Flying extruder, did I take it too far?

[Adam Talbot]

I started out with a Nimble extruder...  Then a NEMA 8 worm drive extruder...  Then BMG dual extruder...  But every option had down sides.  Bowden was the lightest option, but came with it's own host of issues.  Then I stumbled on a picture of a Delta with a flying extruder.

So I quickly designed, and built my first flying extruder.  I love everything about the flying extruder!  But I thought I could do one better.  I thought I could make my extruder lighter.  So I removed the fans from the Effector and moved them up to the flying extruder assembly.  But this presented me with an interesting problem.  What type of pipe/hose/duct should I use to connect the part cooling, and hot end cooling fans to the effector?

I needed a light weight, flexible, and elastic form of ducting to get the air sources from the flying extruder to the effector. After racking my brain a bit, I came up with the perfect, all be it odd, solution.  Condoms.

I think this concept could be taken further by providing a temperature controlled air flow from an external air source.  This would work very well with more exotic materiel's that require much more precise cooling.

[ekaggrat singh kalsi]

just curious what was wrong with a nema8 worm drive? weight or retraction speed?

[Piotr Mońko]

Hi Adam, 

I love your durextruder :) 

I'm just curious what where the downsides of Nimble, worm drive (got any pics), and BMG that flying extruder solves?

--
You received this message because you are subscribed to the Google Groups "Delta robot 3D printers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to deltabot+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/deltabot/CAH_2Ghe%2BRGaEoz2sWMtThjxb2BYHzu18C86ZytWyDxn5TtEX2g%40mail.gmail.com.

[Daren Schwenke]

I went down this rabbit hole a couple years back, so perhaps I can save you some time.

I tried CPAP tubing, but it was way too stiff and heavy.  Like you I ended up slapping a condom on my printer.  :)

The latex condoms will degrade pretty quickly in the air and failure sucks, unless you like to keep you printer well lubed...

They also make polyurethane condoms, which do last longer but they are *much* less elastic/bendy.

Then I tried molding some flexible ducts over some lubed PVC pipe from silicone, but I just used acetic acid cure caulk so they were either too stiff or the thin ones shredded on removal from the pipe.  

So that prompted embedding some fiberglass cloth, which was awesome and let them be very thin but that kinked way too easy.  I did use this idea later for noise/thermally isolating a blower on an incinerator though and it worked great.

So that led to embedding some tool (light mesh fabric) which was the best.  The tubes would still tend to kink near the edge of the bed, but they worked well enough that is where I ended up stopping.

Nothing was ever really as nice as the original latex condoms though.

In the end I abandoned all that as the weight of the ducting I modelled almost offset the gain of moving the fans to the flying bit. 

The lightest solution I found for my flying setup was water cooling the hotend and then using an good sized air pump/ring nozzle of thin wall AL tubing for my part cooling.  

A tiny amount of moving water provides massive cooling and that whole setup filled with water was half the weight of the normal heatsink.

[Adam Talbot]

Lets see if I can comment on all of the responses.  Lots of fun pics attached, and links embedded.

History of my setup.  This is my first 3D printer, and I jumped into the deep end by starting with a Delta.  I knew this was going to be a challenge, before I started. After 20 years of building CNC routers, I do have a rather solid understanding of most the concepts. I tried to buy, what I though was, best in class hardware.  I loosely designed my Delta off the Kossel concept. Please keep in mind I started this build in May of 2019 and I am still learning a LOT, so take everything I have to say with a grain of salt.

Extruder(s) and what led to condoms.....  I started off with the nimble extruder, and my effector was nothing more than a simple CNC'ed plate (Version 1). This got me through MANY of the early learning pains.  The counterfeit stepper controllers (Leadshine DM556's) were the hardest to troubleshoot. Although the uStepperS closed loop steppers missing steppes had some of the most fun out comes. At one point I was even running all motion control off an Arduino Uno! But I always had problem with ghosting at high speeds (+250mm/s).  

My goal is to reach a print speed of 1000mm/s. As my delta matured, so did my effectors, but I was never able to locate the root cause of a minor fluctuation/ripple in the print surface.  It was not ghosting, it was something else.  Said fluctuation/ripple were on all of my parts and was there from the beginning of my machine.  So on a whim, I strapped a BMG extruder onto my effector (V3).  First time in all this time I had a flawless finish quality.  The only change was Nimble to BMG.  So I built V4 of my effector which had native BMG support.  Instant ghosting, even at slow speeds!

Turns out when you add a ton of weight to the effector, you get lots of ghosting.  This lead me to new/bigger belts.  I upgraded from GT2 to HTD3M belts, and solved the ghosting, at slower speeds.  I knew weight on the effector was a BAD thing. I proved that the Nimble was causing some kind of strange issues.  So this lead to removing the nimble drive cable and direct attaching a NEMA 8 motor (V5). The results were perfect, kinda. The 1/4~1/2 of a 3DBenchy was perfect, but the motor would over heat, and freeze up.  Even a bigger NEMA8 suffered the same issues.  But I had hope.

I found a cool worm drive extruder on thingiverse which I quickly adapted to a NEMA 11 motor (V6).  The NEMA11 did not overheat, but still ran too hot for my tastes.  The worm drive also had a gear engagement issue that was again causing surface defects.  To make matters worse, I still had a big heavy motor on the effector.  Then I saw my first picture of a flying extruder. I was able to quickly print out a test of the concept, and fell in love!

My goals with the flying extruder concept.  Remove ALL weight from the effector.  Even better if I can remove any other extra forces, like the constant push of the Nimbles drive cable.  Thus leading to my current setup.

The Nimble.  I love the concept, but I think the drive cable is having issues. But I don't understand exactly what.  I have seen other complaints about this on-line.

The BMG.  I love everything about it, except the weight.

The NEMA 8's all had overheating problems, in my setup.  Please keep in mind I run my motors at 36V.

NEMA 11 + worm drive works, but I was having gear problems, and heat issues.

Now to tackle the current state. The condoms will degrade fast.  It's just going to happen with any natural rubber product.  I have pondered some of the non-latex condoms.  I have found that ultrasound probe covers might be a good non-latex option.  But non-latex condoms are much simpler to come by.  I have also pondered using something like rip-stop nylon fabric.  It has no elasticity, so would require a longer length, thus prone to more kinking.  I am not sure what the best path/option is, yet.

Water cooling has always been on my radar.  But I never found a small water cooled heat block.  Many of the solution just put a jacket around the E3D V6 heat sink, which is kinda pointless.  Daren, what are you using?  Links/pictures?

Please ask questions, or comments. I am always looking for ways to improve.

--

You received this message because you are subscribed to the Google Groups "Delta robot 3D printers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to deltabot+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/deltabot/209ee0d6-8b8d-4ae2-8c9a-406a9688b4db%40googlegroups.com.

[Daren Schwenke]

I don't think the hotend part of my build will be much use to you.  :)

https://hackaday.io/project/20763

My flying extruder.  It uses an assembly line tool counterweight spring tuned to the weight of the extruder from above.  The lines to the carriages just hold it centered and keep it from bouncing around too much.

https://youtu.be/3uK_ZrMR9dQ

You can do the same thing for a single stepper motor with just this:

https://www.amazon.com/KEY-BAK-Retractable-Keychain-Polycarbonate-Oversized/dp/B01CQLJ8I0

As for a cooling block, you can build your own very simply.  Any small cube of AL drilled in two directions will suffice.  

Even just the contact area of poking a hole in some silicone tubing and sticking the hotend through is enough as long as the water can still move.

https://hackaday.io/project/158650

To unsubscribe from this group and stop receiving emails from it, send an email to delt...@googlegroups.com.

[Adam Talbot]

Wow...  Thats a great setup.  I will adapt the "counterweight spring" to my ridge.

Now its time to build a custom hot end with water cooling!

To unsubscribe from this group and stop receiving emails from it, send an email to deltabot+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/deltabot/e6d590de-7fc7-4d72-a831-2199194cc2e7%40googlegroups.com.

[Boštjan Čadež]

The simplest way to make a water cooled extruder is to convert one of these nf-crazy or dragon. I used the trick Marc Peltiers (Zatsit) idea and just put a silicon tube instead of the copper fins (they slide of easily).

The downside of those extruders is that they use copper blocks which are really heavy. On mine I modified an aluminium e3d block and used syringe needles as standoffs. I used the nf-crazy heat brake which complicated things a bit because the thread is shorter and i had to mill (with the printer) a recess so it would touch the nozzle. This heat brake should work better. 

The effector on the pictures weighs 60g which is half the weight of the air-cooled one \o/

To view this discussion on the web visit https://groups.google.com/d/msgid/deltabot/e6d590de-7fc7-4d72-a831-2199194cc2e7%40googlegroups.com.

[Daren Schwenke]

Looks great!

You received this message because you are subscribed to a topic in the Google Groups "Delta robot 3D printers" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/deltabot/n3PImRPtmOo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to deltabot+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/deltabot/99b902d6-a402-4fd8-a7e0-2a1f1ba3fdfe%40googlegroups.com.

[ekaggrat singh kalsi]

would love to see more details on the routing attachment. how heavy is that? and how fast can it mill ?

[Boštjan Čadež]

The weight is 160g.  I made it to mill acrylic which it can do at about the speed of printing. 900mm/min with .1 mm depth of cut. I mill PCBs at 300mm/min. And the aluminium experiment was at glacier moving speed of 10mm/min with a depth of cut 0.05mm (more like grinding :D). 

The speeds are slow but with mag joints and the cheapest possible tools its ok. Its definetly better than not being able to do it at all:). Right now Im waiting for parts to make a better spindle which can actually be replicated by others. The one I have uses a modified dumpster diving part.

I'll post it here when I'm done.

[Boštjan Čadež]

Forgot to post some pics. The middle two holes in the e3d block I messed up by hand later. The four outer ones is what it came out like.