3D printers - fun with FDM printers (Marlin-based)

[Hayfield]

I’ve used butanone (aka MEK) successfully with Anycubic PLA and 3mm Society chairs. It forms a good welded bond.

Cheers,
Paul

Paul

I have used Butanone a few months back, I found it made a very good bond, similar to the one with ply, but not a weld

 

[Martin Wynne]

@gavin

On the other hand, I can't post suggested profile files here if they don't work on the latest printers, and the Klipper-enabled ones are likely to become increasingly common. They also print much faster than Marlin and I know some users are put off using FDM because of the very long print times. It might be possible to speed up significantly the production of timbering bricks on the Klipper printers. The only way to find out is to get one...

... and a Neptune 4 Klipper-driven printer has now arrived, been assembled, adjusted and switched on.

That makes 9 3D printers here -- 7 FDM and 2 resin. Clearly I am in need of some remedial rehab, this addiction to 3D printers can't go on.

With Klipper there is no USB cable connection, so it can't be driven directly from Cura. Instead there is an Ethernet cable connection to the computer (cable supplied), which enables full control via your web browser, in my case Firefox. A page of very detailed controls appears, called the Fluidd Dashboard. It all seems to be working fine, just needing the additional step of dragging a saved G-code file from Cura into the Job list.

If you prefer to use a card, the tiddly Micro SD card socket on the Neptune 3 is replaced with a normal USB memory stick.

Time for a meal, then I will see if I can get a test print going.

cheers,

Martin.

 

[Phil G]

Hi Martin,
Time to rename your home 3D print farm I think
cheers
Phil

 

[Phil G]

Martin,
Does the Neptune 4 still have that magnetic metal bed plate idea from Elegoo?
After a lot of playing around with my Neptune 3 I found the bed temp must be between 65 to 70 c for PLA+ other wise you start to get a partial lifting of the piece, a brim helps but is then a pain to cut off.
I was wondering do you have similar issue with either of your Neptune's
Lastly do you have any issue if I post some STL files for very useful modeling tools on the forum maybe under a new topic, things like a Nano vise, small 3 jaw chuck and some automatic finger type clamps. non are any good for soldering for obvious reasons but great to handling/holding small objects for drilling or even gluing.
cheers
Phil

 

[Phil G]

Clearly I am in need of some remedial rehab, this addiction to 3D printers can't go on.

maybe you need some laser treatment

 

[Martin Wynne]

Martin,
Does the Neptune 4 still have that magnetic metal bed plate idea from Elegoo?
After a lot of playing around with my Neptune 3 I found the bed temp must be between 65 to 70 c for PLA+ other wise you start to get a partial lifting of the piece, a brim helps but is then a pain to cut off.
I was wondering do you have similar issue with either of your Neptune's

---

Lastly do you have any issue if I post some STL files for very useful modeling tools on the forum maybe under a new topic, things like a Nano vise, small 3 jaw chuck and some automatic finger type clamps. non are any good for soldering for obvious reasons but great to handling/holding small objects for drilling or even gluing.
cheers
Phil

@Phil G

Hi Phil,

The PEI (Polyetherimide) surface on the bendy tin printer bed will grip at 60degC and then release as it cools down below about 40degC (aided by differential shrinkage). This means that if you want to use a cooler bed temperature than 60degC for less stringing and lower part shrinkage, PEI can be a problem.*

This is presumably why Elegoo include a PVP glue-stick with their Neptune printers, although nowhere can I find any actual mention of it. For use with cooler temperatures, apply a thin film of PVP glue as you would for a glass bed.

I haven't tried doing this on the textured PEI surface, but I have used it on the smooth reverse side. It works ok, but not as well as on glass (it needs more glue on metal).

*It's also a problem if you use the "resume after power outage" function on many printers. Not much point in resuming if the power outage has lasted long enough for the bed to cool down, and the part has become loose.

---

Sure, by all means post any useful STL (or other files) you have -- that's what we are here for. Just be sure to make clear whether you are talking about something created from or part of Templot plug track, or something entirely separate of your own design. You wouldn't believe how confused some folks are getting in trying to follow all the twists and turns of these recent 3D topics. For STL files, also make clear whether they have been mesh-fixed ready for printing, or are raw STLs straight off your CAD.

Ideally, upload your files in the Resources section of Templot Club. They will then be easier to find in future and create a library of useful files for everyone. Almost nothing has yet been uploaded there, and there is a category specifically for STL files (other). Go to:

https://85a.uk/templot/club/index.php?resources/categories/stl-files-other.14/

Click the Add resource button to upload a file and enter the details. When you add a resource, a topic is automatically created on the main forum to alert everyone to the new resource and for any subsequent discussion about it.

cheers,

Martin.

 

[Martin Wynne]

... and a Neptune 4 Klipper-driven printer has now arrived, been assembled, adjusted and switched on.

@gavin @James Walters

I thought I should start by making a test piece of timbering base using the Neptune 4 printer straight out-of-the-box.

I wasn't expecting very good results, based on doing that with previous (Marlin-based) printers -- which have all required custom settings to get good timbering bases.

In the event the result was excellent, and amazingly fast, with just a few issues to deal with. I don't know how much of that is specific to the Neptune 4, and how much to the Klipper software generally for similar printers, but I suspect the latter. I think I'm going to have to eat some of my words in this topic, when referring to the latest Klipper-based printers.

To which end I think going forward I will mark this existing topic as specific to Marlin-based printers, and start a separate topic for the Klipper-based printers.

It's all part of the learning process with an experimental project, but I hope I haven't misled anyone about FDM printing options. I had tended to regard the Klipper developments as more about hype than substance, but I think I was wrong on that.

So for more about the Neptune 4 and Klipper, see:

https://85a.uk/templot/club/index.php?threads/3d-printers-fun-with-fdm-printers-klipper-based.873/

cheers,

Martin.

 

[Paul Boyd]

A request please - can anything involving external CAD be moved into a separate topic, please? It’s hard enough to follow as it is! It might also give the wrong impression to anyone coming into plug track.

Cheers,
Paul

 

[Martin Wynne]

A request please - can anything involving external CAD be moved into a separate topic, please? It’s hard enough to follow as it is! It might also give the wrong impression to anyone coming into plug track.

Cheers,
Paul

@Paul Boyd

Thanks Paul. That's exactly a point I keep making.

Now done, see:

https://85a.uk/templot/club/index.php?threads/3d-track-fun-with-cad.875/

It's hard work trying to keep the 3D info clearly explained for anyone trying to make sense of it all from the start.

I'm trying to divide it into three separate areas of discussion:

1. the actual Templot plug track experimental software -- a moving target if ever there was one.​

​

2. the real-world practicalities of making Templot plug track using 3D printers, laser cutters, etc.​

​

3. everything else -- which plug track beginners can safely ignore if they wish.​

Some things cut across all three, so that might not be the best way of sub-dividing the topics, but it's the best I can do at present to avoid getting in a muddle.

cheers,

Martin.

 

[Martin Wynne]

.
I'm wondering whether to delete all my previous posts about FDM printing, on the basis that I don't know what I'm talking about.

When I started with the BIBO printer 6 years ago, the stringing was awful. I found that the only way to reduce it was to reduce the nozzle temperature. The theory being that a less fluid material would be less likely to ooze from the nozzle. This was in line with online advice about what to do about stringing. Combined with aggressive retraction settings of up to 9mm the results were quite good. I produced a lot of stuff on there at 180degC, including some usable (just about) FDM-printed chairs in 7mm scale.

When I started with the Neptune 2S I transferred the same settings to it, and the results were excellent, even with the bowden-tube extruder (the BIBO is direct-drive), and much better than using the standard profiles supplied with the printer. So I used the same settings again on the Kingroon printers, with similar good results, and again much better than using the standard profiles supplied with the printer. All these printers use the Marlin G-code interpreter in the firmware.

Fast-forward(!) to the high-speed Neptune 4 printer (using Klipper firmware), and I was surprised to find it printing PLA at 220degC with a small extraction of only 0.8mm. I knew those settings on the BIBO would produce significant stringing. But the results from the Neptune 4 were very good, with no stringing. Initially I put this down to differences in the Klipper firmware, or maybe in the extruder design for high-speed printing.

But maybe not? Today's experiment was to try the same settings on the Kingroon printer. 220degC, 0.8mm retract, 60degC bed constant. And the results from the Kingroon are very good (using exactly the same STL file), with no stringing in the sockets:

It took longer of course -- 81 minutes (using 0.16mm layers) instead of 38 minutes on the faster Neptune 4 (using 0.12mm layers). The Kingroon would be even better (but slower) at 0.12mm layers, so I will try that next.

The possible explanation is that a more fluid material in the nozzle from the higher temperature allows much faster pressure changes in the nozzle, so that a short sharp retraction is all that is needed.

And a conclusion from all of that might be that you can get good results at each end of the PLA temperature range, but not in the middle of the range?

So now I don't know what to think. Should I delete all the lower-temperature profiles I have posted previously and replace them with 220degC profiles? Or suggest that folks need to find their own best settings and ignore any chunterings from me?

Answers on a postcard.

Martin.

 

[Hayfield]

Martin

Thank you very much for your experimenting, your test piece looks so much better. I now look forward to seeing the next test shot

Thank you very much

John

 

[Phil G]

Hi Martin,
Just a though, has the PLA+ also improved over the last 6 years as well? when you look today at the PLA rolls They all talk about the correct working range being 205 to 235 degrees centigrade exact values being brand dependant. I can't comment about temps 6 years ago, because I was not interested in 3D printing at that time.

Modern trends which your calling klipper based firmware, are all about faster printing speeds, which require a more molten and more responsive polymer, its follows to increase overall speed, short sharp retraction values help.
All of which, in turn requires a lot more rapid cooling after the the filament has been deposited.
The question is, was it the PLA that allowed these advancement? Or was it simply realising that's the only way to print faster?

I don't believe you should delete your previous temp posts, after all there is a lot of useful information contained within which we know will work.
Equally there is a horses for course scenario to consider, where your findings (Cura slicer recommendations) will not always be replicated verbatim by others. Either on the same or different types of machines. Maybe that should also be acknowledged.

If you enjoy experimenting with printer settings, than I for one see nothing wrong with that, or posting your results.

On a slightly tangential topic,
I do wonder if you will also start to experiment with other slicers
After all fast Klipper firmware and the Orca slicer certainly seems to be an up and coming trend, when looking to print at faster settings I mean.
Please note I am not suggesting you should. Only musing if you will
Cheers
Phil,

 

[Martin Wynne]

I do wonder if you will also start to experiment with other slicers

@Phil G

Hi Phil,

I'm not really looking for a new hobby of 3D printing ins and outs. They are just a tool for getting jobs done.

As far as the plug track bases are concerned, I would very much like to be using the Simplify3D slicer rather than Cura:

https://www.simplify3d.com

S3D allows different printer profiles at different levels. For plug track only the top 2" scale needs to be printed at the highest quality, below that the print could be done at a courser quality which runs much faster. This would significantly reduce print times, and allow the final top layer to use the finest settings available (but slow).

Cura can't do that. It has a post-slicing script available which can edit the G-code by layer, but it's very limited in what it can do, and tricky to set up.

The snag is that S3D costs $200 and Cura is free. It's no contest for most users.

cheers,

Martin.

 

[Phil G]

I would very much like to be using the Simplify3D slicer rather than Cura:

Hi Martin,
Yes I am very awere your personal preference for a slicer is Simplify3D. which makes sense, however as you have posted its a paid for product.
I also fully understand your main aim here, is to find the best tool for plug track, and it was in that context I was thinking with my post.

I guess the good news is, no matter which free slicer we use, the current Cura or a possible future change, there all trying to reach the standard of Simplify 3D. All be it a slower than would be ideal for us.

The very obvious next step for any slicer is the multi or variable printer profiles for one job, this in theory at least, would help any project on the build plate in all respects.

As you say its outside the reach of the current version of Cura, it can't do that, its also hard to program such a thing for Marlin v2.1, firmware but maybe not so hard for Klipper firmware?

I am also awere there are a myriad of posts online, all of which basically say, "whist Cura can be used to slice for klipper, your never going to get the best out of Klipper if you do that.
This does make total sense, after all both Cura and Marlin are from the Uni-maker stable So why would they take Cura beyond Marlins capability .

Other free slicers are just starting to use the much better processing power avaible when using klipper, which is its real gain, although currently it seams its only targeting faster computational speeds to the stepper motors!!

Now if you assume, which most posts I have read do, mean other slicers are starting to get the best out of klipper, by making more accurate projects.
Then it does follow.

My musing was only will you dabble with another slicer? I know that's a tangent to plug track, but there again so was a klipper based printer.
Maybe it's far too early, and the multi layer capability, would need to be marketed before there is an logic to looking at other slicers. I don't know.

I was not asking you to try anything, Simply wondering outload if you maybe had considered it.
Cheers
Phil,

 

[Phil G]

Hi Martin,
Very much out of interest, I started to do a google search on adaptive layers, it sounds as though Cura does have a very elementary version of adaptive layers, it sound as though you have very little control on when and how the different layer thickness values are applied.
However after further searching I found the attached you tube clip, on this it sounds as though it's a much more advanced version of adaptive layers than the Cura version. The clip is using Prusa slicer, If you get time maybe worth checking the clip out,
just to see how you think it compares to Simplify 3D.
If you do watch it I think you may have to slow the speed slow down a bit.

Please note I am not advocating Prusa slicer here at all, simply enquiring if you think free slicers are getting closer to the Simplify 3D version. When it comes to adaptive layers.
Cheers
phil,

 

[Terry Downes]

.
I'm wondering whether to delete all my previous posts about FDM printing, on the basis that I don't know what I'm talking about.

Hi Martin, I don't see that there is any benefit in deleting historical posts but, I'm wondering whether there is a better way to communicate/share updates, config files etc. I notice that there are no posts or files within the logically positioned Resources Tab.

Maybe, We should create/encourage a couple of easy to read matrices/spreadsheets here. If You agree then, I could begin by uploading/sharing my settings for my AnyCubic Cobra Plus FDM printer.

 

[Martin Wynne]

Maybe, We should create/encourage a couple of easy to read matrices/spreadsheets here. If You agree then, I could begin by uploading/sharing my settings for my AnyCubic Cobra Plus FDM printer.

@Terry Downes

Hi Terry,

Yes please.

That's what the Resources section is for, although I suspect many members never think of looking there. That's why I have set it up so that any additions automatically create a linked topic in the forum.

You can also search the Resources from the Resources tab:

https://85a.uk/templot/club/index.php?search/&type=resource

But a proper A-Z index would probably be more useful. I will see if such a thing is available in XenForo.

cheers,

Martin.

 

[Steve_Cornford]

Speaking personally I am not too bothered about speed, just consistency of results. As Martin has mentioned previously, there are lots of other modelling jobs to be getting on with once you have set the FDM timber base print going. It obviously pays dividends in taking time to get your brick design accurate in the first place.
Steve

 

[Phil G]

Speaking personally I am not too bothered about speed, just consistency of results. As Martin has mentioned previously, there are lots of other modelling jobs to be getting on with once you have set the FDM timber base print going. It obviously pays dividends in taking time to get your brick design accurate in the first place.
Steve

Hi Steve,
This post has me a bit baffled, whilst I don't disagree with your sentiment, you have lost me as to the purposes of posting it. Are you talking klipper vs Marlin or something else?
If your talking about the you tube link I posted. Then I fear you may have missed the key point. It was never about printing faster, its about being able to vary the layer height to get a much more accurate product. Something Simplify 3D can do, and something Martin has commented on previous posts, would be advantageous to plug track production.
Cheers
Phil,

 

[Phil G]

Hi Martin,
I have just uploaded into the resource page, A 4mm scale code 75 knuckle bending jig, it consist of 7 stl files and a PDF with instructions for assembling and using the using the code 75 rail knuckle bending jig.

I also have these files as STP files and was going to upload these as well (to allow anybody with a cad knowledge to rescale to other gauges) or maybe make improvements.
unfortunately its reporting, STP files are not supported in the resource area.
I am happy to post these, but where too I don't know, could you please advise?
cheers
Phil
PS only posting here because of the recent posts with Terry above.

 

[Martin Wynne]

Hi Martin,
I have just uploaded into the resource page, A 4mm scale code 75 knuckle bending jig, it consist of 7 stl files and a PDF with instructions for assembling and using the using the code 75 rail knuckle bending jig.

I also have these files as STP files and was going to upload these as well (to allow anybody with a cad knowledge to rescale to other gauges) or maybe make improvements.
unfortunately its reporting, STP files are not supported in the resource area.
I am happy to post these, but where too I don't know, could you please advise?
cheers
Phil
PS only posting here because of the recent posts with Terry above.

@Phil G

Hi Phil,

Sorry about that. STP and STEP files now included in the Resource options.

Many thanks for uploading the files -- I have added a screenshot in the description.

cheers,

Martin.

 

[Steve_Cornford]

Hi Phil,
In the video you posted in the first 30 seconds it mentions printing thicker layers to save time where you dont need detail. Call me old fashioned, but i equate saving time to printing faster.
Actually I was really referring to Martin's post which mentioned 81 minutes printing 0.16mm layers on the Kingroon (which I have) and his statement

The Kingroon would be even better (but slower) at 0.12mm layers, so I will try that next.

I would assume that printing at 0.12mm would take 33% longer?
I should have been more specific, sorry.
Steve

 

[Steve_Cornford]

Hi Martin
In Templot, Shapes > Boundary > pre-set brick boundary dimensions, is it recomended to define a boundary size that will fit just inside the perimeter fence that has been defined in your machines Start G-code ?
For instance on the Kingroon KPS3 Pro S1, x=188mm, y=189mm
Steve

 

[Martin Wynne]

Hi Martin
In Templot, Shapes > Boundary > pre-set brick boundary dimensions, is it recomended to define a boundary size that will fit just inside the perimeter fence that has been defined in your machines Start G-code ?
For instance on the Kingroon KPS3 Pro S1, x=188mm, y=189mm
Steve

@Steve_Cornford

Hi Steve,

Still so much to explain and write about when I'm still learning myself.

Do you mind if I use para numbers -- it helps to clarify my thoughts.

1. On the Kingroon Pro S1 the physical bed area is 210mm x 210mm.

2. It would be good to be able to use all of it, but the usable printer area is limited to 200mm x 200mm, i.e. leaving a margin of 5mm all round.

3. This printable area is set in the printer firmware. If you try to exceed it the overlap will simply be scrunched back into the printable area and wreck the edge of your print. The limit could be changed by modifying the firmware config settings, but that's not a simple thing to do.

4. For levelling I set a perimeter priming run at 190mm x 190mm, i.e. a further 5mm all round inside the printable area. It would have been better at 150mm x 150mm with the corners directly over the levelling wheels, but that would have been too much of an obstruction to some test prints.

5. Running a timbering brick over the top of the priming run doesn't matter too much. It is only 0.2mm thick and the timber webs are 0.96mm thick in 4mm/ft scale. If the priming run obstructs any sockets or connector clips, it is easily cut away with a craft knife.

6. However, if you are getting consistent good results and not needing to adjust the levelling every time, it's a waste of space and filament to do a perimeter run every time. Consistent good results means you can come back to the printer after a gap of a week or two and get good results straight off without adjusting anything.

7. The easiest way to omit the perimeter run is to create a second custom printer with all the same settings as the first one, but with different Start G-code. To create a second custom printer just do the same as for the first one, but give it a different name.

Mine is named Templot KN ProS1 - no perimeter. For how to set up a custom printer in Cura, see:

https://85a.uk/templot/club/index.php?posts/7361

Set the size as 200mm x 200mm and the most important setting which is on the Extruder tab at filament diameter = 1.75mm.

8. Choose whichever of the two printers you wish to use on the top-left dropdown list, in PREPARE mode. Without a perimeter run you can set the brick boundary rectangle at 200mm x 200mm. Or maybe 199mm x 199mm to allow for the FDM shrinkage settings.

9. Even without a perimeter run it is still necessary to prime the nozzle before starting a print. Here is some Start G-code you can use for the second printer. This runs up and down the extreme left edge of the print area, so ideally keep the brick content 1mm away from the left edge of the boundary rectangle.

;Kingroon S1 prime left margin line
M413 S0 ;Disable power-loss recovery
G28 ;home
G92 E0 ;Reset Extruder
G1 Z1 F500 ; up
G1 X0 Y5 Z0.1 F500.0 ; to start position and touch - prevent backlash
G1 Z0.3 F500.0 E1 ; up to priming line
G92 E0 ;Reset Extruder
G1 Y195 F1500.0 E10
G1 X0.4 F1500.0
G1 Y5   F1500.0 E20
G1 X0 Y0 Z0.1  F500.0 ;touch origin again for backlash
G92 E0 ;Reset Extruder

You should be able to copy and paste that to replace the original Start G-code after creating the new custom printer. It is also attached below as a text file.

10. Also attached is the latest profile I have been using on the Kingroon. I spent a lot of time comparing test prints after my recent experiments with 220degC and decided in the end that I preferred the results from my original 190degC settings, with 60/45degC bed and 0.12mm layers. But I reduced the retraction to 0.5 seconds (2.5mm @ 5mm/sec), which saves some time. I also saved some time by increasing the first layer to 0.36mm instead of 0.24mm. If you import this profile, don't forget that you must then select it in the drop-down list in order to use it. It will appear in the list only if you have one of the custom printers selected.

• Please note everyone that these settings are not the final word on the subject, or even what I definitely recommend. Just the latest settings which I have been using recently and getting good results with PLA+ filament on the Kingroon printers. It's quite possible I may find some better settings in due course.

cheers,

Martin.

 

[Phil G]

I would assume that printing at 0.12mm would take 33% longer?
I should have been more specific, sorry.

Hi Steve,
Firstly there is no need to be sorry, I too am no expert on FDM printers or adaptive layers. I also agree the clip does talk about faster prints amongst other things. The key take out for me was you need a slicer that allows the person to decide where the layers will be thinner (.008) or thicker. I believe up to 0.35 or maybe a bit more.

With the very limited option, Cura gives you, on adaptive layers "accessed from the experimental section". (It's limited because its fully automatic from what I have tried) It also seems as though its only about saving time not improving quality.

On the other hand the Prusa slicer has the manual adjustment slide bar on the right hand side of the screen, here you can over right the automatic feature, to add thinker or thinner layers where you want them, Which has to be at least an improvement.

Also as I understand it there is more than just saving time at play here. It's desirable to have the right thickness of PLA exactly where you need it, too much can be almost as bad as too little. I believe (again not being an expert, just read a lot) that's all about how the PLA cools and the possibilty of internal stresses caused by poor cooling thus deformation. The outake is too much plastic can ruin your model.

I know Martin has the simplyfy3D which he says is better, but costs money, so I was just curious if the Prusa version of adaptive layers came even close to the S3D version? Given Prusa slicer is free software I for one would give it a go if there were gains to going that route.
cheers
Phil

 

[Steve_Cornford]

Hi Martin,
Thank you for your very detailed reply and the attatchments. I will give them a go, although I have started to get fond of my growing pile of perimeter fences
Steve