Jump to content
Ultimaker Community of 3D Printing Experts
  • Sign Up
Sign in to follow this  
lars86

Complete Hot-end and Print Head Redesign

Recommended Posts

Posted · Complete Hot-end and Print Head Redesign

I find that my more or less stock, but well tuned, printer can definitely gracefully support higher throughput than the hot end can deliver, and in comparing my results with others, it seems that both my mechanics and my throughput are representative of many other users. If the limiting factor for you is the printer mechanics, then I'd definitely try and figure out what's different about your printer.

Getting sufficiently fast extrusion is a problem that I deal with on a daily basis. In my experience helping users on here, and offline, I'd say it's the single largest problem that trips up new users and leads to blocked heads and stripped filament. With the wrong settings, you can very easily exceed the throughput capacity of a standard nozzle at very mediocre linear speeds.

As Joergen noted, while you might want to print perimeters at more modest linear speeds, you can certainly print loops and infill at very high linear speeds, and all of them at high volumetric throughput, in order to reduce print time without impacting finished print quality or strength.

 

 

So, can you post up a video of your printer pushing 30mm^3/s with a decent surface finish?

Even with thick 0.15mm layers, 250mm/s printing is only 15mm^3/s. I was able to hit that with perfectly fine extrusion, but circular perimeters were terrible. It looked like a control issue; uneven jerky speeds. When using slower perimeters on the new Cura, it underextrudes the infill.

Maybe we just print with different expectations.

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

I tried the same gcode with white PLA from German Reprap today. The speed difference to the colorfabb PLA/PHA is impressive. Infill is only fine up to approx 170 mm/s, even with increased temp and flow rate.

Lars, it would be interesting to see how the PLA/PHA filament works with your hot end redesign some day. It's less viscous and therefore more prone to the issues you are adressing.

 

How do you like the PHA overall? Consistent extrusion and good bonding?

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

So, can you post up a video of your printer pushing 30mm^3/s with a decent surface finish?

Even with thick 0.15mm layers, 250mm/s printing is only 15mm^3/s. I was able to hit that with perfectly fine extrusion, but circular perimeters were terrible. It looked like a control issue; uneven jerky speeds. When using slower perimeters on the new Cura, it underextrudes the infill.

Maybe we just print with different expectations.

 

No, precisely because I can't melt 30mm³/sec. That's two or three times what the standard print head/extruder can extrude.

Having a large nozzle makes it easier to sustain in the 15-25mm³/s range, but that's about as far as I can reliably go with PLA. With a 0.65mm nozzle, and 0.2mm layer height, that's about 170mm/s linear speed. I don't have to even get close to 250mm/s to start bumping into the limits.

Which was the point I was making in the first place.

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

How do you like the PHA overall? Consistent extrusion and good bonding?

 

It's a PLA/PHA compound and I suppose the PHA share is quite small. But it definitely alters the material characteristics. The manufacturer emphasizes the extended elongation. Most people in the colorfabb thread praise the colors and the somewhat smooth touch. That's true but not the point for me.

Heated to recommended temperature (210-220°), the viscosity is significantly lower than PLA. It is almost liquid and drops out of the nozzle if the printer stands still. Once you have started printing it extrudes very consistently and I assume the extrusion process window is considerable larger. But PLA/PHA is more prone to blobs and stringing. Bonding between layers is like normal PLA, but the surface is less sticky to my heated glass bed.

With PLA/PHA my standard hot end setup comes close to the 10-15mm^3/sec barrier Joergen mentioned (like in the stress test video). With normal PLA unfortunately, it stays far away from it.

Anyway, for high quality prints I set perimeter speed to 50-100 mm/s because of the motion quality. Extrusion speed is not an issue. That is the point you are emphasizing. But frequently infills make up 50 to 90% of the extruded volume. Printing with 200 mm/s instead of 100 mm/s there would significantly reduce overall printing time. My mechanics could do that, my standard hot end can't (with PLA).

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

No, precisely because I can't melt 30mm³/sec. That's two or three times what the standard print head/extruder can extrude.

Having a large nozzle makes it easier to sustain in the 15-25mm³/s range, but that's about as far as I can reliably go with PLA. With a 0.65mm nozzle, and 0.2mm layer height, that's about 170mm/s linear speed. I don't have to even get close to 250mm/s to start bumping into the limits.

Which was the point I was making in the first place.

 

We are just trying to solve two different issues. You are looking to push the threshold of maximum speed, I'm assuming for larger objects. On the other hand, I am willing to wait longer for a print that has very nice details, great surface quality, better overhangs, bridging, and bond strength. I don't see myself moving to a larger nozzle, printing with 0.2mm layers, or with print speeds even approaching 170mm/s.

I showed that my new print head can easily push 15mm^3/s, which it sounds like some stock setups cannot do. All while cutting the melt zone essentially in half. This lends credibility to my theory that overcoming friction in the filament path and the inherently higher nozzle pressure, with the stock filament drive, has more to do with max extrusion rate than melt zone length (not to say that it doesn't play a role). I'm telling you; if you design or retrofit a better filament drive, you will print faster and more reliably. The drive system is on my hit list, but I've been too busy with work. I think that the Makerbot style drive gear is a better design. The larger diameter and true radiused groove, allow much more contact area. This gives you less filament deformation through the rollers (less frictional loss and more reliable cross-sectional area), and more potential driving force before slippage. I have one here to design something around. I just wish they would have cut that groove with a dimension closer to 3mm (it's larger). A twin drive system would be trick.

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

Lars - I think we both want the same thing - reliable, good quality prints. :-)

In my experience, the print head movement system is already better than the extrusion system can fully support. We print slower than we might otherwise do, because the print head can't provide enough plastic at the right temperature when it should, and provides too much when it shouldn't. It sounds like your new head design addresses at least some of that, and if you can sustain 15mm³/s then that's also a step forward, so yay! As you say, reworking the extruder drive mechanism is also going to be a big part of a long term solution.

I'm interested in understanding what prevents consistent printing at, say, 300mm/s, however, because as far as I can tell, even then it's not the mechanics of the print head movement but there do seem to be some extrusion/temperature related problems that crop up at those head movement speeds that aren't directly related to throughput volume per se. If we can solve them to enable high speed printing, I think that can only help improve quality even further at lower speeds.

There's another issue that you touched on earlier that I think is worthy of further exploration, too. You talked about circular perimeters being particularly poor, in terms of the quality of both finish and head movement. I also saw this in some 300mm/s+ print tests I was doing. In a lot of ways, circles are the hardest thing for the UM to print, because they require the largest number of very short segments, while also providing the lowest velocity change at each junction - allowing for more speed to be carried through each transition. So they require the most segments to be printed in the least amount of time. While that could tax the mechanics of the printer, I think that in practice it's taxing the firmware more than anything - and, indeed, overtaxing it.

I tried printing a 130mm diameter high-res cyclinder at 350mm/s. It basically worked, but the movement was slower than requested, and the movement wasn't very smooth. Looking at the resulting gcode, and doing the math, the end result was the printer was going to have to process about 130 line segments a second. So, about 0.008s per segment.

I think the problem here is that the firmware simply can't plan line segments that fast. When the buffer starts to empty, Marlin slows down segments to take at least 20ms to print. If the buffer gets completely empty, then it pauses momentarily until it can refill. I suspect that that when printing circles fast, we're seeing those slow downs and pauses, resulting in erratic movement.

I'm planning on adding some instrumentation to Marlin, to see how long segment planning takes, and try to figure out if there's a better way to manage the buffer to make it degrade more gracefully. Perhaps a larger buffer and/or longer delays would give better results - or a more intelligent process that gears the timing of a segment to how long segments are currently taking to process (I presume the time goes up the buffer fills, due to the need to scan the buffer to keep all the segment entry and exit speeds consistent).

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

Looks pretty good. But, how is the retraction working? So far all "all-metal" hotends have failed our retraction-hell test, and blocking up after a few hours.

 

Hi there -- was on holiday so was quiet about this thread ;)

I also experience clogs with my own all-metal hot end. I think it boils down to a TOO drastic thermal break in my case (something like 40° to 230 in 3mm?). When the filament is pulled back, I guess it "freezes" immediately.

It then probably adapts and sticks so well to the inner metal tube tiny scratches that it will not resume even when I push it back by hand... I have to stop the dedicated cooling fan to let a longer part of the hot end get hot enough to resume...

I though that the filament would contract when cooling down, but if it does it is not enough... It just fits perfectly and seals the colder upper tube on a few mm. It works with a very short retraction like what does Lars but I'm left with some stringing on long hops though on my side (esp. as my bowden tube is 4mm wide inside)... :(

I still need to figure this out... My best idea so far is to make the thermal transition longer (what a pity!), so the filament will not freeze so quickly!

I am sure that we can get rid of this damned expensive PEEK ;)

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

Also, not to hijack the thread, but I'd be interested to hear your experiences (and anyone else's) experience with Jeremie's blocks, as I'm considering installing those when I take everything apart to install the new pulleys.

 

Not exactly the answer you expect since I may have a skewed opinion on the banana blocks ;)

Now I hear the danger of over-tightening the bushings... May be the parts printed very nicely on my printer but I really had no issue. Now, my bushings are quite shorter than the "real" ones (only 16mm long, and same 11mm dia).

Still, I have really no trace of wear and the thing moves better than before (partly due to a better alignment may be), so I don't feel I would even buy harder-to-find longer ones if they get damaged in a few years...

What would be nasty is if ever some blob is left on the printed parts, that dis-align the bushing on the X/Y rods indeed. Careful inspection and cleanup is required I guess indeed.

I also initially had lips around the bushing (you'll see the "shoulder" variable in the openscad source), but it did not prove useful in my case -- I can't remember why but I thing it was conflicting with the stop switches, eg. Actually when I finally printed and used the block near the Z screw, I just left the standalone screw lose, as it looks like there is really no reason to tighten the bushing at all. IMHO I really should make a new revision without it at all.

Now I don't like much the idea of a spherical bearing, just because I would need to buy some ;) Seriously and for sure, it would add more freedom... but there is no reason if the rods are properly tuned and squared... I don't like the idea that my (short) bushing is allowed too much rotational freeplay on the rod, as I suspect it could really induce *more* wear on the rod each time the direction is changed... It still would be worth a try though because this analysis may be plain wrong -- I am fully self-made in mechanics -- ;)

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

I print ABS nearly exclusively and I usually print solid parts. The situation you cite is typical for me. The rate that I can feed, melt, and force plastic through the nozzle is usually the factor limiting my print times. In addition to the possibility of converting to 1.75mm filament, I think some speedup (or. alternatively, sub 0.40mm nozzle use) is still available for the taking with a stronger filament feeder.

I'm using a 0.50mm MakerGear BigHead nozzle and, much like Joergen's somewhat modified setup, my hot end can handle higher temps than a standard Ultimaker. I'm not inclined to use a larger nozzle because the quality (resolution) of the visible printed surfaces suffers. I haven't suffered from jams using quality ABS since making the minor changes to my hotend so I'm not interested in a shorter melt zone, if it cannot melt ABS as quickly as my relatively stock setup.

BTW, I made a passively cooled, short stainless steel hot section on my lathe once. It didn't work. After several minutes of printing, the ABS filament would always begin to resist feeding and eventually come to a grinding (literally) halt, or nearly so.

 

Until you run into a situation where (at the speed you are after) the motion quality of the print head is acceptable, but you can't extrude fast enough, the solution to faster printing is either in the slicer program and/or the UM's control system.

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

Hi there -- was on holiday so was quiet about this thread ;)

I also experience clogs with my own all-metal hot end. I think it boils down to a TOO drastic thermal break in my case (something like 40° to 230 in 3mm?). When the filament is pulled back, I guess it "freezes" immediately.

It then probably adapts and sticks so well to the inner metal tube tiny scratches that it will not resume even when I push it back by hand... I have to stop the dedicated cooling fan to let a longer part of the hot end get hot enough to resume...

I though that the filament would contract when cooling down, but if it does it is not enough... It just fits perfectly and seals the colder upper tube on a few mm. It works with a very short retraction like what does Lars but I'm left with some stringing on long hops though on my side (esp. as my bowden tube is 4mm wide inside)... :(

I still need to figure this out... My best idea so far is to make the thermal transition longer (what a pity!), so the filament will not freeze so quickly!

I am sure that we can get rid of this damned expensive PEEK ;)

 

I would definitely suggest finding a bowden with a better fit to the filament. This is a crucial element to having a direct relationship between drive movement and extrusion, and minimizing the sacrifices compared to a drive-on-print-head setup. Think of it as being similar to having very loose belts which introduce lag and inaccuracy between the stepper and print head.

Also, I think this actively cooled, stainless setup is sensitive to wall finish. If you have gouges/scratches throughout the stainless tube, it increases the potential to jam quite a bit. Use a Dremel or drill, glue some fine grit sandpaper to a cylindrical shaft and polish the tube.

 

Not exactly the answer you expect since I may have a skewed opinion on the banana blocks ;)

Now I hear the danger of over-tightening the bushings... May be the parts printed very nicely on my printer but I really had no issue. Now, my bushings are quite shorter than the "real" ones (only 16mm long, and same 11mm dia).

Still, I have really no trace of wear and the thing moves better than before (partly due to a better alignment may be), so I don't feel I would even buy harder-to-find longer ones if they get damaged in a few years...

What would be nasty is if ever some blob is left on the printed parts, that dis-align the bushing on the X/Y rods indeed. Careful inspection and cleanup is required I guess indeed.

I also initially had lips around the bushing (you'll see the "shoulder" variable in the openscad source), but it did not prove useful in my case -- I can't remember why but I thing it was conflicting with the stop switches, eg. Actually when I finally printed and used the block near the Z screw, I just left the standalone screw lose, as it looks like there is really no reason to tighten the bushing at all. IMHO I really should make a new revision without it at all.

Now I don't like much the idea of a spherical bearing, just because I would need to buy some ;) Seriously and for sure, it would add more freedom... but there is no reason if the rods are properly tuned and squared... I don't like the idea that my (short) bushing is allowed too much rotational freeplay on the rod, as I suspect it could really induce *more* wear on the rod each time the direction is changed... It still would be worth a try though because this analysis may be plain wrong -- I am fully self-made in mechanics -- ;)

 

Once the pulley pairs are indexed and locked down, there wouldn't be any real change of direction in the spherical bearings. They would just allow the small angular misalignment present when you adjust the rods to be perpendicular. I have made some nice little clip-on alignment tools, and still think it's tough to achieve alignment perfect enough to prevent all bushing bind. There is likely misalignment even in the 8mm perimeter rods (don't form a perfect square). The bearings would allow for the bushings to always be free from a bending moment against their shaft. This would definitely reduce wear and friction throughout.

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

I would definitely suggest finding a bowden with a better fit to the filament. This is a crucial element to having a direct relationship between drive movement and extrusion, and minimizing the sacrifices compared to a drive-on-print-head setup. Think of it as being similar to having very loose belts which introduce lag and inaccuracy between the stepper and print head.

 

Actually, getting back to a "tight" bowden would be problematic given the weird stuff I extrude (laywoo, trimming lines, etc...). But I agree I could switch to the appropriate one each time -- now, I never had clogging issues so far with my big bowden so this is a separate issue imho

 

Also, I think this actively cooled, stainless setup is sensitive to wall finish. If you have gouges/scratches throughout the stainless tube, it increases the potential to jam quite a bit. Use a Dremel or drill, glue some fine grit sandpaper to a cylindrical shaft and polish the tube.

 

You confirm my fears and I need to try your hint - though I did not expect a lot of success on a stainless steel tube (are many tiny scratches better than one helicoidal one?).

Also I finally received my brand shiny E3D hot end I initially expected sooner (hence my own attempt that I now want to finalize!)... It was machined to a much higher level, so I will probably try their tube on my design first.

 

Once the pulley pairs are indexed and locked down, there wouldn't be any real change of direction in the spherical bearings. They would just allow the small angular misalignment present when you adjust the rods to be perpendicular. I have made some nice little clip-on alignment tools, and still think it's tough to achieve alignment perfect enough to prevent all bushing bind. There is likely misalignment even in the 8mm perimeter rods (don't form a perfect square). The bearings would allow for the bushings to always be free from a bending moment against their shaft. This would definitely reduce wear and friction throughout.

Actually, my xy blocks could also clamp the rods with a conic shape instead of a cylindrical one as now... It would give the rod a slight rotational play without the need for spherical bearings. Not sure it would clamp it tightly but it is worth a try.

I also got the thing the wrong way in the first place, eg. "When the head changes direction, I suspect it could drag and lead to my (short!) bushing to rotate a bit on the Z axis, hitting the sliding rods each time" --> not so true since the head is dragged by the XY blocks and not the opposite. Well, I usually drag and move the head directly with the hand and I should better grab the sliding blocks instead!

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

HI Lars86, could you please share with us the drawings of the metal parts (heatsink & stainless steel tube)?

Thanks in advance,

 

Share this post


Link to post
Share on other sites
Posted · Complete Hot-end and Print Head Redesign

HI Lars86, could you please share with us the drawings of the metal parts (heatsink & stainless steel tube)?

Thanks in advance,

 

The heatsink is 1.5" in diameter, has (6) 1mm fins, spaced 2mm.

The stainless tube is basically an extended version of stock. It's a stainless M6 bolt, with a thin wall section just above the heater block (~0.6mm wall thickness, basically a removal of the threads). I truncated the threads in the heater block to reduce surface contact, and used thermal paste on the threads joining the heat sink.

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×
×
  • Create New...