Fastest 3D-Printer in the World

The functionality is the judge, thats the nice and simple thing on this part. It wont snap flat, then the quality is too low.

If you want to see it with better surface, then i just print a bit slower and its still faster then anything else i guess.

Print it with similar quality in under 38min, proove it, and the title is yours.

---

Homepage

But thats why i was wondering what 200mm/s looks like? coz if thats bad it helps to judge it against others.
What speeds are people hoping their Core XY will do?

Everybody thinks mm/s is important, but it is not. All that counts, is ACCELERATION and LOW MASS.

And this is why i can print so fast with my lame 130mm/s infill. The stiff and lightweight frame i designd makes this possible, the other parts are just cheap ramps standard, more isnt necessary.

I already printed at nearly 300mm/s @ 0.1mm layer, but at this speeds, the mass of the effector / hotend gets so high, it will wobble. Therefor i make a new hotend with less than 1/4 the mass of my actual one, but dont know when its finished, because i´m a lazy pig.

Edited 1 time(s). Last edit at 01/25/2017 07:38AM by Kolbi.

---

Homepage

Politicians answer, I havent seen your other prints at slower speed, so where can i see one at around 200?(so you printing slower but with fast accel?) which cant be that fast for a Delta but is 4x more than i'm printing so reduces time to a good rate, but if it effects quality then i'd have to go slower.

Edited 1 time(s). Last edit at 01/25/2017 07:56AM by MechaBits.

I dont understand completely (my english), do you think i printed with 200mm/s in the video? It was maximum 130mm/s for solid and infill.

---

Homepage

No, I had no idea what speed, I thought you where going at something like 400 or 800mm
but you say it tops out at 130, so it looks fast because of how fast you get there.
So I didnt think 130 was considered the fastest possible.
While you have printed what look like some good overhangs/slopes
the part looks bad, so if you dial it down to get a good print what is the speed?

Edited 1 time(s). Last edit at 01/25/2017 08:04AM by MechaBits.

Mostly, i turn down to 80-70% for much better surface and edges. Cooling makes the most part of overhangs quality, speed is nearly no matter. In the video you can see a lot of bended up edges in one part, this is because i forgot to heighten the air cooling nozzles. In the next cut you can see the nozzles higher, and no more bending.

---

Homepage

Quote
Kolbi
The functionality is the judge, thats the nice and simple thing on this part. It wont snap flat, then the quality is too low.

If you want to see it with better surface, then i just print a bit slower and its still faster then anything else i guess.

Print it with similar quality in under 38min, proove it, and the title is yours.

lol, new year new "who has the ..." comparison ?

this comparison is useless with real rules, all written in this thread yet is so soft and not regulated, it can be cheated very easily.

for a real comparison you need more specific rules !

* same nozzle diameter set at software (note,slic3r with auto diameter is cheating there, you have to set all diameter values to the same value !)
* same layer height at the whole part.
* same print part dimensions = picture of the print with a caliper (it is possible to cheat with a different print size after gcode !)
* same Infill type/%, for sure Line Infill is the fastest
* same software should be use, and better define a setting (available for download) that has its core parameters that are allowed to be changed and which are not allowed to be changed
* gcode used has to be made public for fake/settings check.
* filament material does not matter
* filament color does matter, only colors that are easy to take pictures with and to compare sould be allowed (no natural white/transparency, translucent, sparling,... best are pure colors which are not too bright)

* in my opinion parameters that should be open for tuning are, speeds, acceleration, yerk, ....? any other parameter should be locked down to a specific value.

last but now least , your printer isn`t the fastest, trust me i know a couple of printers that are faster

And as the_digital_dentist , high speed is overrated, the real limit for nice prints is set with the filament viscosity and no one can cheat that


Chri

---

[chrisu02.wordpress.com] Quadmax Intel Delid Tools

Do they click together?

I just tried slicing this part with my normal settings, and got 57 minutes. Tweaking things got down to 45 minutes, but this is still just a slicing test, not a real speed test. I don't think I'd be able to get "acceptable to me" quality at lower than 50 minutes.

I must admit, I'm on the side of the Digital Dentist when it comes to quality. Printing time for me is: it takes as long as it takes. Faster can be better, but only when you don't sacrifice quality.

But that's just my opinion

Ah i forgot some things:

* not only same software has to be used, also same version/build, only official build`s are allowed.
* Heavy duty /break test, its also possible to cheat via flow (optional?)
* when making a videoclip, it has to be from beginning until removing the part, including a watch(timer) which is not manipulateable to be visible the whole time, or even better play a movie (opensource) at a display in the background, that is partly concealed by some of the printers moving parts.

Chri

---

[chrisu02.wordpress.com] Quadmax Intel Delid Tools

I think there are useful conversations to be had around printing speed and time. Print times are frequently long, and if you have a large printer, print times can become a limiting factor to its usefulness. Multi-day prints are difficult if you consider it unwise to leave your printer unattended.

On the movement side, I'd be interested to see a discussion of which design is best suited to high speeds. Delta looks like it could have the lowest moving mass, and I like the symmetry of the movement. On the other hand, they always look a little flimsy. Perhaps coreXY with it's more robust construction can handle the acceleration better despite the likely higher moving mass? And Viktor has shown us that even a Prusa style moving bed can be made fast if you throw enough force at it.

Then there's the question of laying down plastic. What limits our ability to place plastic accurately at high linear speeds? Position of the nozzle is covered by the movement section above, so here I'm thinking about things like melt-rate, liquid flow and layer adhesion. Are there any factors of nozzle design that can help improve high speed performance?

As Chri pointed out above, printer performance is very multi-dimensional, and the concept of 'fastest' is frankly silly. However, I have needed to get shorter print times recently while churning out big parts for the new build. My printer is Prusa style with a heavy dual direct drive extruder mounted on bendy 8mm rods, so high speeds and accelerations are out for me. I've been exploring large nozzles as an alternative, which more than halved the print times for the parts I needed. Large nozzles aren't suitable for everything, but they are good for structural parts as the resulting prints are strong.

I was quite surprised by the caliblock part that DJ pointed us to. It's a very clever design and prints well with a bigger nozzle. I ran some tests last night and this morning, and I can get functional prints of the part in less than 15 minutes (not including heat-up time) using a 0.9mm nozzle and printing at 19 mm³/s.

Quote
Chri
this comparison is useless with real rules, all written in this thread yet is so soft and not regulated, it can be cheated very easily.
for a real comparison you need more specific rules !

I agree, but the many rules you posted makes no sense, you only need a simple classification, otherwise you´ll have 10000 cathegories in short time.

Just use

- Operating principle (filament, liquid, etc)
- Motor Size
- If you want you could also classify voltage (12,24,...V)

Then its fun and many can join the competition.

Quote
Chri
your printer isn`t the fastest, trust me i know a couple of printers that are faster eye rolling smiley

Dude, i trust you as far as i can throw you, show me the proof!

Quote
Chri
when making a videoclip, it has to be from beginning until removing the part, including a watch(timer) which is not manipulateable to be visible the whole time, or even better play a movie (opensource) at a display in the background, that is partly concealed by some of the printers moving parts.

The fact is, you could easy manipulate clocks and play films slower.

The only proof of concept is when you walk around and talk a lot in the video, because its hard to fake your mouth movement i think. Also its very simple, everyone can do that proof.

Quote
nebbian
Do they click together?

Yes, better then the first ones. If you do not believe, i can make a video.

Quote
JamesK
Print times are frequently long, and if you have a large printer, print times can become a limiting factor to its usefulness. Multi-day prints are difficult if you consider it unwise to leave your printer unattended.

Totally agree, it was my big 12mm rod Multicolor Rostock, that made me think of fastest possible speeds. Couldnt sleep for days when printing big parts because i had fear it will burn and kill me in sleep.

Quote
JamesK
On the movement side, I'd be interested to see a discussion of which design is best suited to high speeds. Delta looks like it could have the lowest moving mass, and I like the symmetry of the movement. On the other hand, they always look a little flimsy. Perhaps coreXY with it's more robust construction can handle the acceleration better despite the likely higher moving mass? And Viktor has shown us that even a Prusa style moving bed can be made fast if you throw enough force at it.

I think the delta is always a bit easyer to make it fast, especially in Z-dir for fast retracts. Also it seems to have a weight advance the bigger it gets. I bet i get any kartesian machine go this fast with few mods, just keep mass low as possible and you win.

Quote
JamesK
What limits our ability to place plastic accurately at high linear speeds?

Linear Speed Limit: Weak Extruder and/or bad hotend or settings
Acceleration/Dynamic Limit: Clamping filament (bad hotend/extruder/spool) and geared extruders, dont use them!

Quote
JamesK
Are there any factors of nozzle design that can help improve high speed performance?

Not really, it just should be machined well.


The question should be "Are there any factors of hotend design that can help improve high speed performance?"

Then i would answer 4 things:

1.) Low Mass
2.) Accurate and smooth drilling of hotend inside
3.) Powerful Heat-Barrier for maximum nozzle heat
4.) LOW MASS

Quote
JamesK
I ran some tests last night and this morning, and I can get functional prints of the part in less than 15 minutes (not including heat-up time) using a 0.9mm nozzle and printing at 19 mm3/s.

Sounds interesting, can you proove it and show us please?

---

Homepage

Quote
Kolbi
Sounds interesting, can you proove it and show us please?

I'm not interested in posting videos and all that silliness, if that's what you mean. I have photos of the parts and the host log for the record that I can add this evening. Other than that you can either take my word or not, as you wish.

And i have photos of an ufo and the log files from the flight record that I can add....

Dont take it personal please, just fun.

But if you want to be serious, you need a video proove. Film a few layers and a person takling, click test, and it should be enough to calculate if its possible.

---

Homepage

Quote
Kolbi
And i have photos of an ufo and the log files from the flight record that I can add....

Dont take it personal please, just fun.

But if you want to be serious, you need a video proove. Film a few layers and a person takling, click test, and it should be enough to calculate if its possible.

Like I said, not interested.

Fastest is fastest. Why should the be any restriction on motor size, print size, printer architecture, printed material, etc., if there isn't any restriction on final quality?

---

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

Agreed DD, a 3D printer is a 3D printer maybe we should be only comparing FDM machines printing the same object which has to function to be accepted, but I can't see why there has to be any other stipulations.

Well said James, however I am willing to take the bait albeit sarcastically. Here's my video:

[www.youtube.com]

Sorry I did have to put the watch (which is made from wood BTW) , back as it fell down after a few minutes, the time was not changed.

44 minutes. Printed in ABS with 0.4mm nozzle on e3d v6, 2 perimeters and 10% line infill. Acceleration 3000, Jerk 1200, normal printing speed for perimeters 60mm/s, but printer set at 400 to 500% speed during this print. I could go a bit quicker as I ramped up from 150% for the first layer to 400, then 450 after about 5-6 layers and switched to 500% nearing the end, but it probably could do the whole thing (minus layer 1) at 500% speed.
As you can see at the end the cal blocks click together, the white one was done in 57 minutes previously so I have a new PB to be proud of.

I think I ought to be calling this the second fastest 3D printer in the world, yes?

In all seriousness its not a very pretty part, I did switch the acceleration to 6000 but at that speed the perimeters started to become ragged, so I've hit the mass/inertia limit for my effector as it is currently setup at that speed. Sure I could remove the fans, and switch back to a much more minimal effector. I've got a micro-delta running a deltaprintr mini hotend which is tiny (and very low mass) but its heater block is too small to melt plastic at that rate, so it wouldn't go any faster and actually print anything.

I also tried this on my corexy but something curled up and I missed some steps and got a spaghetti ball instead of a cal block.

So are we all done seeing how far we can p1ss?

Edited 5 time(s). Last edit at 01/25/2017 01:18PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
digital dentist
Fastest is fastest. Why should the be any restriction on motor size, print size, printer architecture, printed material, etc., if there isn't any restriction on final quality?

There have to be regulations, otherwise teams will form up, buy strongest drives available for big money, and no normal person can join competition anymore. Also energy efficiency on such systems is very very shitty, so not interesting for production.

Quote
DjDemonD
44 minutes. Printed in ABS with 0.4mm nozzle on e3d v6, 2 perimeters and 10% line infill.

Great job! 2nd fastest printer it is. Clever to save infill, will try that too.
Pitty there is no real time sound, just for personal issues i´d like to hear the sound of your machine.
You could try to increase print acceleration to 4000. You´ll have more ghosting, but save a lot of time.

---

Homepage

Your Delta's(DJ) much tidier, zippy too, How many grams was the print so I can compare SIZE with my last print.

Edited 1 time(s). Last edit at 01/25/2017 02:39PM by MechaBits.

Kolbi, you can barely hear my printer running, its based on a duetwifi and so interpolating to 1/256th microstepping. I did try 4000 acceleration but wasn't happy with it, could have worked but my desire to have a good object at the end was too strong.

As for regulations I'm not sure this is an official contest of any sort, you want one set of conditions, those which sort of suit your setup, other people want it stricter and some looser (my preference). My machine has nema 17 motors, but it also is 24v, tooling plate bed, mains heater, duet wifi etc.. all in probably cost me £800-£1000 to build, the time involved was ridiculous. So not a fair contest under those rules anyway. However its not built to win a speed race, but be the best delta I could afford to build, and it still lacks a metal effector and metal carriages to increase the precision even further. If anyone is good with CNC let me know. I think its quite normal to want more and more quality, but only a more modest increase in speed. But I take the point that if 3D printing at high quality was much faster it would change the nature of it from prototyping and hobbyist use to some production.

Thanks Mechabits, the blue cal-block was 12.97g, the white one I did a few months ago in 57mins was 16.45g.

After James' thread with the interlocking cube challenge it might be interesting to see if the cal-block can be done with 1 perimeter?

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
DjDemonD
After James' thread with the interlocking cube challenge it might be interesting to see if the cal-block can be done with 1 perimeter?

It can, and no infill. Although my perims are 1.2mm wide

Yeah might have to break out the big-ol' volcano nozzles, if I wasn't snowed under with customer prints to do - which I have been neglecting doing this speed challenge today, then I might have a go.I wonder if it will work with my 0.4 nozzle?

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Yes, I should think so. Keep the layer height to less than half the extrusion width and I think this will print fine. You might get into some strength issues with the connecting lugs though.

Quote
DjDemonD
...However its not built to win a speed race, but be the best delta I could afford to build, and it still lacks a metal effector and metal carriages to increase the precision even further...

Would you consider an epoxy glass PCB effector and carriage adaptors? We may have some beta ones available soon.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Yes definitely interested to know more. I presume they are considerably stiffer than printed parts?

I'm using Haydn's magnets so I'd need something which would be able to adapt to that system. I'd definitely try it, but with development on the piezo sensor z probe, I'm still a little in flux deciding on the best way to go. Am thinking of adapting Lykle's effector for the piezo sensors to position a v6 hotend above the effector to reduce (theoretically) tilt even further.

Edited 1 time(s). Last edit at 01/25/2017 03:38PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
DjDemonD
As for regulations I'm not sure this is an official contest of any sort, you want one set of conditions, those which sort of suit your setup, other people want it stricter and some looser (my preference).

I totally agree, this is why i started the thread, we need simple regulations to make a few categories for contests that are seriously.

For now, this is the Nema17 Filament Contest with parts everyone can afford.


Machine Rules:

- Filament
- Nema17 steppers
- 12 TO 24V, not more!


Proove Rules:

- Video with moving and talking people in background OR you show a handy and open google, then play a random video on youtube. Clocks are not allowed because of very easy cheating.

- There must be a snap test, that shows that parts can be removed without breaking. At least one side must be snapped and removed 3 times to prove the parts are strong enough, and you must show that the sides lay even on each other.
The important thing is to make the video in a fake proove way. Best is one complete, uncutted video. If not possible, try to make a video where you show your printed part detailed during print for multiple times. On the basis of this, it can be recognized when you pull it off the printer and make the snap test without cuts.

- Optional: Break the part to see how strong it was


I think rules cant get easyer, or does anybody have another simple ideas?

Edited 2 time(s). Last edit at 01/25/2017 04:58PM by Kolbi.

---

Homepage

The effector and carriage adaptors are designed to use Haydn's rods. The first version of the effector will incorporate a piezo sensor, but until we have the prototype we won't know if we have got the stiffness right. We are also looking at accelerometer and strain gauge versions in case the piezo version proves troublesome. To save hijacking this thread further, I suggest we continue this discussion elsewhere if need be.

Edited 1 time(s). Last edit at 01/25/2017 03:47PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

So my last print was 29g(0.4) in 2hr15 not fast only about 40mm/s with smaller parts slowing a lot, nice print. Did you use a smaller nozzle?

Sounds great David happy to be involved in any way. Piezo sensor isn't far off, just waiting on some acetal and ptfe rods from Doug to see if I can just eliminate that last bit of nozzle wobble before settling on a basic design for the module, that can be incorporated into whatever comes later on.

Thanks Kolbi for the clarification but I think it's run its course for me.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
MechaBits
So my last print was 29g(0.4) in 2hr15 not fast only about 40mm/s with smaller parts slowing a lot, nice print. Did you use a smaller nozzle?

No 0.4mm but 2 perimeters, 2 bottom and top shells and 10% infill.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions