Jump to content

geert_2

Ambassador
  • Posts

    2,084
  • Joined

  • Last visited

  • Days Won

    34

Posts posted by geert_2

  1. Thanks for the feedback, I appreciate it.

    I have never used Faberdashery PLA, so I can not comment on that. But it is entirely possible indeed that some PLA-brands are more sticky than others. If they stick reasonable to an untreated glass (no glue, no salt, nothing), then they are likely to stick better with the salt method too, in my experience. And vice-versa.

    Concerning the glue stick: it seems that different brands give different results. For my difficult models, I had poor results with the default glue stick in the beginning: corners did lift, but fragile models got damaged when removing them.

    That is why I searched for an alternative method and came up with the salt method, which (for me) did not have both problems.

    But that was before I read that you could spread the glue out with a wet tissue, and it may have been a different brand of stick. And it was long before I read about neotko's method of using window-cleaner to remove models.

    So I think there is no "best method" that suits for everyone, every filament, and every model. But there are lots of very good and usable methods for different circumstances. So you should try a few and then use the one that suits your models, your environment (moisture, temp, etc.), and your filament best. And important: choose one that you like. :)

  2. I think they look really great. They look like high quality casts rather than 3D-prints. :)

    Concerning a few occasional imperfections: if you look at bronze statues, they also contain lots of imperfections: scratches due to removing moulding or welding seams, circular defects from the removal of support pins from the moulding, sanding marks, washed-out details because of the polishing, and so on. While it may hurt to the eye because you know they are there, I consider it part of the job.

  3. I have never changed jerking or acceleration settings, so I can't comment on that. But generally, printing slower (and thus also cooler) greatly reduces this effect: for small high quality parts I use 20 or 30mm/s instead of 50mm/s.

    Also, some materials hide this better than others, if they are slightly translucent. But not too translucent, because then you start to see the inner print lines too much.

  4. In my UM2 (non-plus) with original firmware, the calculation goes wrong when the tune-menu is kept open. For example when manually adjusting temp. If this tune-menu has been open for some time, then the remaining time sometimes increases to hours extra (on a print of a few hours). It looks like the calculation is paused or so? After the tune-menu has been closed, remaining time slowly recovers to reasonable values. But I don't care, since it doesn't affect print quality. :)

  5. ...

    @geert_2 - you've answered all the other questions brilliantly as usual. May I take this opportunity to say thank you for all the times you have helped me. Your posts always pop up whenever I search the forums!

    ...

    Cheers,

    Matt.

    You are welcome!

    In my experience, helping others is well spent time: it requires me to think about things, and to see the situation from lots of different viewpoints. Whenever a solution is found (it doesn't matter who finds it, me, one of the other people here, or the original topic starter), we also know that anwer and we can use it four ourself. So this is a very fast way to learn, with a high "return on time investment". :)

    • Like 2
  6. ...

    I wonder of the coating is conductive?

    ...

    If you would like to try it, maybe you could ask a couple of manufacturers to send a few worthless scrap parts with defects or so, for free? Just to evaluate the surface finish and quality.

    I have also wondered if this is a layer of metal, or metal flakes, chrome powder, or some special optical effect in plastic? But it seems that it has to be covered with a sort of varnish, so probably won't be conductive afterwards.

    I also wondered about scratch-resistance: better or worse than normal paints? The old method in which a shiny chrome layer was put on with vapor, in headlamp mirrors, was extremely weak and susceptible to scratches. You couldn't even wipe it with a soft cloth. But this method seems to be different?

  7. This is for an UM2, but maybe it can be used for an UM3 too?

    I disconnect the bowden tube at the nozzle, and pull the tube out. This breaks the thin strand of filament. Then I move the filament forward a bit, and I cut off the irregular tip like this:

    DSCN5251.thumb.JPG.0fb04621832303211d12c08aa7c7d534.JPG

    And then I manually pull the filament out from the back. Sometimes this can be done while the printer is on, but sometimes the motors block this, and I have to switch the printer off.

    In this way I never have pieces getting stuck in the tube or feeder. Of course, I learned this the hard way. :)

    Then I do one or more atomic pulls to clean the remains out of the nozzle, and replace bowden tube.

    To load, I insert the new filament manually, which goes easy with the tip cut-off in an angle, and using "maintenance > advanced > move material" I move it until it fills the bowden tube and is extruded. Then a small retract. And that's it.

    It is more difficult to describe than to do.

    • Like 1
  8. Hoi Sander: the concept of spanners that grip on the flats of bolts, instead of on the corners, exists for years in hardware tools. Especially in car shops, since those bolts and nuts get corroded and dirty, and require huge forces to remove.

    But then it still takes someone to have the bright idea to teleport this concept to 3D-printing, where it is equally needed. :)

    There do also exist similar spanners that can handle both metric and imperial bolts and nuts, based on this concept, the so-called "metrinch" spanners. Next time I am in a hardware shop, I will pay more attention to it. :)

    For example, see:

    https://sites.google.com/site/harkosworkshop/technical-posts/bihexsocketsvsinglehexsockets

    http://www.metrinch-tools.com/website/introductie.php

    • Like 1
  9. I never go over 210 to maximum 220°C for PLA. This would cause the PLA to get burned, thus worsening the amount of burned material inside. If you have printed with PET or ABS before, and you do the cleaning with PLA, you might try higher temps indeed to melt their residu, but only for a very short time. If you use nylon to do the atomic pulls, you will need a higher temp. It depends on the materials.

    The brute force of the traditional atomic pulls is indeed why I developed this more gently method. I did displace the nozzle and teflon coupler a few times, and I worried about bending the rods, or doing other damage. There is no such risk with the gentle method.

  10. "M4" defines the thread's outer diameter in millimeters, in this case 4mm. Thus a bolt with an M4 thread will just slide in a hole of 4mm diameter. An M6 will slide in a hole of 6mm. The thread will fall through, not lock. Thus the inner thread of a nut is smaller, to get the screwing action, but I don't know the dimensions by head.

    The outer dimensions of the hex heads are bigger: about 7mm from flat to flat for an M4.

    Google for "metric thread dimensions", for tables with the full specs. These also show the required drill sizes before tapping the thread.

    If you want dimensions of plastic screws, see the site "www.skiffy.be". This is the biggest European manufacturer of nylon and other plastic screws. Their site shows all dimensions per product.

    @ purps: your somewhat rounded "flats" are an excellent idea. I have seen hex wrenches with such shapes, but it has never occured to me to do use it in 3D-printing too. Next time I need one, I am definitely going to try this.

    For "moulding in" nuts and bolts in a 3D-printed model, another option is using openings with small undercuts, or a sort of snap-fit clips. So the nuts can be pushed in with brute force, but they can not fall out. Then the models can be printed unattended.

    • Like 1
  11. To make the site faster, and to reduce your server-load, I think it might be a good idea to make all animations and all films "on-request", instead of autoplaying. Provide a button which the user has to push. And only then the film or animation should be loaded from the server and executed. Provide a static low-resolution JPG-image as placeholder: this loads very fast.

    The first time people arrive at your site, they will want to see the animations. But after they have seen it once, it becomes an annoyance.

    Also, mobile devices may not have enough resources and RAM to play videos or big animations; this could be the reason it doesn't work for some.

    Further, I would suggest you use standard fonts that are available on every computer (Windows, Mac, Linux), instead of loading custom fonts from the web. This will also greatly reduce loading times, and server load. And it will increase stability.

    For example, define fonts as: "Arial, Helvetica, sans-serif". Or similar. I prefer a boring Arial that loads fast and is guaranteed to show up correctly, rather than a nice font that is slow and may not display correctly on my system or browser.

    • Like 1
  12. A bit late, but still: in car tuning shops, they do have spray paints that give a chrome or gold effect (or any other desired colors with mirror-effect, as in christmas balls).

    On Youtube you find a lot of demo videos. Search for "spray on chrome", or "spray paint mirror chrome effect", or similar.

    It seems to consist of cleaning, spraying on an activator, rinsing with water, spraying on chrome, rinsing with water again, spraying on a sealant varnish, rinsing, drying with compressed air. It also seems to work well on flexible materials, even leather.

    For example these:

     

     

    • Like 1
  13. On my computer they work. Windows 7 Pro, SP1, Pale Moon browser (=derived from Firefox, but still has the menu-bar and statusbar).

    But it is going very slow and stuttering, and the CPU is at 100%. So it seems that they consume a huge amount of resources. This might be too much for some systems, or if running in a virtual machine.

    Are these defined as videos, or animated GIFs? If videos, I think it would be best to remove the "autostart", and add a start-button per video. And set the video to play once, and then require another manual restart if the user so desires. This would greatly reduce resources-consumption, I think.

    • Like 1
  14. I am glad the salt method works for you too.

    @prb4: try this link: https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/

    First thoroughly clean the glass, and then clean again with plain warm tap water only (no soaps, no window cleaners). This outside of the printer, of course.

    Then I just put a few drops of salt water on a paper tissue, and gently wipe the glass plate. Gently keep wiping while the water evaporates. So it leaves a thin mist of salt stuck to the glass, nice and equally distributed, but almost invisible. It looks like the glass is just a bit dusty.

    On the second application, I just redo the wiping in the printer, without prior cleaning, and without taking the glass out.

    The disadvantage: you have to wait until the models cool down, to get them off. And it requires a heated build plate.

    For me this works very well for Ultimaker and colorFabb PLA, and a bit less but still okay for ICE PLA. But it does not work at all for ABS. For PET it works a little bit, but less than dilluted white wood glue (gr5's receipt: ca. 1 part glue in 10 parts water).

    If you try other materials, let us know the results (or the lack of).

    • Like 1
  15. I am not sure, but can't you do without spring? And replace it with a piece of metal of fixed length and suitable design? So that you can still finely adjust the position of the pressure wheel, to get the desired bite-depth of the teeth into the filament?

    I have always wondered why there are springs in the feeders of almost all printer brands? Yes of course, officially they are needed to push the filament against the knurled feeder wheel, so it gets a good grip. But does it?

    If the spring tension is too high, it flattens the filament into a plate, instead of a round. Then the filament gets stuck in the bowden tube, or nozzle, or at least causes high friction. If spring tension is too low, it causes slipping, or the feeder-teeth jumping out of the filament, instead of biting. So we are supposed to find the nice spot in-between those extremes. But is there always an optimal spot in-between?

    If the feeder would have fixed setting-screws, so we can finely adjust how deep the teeth bite into the filament, but without spring, then I guess that should reduce both effects: it should stop flattening the filament, as soon as the teeth have bitten into the filament for the desired depth (because at that moment there is no more pressure)? And it makes it more difficult for the teeth to jump out of their bite-pits (as there is no soft spring that can give way)?

    Or am I overlooking something?

    Maybe in Robert's feeder this would be easier to test than in the original feeders?

  16. What happens on the next layers if you continue printing? Does it keep extruding so irregularly, or does it level out nicely?

    I sometimes see a bit of irregular extrusion too in one of my printers, but only on the first layer, although far less than yours. Then it smooths out. So I don't really care, as it is only one layer.

    In some brands or colors of filament, this effect is much stronger than in others. Also, when the nozzle is leveled rather close to the bed, to get a smoother bottom layer, the effect is stronger.

    I am not sure what causes it, but I guess that in the beginning the temperature is not yet totally stable, or that there are variations in molten volume in the nozzle, since you don't have a nice steady stream of filament yet in the beginning. I guess it is like in a chemical plant, where you always have process variations and swings on startup, until all distillation columns and equipment is on temperature, and a steady flow is established? Then things smooth out.

    But of course, there could still be other causes, such as a worn-out coupler, or variations is filament diameter, or filament that has trouble sliding through the bowden tube smoothly, so that pressure builds up irregularly, and then it jerks forward? Or something similar?

    What happens if you print a skirt of 30 lines around the object, just for test?

  17. Normally nylon can easily be dried in an oven.

    But I have the impression (this is a guess, not necessarily correct!) that nylon for 3D-printing sometimes contains modifiers to reduce moisture absorption, or to chemically bond it, or to mechanically bond it in the molecular structure (like the gels in pampers)? Maybe this could cause this?

    Anyway, store all filament in a closed box, with a huge bag of disseccant (the sort with color indicator: blue=dry, pink=moist) that is also used to dry car interiors, and that can be dried in a microwave. For really sensitive materials, keep them in such a box while printing: make a spool holder in the box, drill a small exit hole for the filament, and place that closed box behind your printer.

  18. I don't know Tinkercad, so I can't say anything about it.

    If the purpose is for 3D-printing, I would suggest that you try DesignSpark Mechanical. This is a free but limited version of SpaceClaim. It is very easy to learn, and there do exist a lot of good tutorial videos. The basics can be learned in a few hours. Intermediate level stuff takes a few more days. It is distributed by the big RS-components electronics company. It does require registration, but if you want to do a whole classroom, maybe you could find a solution with RS to make that easier?

    If the purpose is only modeling, but not 3D-printing, you could also try the older SketchUp versions, which were freeware. This has a similar same push-pull interface as DesignSpark Mechanical. But SketchUp does not generate good "solid" and "water tight" 3D-models, so it causes trouble when 3D-printing. Also easy to learn.

  19. Excellent!

    You can't start early enough to get kids interested in technology.

    In traditional education it is sometimes forgotten that the purpose of studying is to be able to actually use that knowledge. The only real knowledge is being able to do things, thus "active knowledge".

    So-called "passive knowledge" - thus recognising that someone else can do something, but you can't - is no knowledge; it is only the first step towards knowledge.

    I believe that all education (math, physics, chemistry, language, whatever), should be given with practical goals in mind. If students understand what they can do with the subject, in their own life, they will be much more interested.

    • Like 2
  20. Have a look at the nozzle while printing. It could also be that molten material accumulated on the outside of the nozzle, especially on small objects, or when you have a bit of overextrusion. And then that starts to burn, sinks, and gets deposited at random spots on the print.

    In my experience, thick light-brown blobs usually come from the outside of the nozzle, while thin black flakes usually come from the inside, with PLA and PET. But this may be different for other materials of course.

    Anyway, regularly doing an atomic pull never hurts. And you should definitely do one whenever changing filament.

  21. I don't know the size of this car. But if it would be rather large, it might be cheaper, easier and safer to buy a completely new toy?

    It might take days to get a few wheels 3D-printed, and cost a lot in shipping. And they are likely to be less strong than injection moulded wheels.

    To estimate the costs, you could have a look at pricing at specialised 3D-printing services, like Shapeways (UK), Materialise (Leuven, Belgium), or 3D-hubs. They give quotes based on dimensions, material, size and volume. I guess you should expect a few hundred euro?

    Another solution, which you can easily do yourself, is remove one wheel, make a silicone mould of it, and cast it in flexible polyurethane. There are a lot of very good tutorial videos on mould making and casting on Youtube, including all required material specs, etc. This would probably be the cheapest option if you want to keep this old toy. And it would be the most educative.

  22. An idea that just occured to me: what about making the prime-method adjustable in Cura? For example let people choose between:

    - Prime as extra skirt lines around model (same thickness as first layer).

    - Prime as spiral (same thickness as first layer).

    - Prime as blob.

    - Prime in the air.

    In the last three cases, with additional parameters to specify the location where priming has to occur:

    x=...,

    y=...,

    z=... (height above glass plate only for priming in the air)

    All with the same material-volume of priming as now.

    For each printer, you could set the defaults that work best in your testing lab. And for old printers, you could chose as default the method with which the printers were delivered. For example: "prime in the air" for the UM2.

    In this way, people can easily adjust the priming method to what suits their situation best. Maybe there does not exist a "universal best priming method", but only "good priming methods for specific circumstances".

    These settings could best sit under advanced stuff, to not overwhelm new users.

    Could this be a solution?

  23. And how does plaTec behave over a longer time?

    Normal PLA gets harder and more brittle: after a year or so, it may snap easily when stressed (like you would do with hooks, keychains, etc.). And it may lose its shine and get duller.

    Does this also happen with plaTec?

  24. If you measure on a modern power supply ("switched mode"), then make sure it sees some load: use a lamp or a suitable power resistor (heater) as load.

    Most modern power supplies do not work or do not work correctly without load: they may output nothing at all; or a far too high voltage. For a correct measurement, it would be best to use a load similar to the real load it will see when in use in the printer. Connecting a few car lamps might do. And then at least you have some visual feedback too. :)

  25. I guess you are always going to have a temperature difference in the plates. A huge temp difference between glass (+60°C, here almost 100°C) and room air (~20°C) causes a strong upwards draft: it will pull in cold air from the sides, and blow off hot air upwards in the middle of the plate. Like the "thermals" outside in nature, in which glider planes keep circling: they can be strong enough to lift a plane from 500m to 1000m in a few minutes. So that will cool down the sides much more than the middle. But the distribution is pretty nice and even, I think.

    If this ceramic glass would cause less warping due to temp differences, that would be good for accuracy, especially when printing in the corners of the bed.

    Concerning the prints coming off: have you tried different bonding methods? Like dilluted Elmer's wood glue (gr5-method: 1 part glue in 10 parts water), 3DLAC, strong hairspray (neotko-method: spray on tissue, wipe glass plate), my salt method (for PLA only: wipe glass plate with tissue moistened with salt water), or ABS-aceton slurry (ABS-only)?

    Normally, some of these bonding methods should work well. Otherwise there may still be something else going on, such as the build plate sitting too far away from the bed?

×
×
  • Create New...