The FLSun S1, the new speed king for 3d printers

[Alan Timm]

Just when you thought we were at peak 3d printer...

https://us.store.flsun3d.com/products/flsun-s1

https://www.youtube.com/watch?v=EXL7ZqsYrYw

https://www.youtube.com/watch?v=UQor7KCzhJw

tldr;

• 1200mm/s print speed?!?!?!?
• 40000mm/s2 acceleration?!?!?!?
• 320mm x 320mm x 430mm build volume
• 350c hardened steel nozzle
• completely enclosed
• closed loop steppers (no more layer shifts)
• load sensor on spool holder to estimate remaining filament
• direct drive extruder
• auto leveling
• heated filament dryer box for spool
• camera, ai print evals (looks for spaghetti)
• structured light sensor for print calibration?
• built-in imu for input shaping calibration
• air filter

[Alan Timm]

For anyone wondering how they do it, here are the s1 slicer settings.

tldr; 1200mms is used for travel speed only, 500/500/800mms is still crazy fast for everything else.

[Chris Albertson]

This is one VERY impressive printer. The idea of building a filament drier into the printer is great. And I think my next printer will also have closed-loop steppers. But as for speed, we reached a limit.

I have a “fast” printer too. It is a very different kind and likely much less expensive. (It is a Veron V0.2) But all of these high-performance printers have the same bottleneck: How fast can you extrude plastic from the 0.4 mm hole in the nozzle?

To extrude the plastic first it has to be melted so the question becomes “How fast can you melt plastic? There is a chamber of liquid plastic above the nozzle and the printer pushes solid plastic into the chamber to force out the liquid. You speed this up by making the heater and the chamber larger but the bigger you make it the less predictable the result because the incoming solid plastic is more disconnected from the liquid at the nozzle. With a small chamber, the coupling is more direct and the process is easy to control and you print a more accurate part.

The extruder I use is a compromise and can do up to about 13 cubic mm of plastic per second. So if my part needs 13,000 mm^3 of plastic, there is no way to print it in less than 1,000 seconds. But this is hugely optimistic as most of the time you have to go slower than 13 mm^3 for any of a dozen reasons.

Years ago linear motion may have been the bottleneck but today we are seeing that the better printers are operating at nearly the theoretical limit of the extrusion process.


I think today it does not even make sense to think about building a robot if you don’t have access to a 3D printer. The top line printer like this FL Sun are attractive, but you can still buy a usable printer for $200.

[Alan Timm]

Yeah, I thought I knew the limits of slinging plastic until I bought my first flsun v400.

That monster can and will render parts at 400mm/s, but at reduced quality.

Actually they updated the specs and the same machine now maxes out at 600mm/s with 20000mm/s2 accelerations.

I quickly figured out that if I slow it down to 25% to 50% of the max speed I get amazing print quality, 200mm/s for large parts and 100mm/s for smaller parts.

I'm expecting that it's possible to run the S1 similarly, which suggests 200-400mm/s amazing prints.

Alan

[Chris Albertson]

With a layer height of 0.2 mm and a 0.4 nozzle, 200 mm per second is 16 cubic mm per second. that is the near limit extruder rate of standard extruders. I set the rate to 13.5 mm^3/second to the printer only moves at 160 mm/s while printing

I can print faster with a 0..15 mm layer but this does not change make the part finished faster

[Alan Timm]

For anyone thinking of picking up an S1 you may want to wait.

Their discord has multiple accounts of new users complaining of the surface print quality.  it shows resonance patterns that only get worse the slower you print.

So far the prevailing theory is that the closed loop steppers slightly under and over compensate resulting in a noticable surface pattern.

Here's just one example.   

[Chris Albertson]

This is a question that should interest anyone working with closed-loop systems.  That means anyone working in robotics.   The basic PID algorithm we all know can only react to measured error and can oscilate about the correct output.   I don’t think reactive control like this can ever be perfect.   

In the case of closed-loop steppers, the number of motor poles is very different from the number of sensor pulses pre-revolution.   I’d agree you should expect a periodic error.

To investigate this you’d need a very high precision rotational sensor.

On Jun 20, 2024, at 3:04 PM, Alan Timm <gest...@gmail.com> wrote:

For anyone thinking of picking up an S1 you may want to wait.

Their discord has multiple accounts of new users complaining of the surface print quality.  it shows resonance patterns that only get worse the slower you print.

So far the prevailing theory is that the closed loop steppers slightly under and over compensate resulting in a noticable surface pattern.

Here's just one example.   

<S1 bad surface.jpg>

On Wednesday, June 12, 2024 at 5:37:04 PM UTC-7 albertson.chris wrote:

With a layer height of 0.2 mm and a 0.4 nozzle, 200 mm per second is 16 cubic mm per second. that is the near limit extruder rate of standard extruders. I set the rate to 13.5 mm^3/second to the printer only moves at 160 mm/s while printing

I can print faster with a 0..15 mm layer but this does not change make the part finished faster

> On Jun 12, 2024, at 3:10 PM, Alan Timm <gest...@gmail.com> wrote:
> 
> Yeah, I thought I knew the limits of slinging plastic until I bought my first flsun v400.
> That monster can and will render parts at 400mm/s, but at reduced quality.
> Actually they updated the specs and the same machine now maxes out at 600mm/s with 20000mm/s2 accelerations.
> 
> I quickly figured out that if I slow it down to 25% to 50% of the max speed I get amazing print quality, 200mm/s for large parts and 100mm/s for smaller parts.
> 
> I'm expecting that it's possible to run the S1 similarly, which suggests 200-400mm/s amazing prints.
> 
> Alan

-- 
You received this message because you are subscribed to the Google Groups "RSSC-List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rssc-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/rssc-list/c1d73941-855b-465a-a689-115131c6f935n%40googlegroups.com.
<S1 bad surface.jpg>