. those few seconds just right, i did see the water nearly "clinging" to it. and going back and forth between the 4 winds of coil.

i took some flour and added it to a cup of water to make it a bit thicker. and i got same results as just water. but i got to a point of adding a bit more flour were it just "slumpped" off to the side. and i gave up. not about to find a pipe to fit the turd into. and see what happens.

it does look like quick test it would "mix" fairly well. but that is just pure messing around. liquid filament and higher pressure and speeds may be a different story.

this is a quick sketch. 4 coils running in opposite direction (every other one) 1mm diameter wire, 4 turns over 20mm. the drawing makes it look neat. the real thing without having perfect turns (meh).



Edited 2 time(s). Last edit at 11/24/2020 05:31PM by boggen79.
Re: My 6 input (5+1) mixing hot end
November 28, 2020 05:37AM
@boggen79

May I respectfully suggest that you start your own thread, rather than keep posting your thoughts and ideas in this thread. I've been on this planet for some 67 years, the last 50 or so practising engineering in one form or another. I've spent much of the last 5 years or so doing multi colour printing, and most of the last 4 years using mixing hot ends. So I'm already fairly conversant with the basic engineering principles and fundamental material properties of what we are dealing with here.

I started this thread to inform others of the work that I am doing, in the hope that my practical tests and the empirical data that those tests and designs produce might help others. Anyone looking for that information will now have difficulty finding it amongst all the theoretical hyperbola that have now been added by your lengthy posts.

Even better still would be if you actually built working prototypes, ran some tests of your own, and posted the results in a forum such as this. That would really benefit the RepRap community (which is what these forums are about).


[somei3deas.wordpress.com]
Re: My 6 input (5+1) mixing hot end
June 01, 2021 04:39AM
great work . I have been following your work since long time. Cant wait to see it work..

Have you seen this?

multi color active mixer .

this guy has it figured out but now has kind of disappeared from the web . The impeller in this seems to be the secret sause . thats y there are no images of the impeller cool smiley
Sorry, only registered users may post in this forum.

Click here to login