Jump to content

danilius

Dormant
  • Posts

    362
  • Joined

  • Last visited

  • Days Won

    8

Everything posted by danilius

  1. If you can upload the model, I will take a look at it.
  2. This will probably come as a surprise to most CAD users - but I use Blender for a lot of CAD work, simply because it's the tool I use the most for my day-to-day work, and whatever package you use, the files end up as polygons anyway. Also, I can show the client a photo-real render or animation before I print out whatever it is I have designed for them, all in the same program. Here is a warning though: Blender has an insane learning curve. Yes, it does organic shapes brilliantly, can sculpt, and even does CAD in a strange way, but it is very, very difficult to get up to speed with it quickly. Why mention it then? Because I produce useable and printable designs with it very quickly, since it lends itself to be used as a CAD sketchbook, so to speak. A basic idea takes very little time to create. With experience, you will be able to refine the initial "sketch" and improve it until it is done. So, Blender is not your obvious choice for CAD, but you might want to consider it for the above reasons. Oh, and it's open-source and has a great community.
  3. A nice option would be to be able to specify geometry as being support, that way one could create a solid object that CURA will convert into support. Sometimes CURA generates no support at all even though it is needed, or the supports generated are a poor choice. This is hardly surprising since we cannot expect the programmers to be able to figure out every possible scenario, and it means making custom support much, much easier to remove.
  4. In the video you posted it looked like the bed gap (i.e. first layer height) was way too off. I bought feeler gauges off ebay for £2, and they are a worthwhile investment, but learning to level the bed and getting that gap just right is critical. Usually, once you have levelled the bed and fine-tuned the gap, you will not have to redo that for months. A good trick to help doing this is to first loosen the screw at the back of the bed - there is only one in the middle. One full turn should do it. Then print a 10cm hollow square with a brim that has around 20 - 30 lines. This will allow you to fine-tune the bed levelling by eye with something that is actually printing. You do the adjustments for all three screws while the priting is happening. This will help make things very accurate. For larger print jobs, I have the bed tuned well enough to allow printing the first layer of PLA onto glue stick at around 35mm/s, 0.2mm layer height.
  5. In the latest firmware if one starts a material change, then cancels and immediately starts a material change again, my UM2 tries to push the head off the front, and makes a horrible grinding noise.
  6. I use two different types of glue on the heated bed - UHU glue stick and some really cheap no-brand garbage. For smallish models the cheap stuff is perfect, and once the bed is cool the models usually come off very easily. The only thing is that overprinting the glue lasts only a few times, and then things get harder to pull off. So, I usually clean the plate once or twice a week and re-glue when the plate is hot (65C). The UHU glue I use for bigger models and ABS. The problem is that the parts stick on to the glass like they will never come off. Mechanical removal is fraught with the possibility of slicing off essential parts of one's anatomy, or at least bits I still want to hang on to. So then I resort to freezing the glass and model once the glass is at room temperature, and that works a charm.
  7. I would sit by the machine and watch it printing. No need to print the whole thing, just the bit that sticks up, with some support at the base. That will give you a big clue as to what is going on. It does not look like its wobbling, since the horrible gloop is only on the corners, and below the bend the outer corners are glooped, and above the bend the inner corners are glooped, but on the bend itself everything is relatively OK. But the layers appear to curl up, and increasingly as the height goes up. How thick are the walls? The layers curling can be easily explained as lack of support, so that part wobbles more and more as the height increases. You can make your own break-away supports very easily, which will keep that part rigid during print and not use too much filament. I made the ones below for a really tricky model (arm sticking out on one side, other stuff as well) and embedded the tips around 0.15mm into the model. They snapped off very easily and worked a charm - I printed 12 copies of the same model and it worked every time. Of course it did leave a small mark at every snap off point. The horrible gloop could be the fans running too low or off, but it seems a bit weird you are getting that only on the corners of one side which then flips over as the bend is crossed.
  8. Cloud based slicing makes sense if you can offer the user a superior experience; say waaaaaay faster slicing, and more/cooler options than they would get with CURA or its ilk. In most cases, user who want better/faster/cooler slicing want it on their computer because of irrational reasons, like ChrisR (not saying you are irrational, but you are talking about a "feeling" thing, not something concrete). It ties in with what I wrote before: the New Economy is about putting things into people's hands, not trying to tie them up in user agreements, paywalls and services they simply have little if any use for.
  9. This is going to take all of 10 minutes to hack and rip the gcode right off the Arduino. So I for one would not trust this for a second with my IP if I was concerned about it. Secondly, there are very few designers around compared to the rest of the world's population. Most of the new, upcoming breed of designers are being fed a steady diet of open-source and free distribution of IP for non-commercial use. So this site will appeal to a very small niche of designers, who already have various avenues for selling their designs, such as their own paypal-driven websites. And they rely on the buyers to be honest and not distribute their source files. But these are few and far between. Then you have thingiverse and youmagine which are flooded with so many designs that getting your site noticed is going to be a really tough sell. So, what the 3D printing revolution is really about is not simply printing, but a change of culture in the distribution of two types of wealth: money, and knowledge. It is a New Economy, a socio-capitalist blend in which the correlation between money and knowledge is much looser, less defined and relies more on people playing nice than enforcing rules. I know how to build a 3D printer, I will tell you how to do it, and if you manufacture one and manage to sell it - good for you. Send me a beer. This is really set to undermine the stranglehold the vast conglomerates currently have over our economies. By distributing knowledge freely (Ultimaker being a superb example) and allowing everyone to have a go at making money, everyone benefits at some level, and the distribution of knowledge prevents the Makerbots of this world from running off with everyone's pooled knowledge and dominating the market. Now that I have my political manifesto off my chest, you can understand what I mean when I say your website is Old Economy trying to leverage some money out of the New Economy. It's a good trick if you can do it.
  10. That's a seriously nice design. It would be interesting to see the sort of accuracy/repeatability you get with this setup as-is.
  11. They are going to shoot you for that, and I will stand by and cheer :-P Super news then. Is it too early to ask what the price will be?
  12. I'm coming a little late to the party here (force of habit :-P ), but where is this thing up to? Is anyone selling the block yet, or is it all still hush-hush?
  13. On the matter of openness, I commend you. I think you guys (and gals) are doing a great job there. The decision to drop dual head printing makes a lot of sense. I do not disagree. The quality of the machine is superb. In general, I still think Ultimaker is a great company with a great philosophy and great machines. There are two very different things here: telling everyone what's cooking (very important, especially in a young company like Ultimaker) and promising something. One has nothing to do with the other. The UM2 was marketed with the dual-head upgrade as a future option. In hindsight I think you can agree that this was a big mistake. What Ultimaker should have said was: "We are working on a dual head solution for the future. Right now, it does not work. There is no guarantee it will ever work. But right now we are confident it will, and we are investing time and money on it." By "work", I mean to the quality we all expect from Ultimaker. Now, I don't want to flog a dead horse. Ultimaker made a mistake, and that can happen. I am not angry. I am not annoyed. I am let down, because Ultimaker is setting the standard for the New Economy, and I wanted to be a part of that. Now I feel a little foolish. So maybe I am more annoyed with myself than anything else. I fell for a marketing gimmick which I thought would have no place in the open source community. In your place I would ask myself: what can I do to make this up to my customers? No, this is marketing-speak. We know what went wrong, by and large. That was very clearly explained, and I'm sure we all agree that this was the correct decision. You need to come up with something concrete and tangible, although what it should be I have no idea. Perhaps others who feel the way I do can come up with something that makes commercial sense for Ultimaker and assuages the customers. On the other hand, if only a minority of customers feel this way I would be agreeable to ignore the whole thing and move on. I can be quite grumpy at times. :-)
  14. For plastic, a cheap set from ebay will be fine. I still have my Vallorbe files from my days as a bench jeweller, and you can buy a whole set of jewellers files fro the price of one of those. On plastic, the difference is almost unnoticeable.
  15. I have to say that when I came across this I was very disappointed for two reasons: 1) I bought the Ultimaker with dual-head printing in mind (not as the main reason, of course, but because I expected to have it added on later at a price) 2) I would have missed this entirely had I not come across a mention of this as an aside in an entirely different thread. This is a major decision, and I would have expected to have received an email to that effect, at the very least. It would be wise of Ultimaker to learn a lesson from this: do not promise anything you cannot deliver same-day. So, if you do not have a perfectly functioning dual-head UM2 in your labs, don't market your printer saying that it is dual-head upgradeable at some point in the future, because now you have lost a great deal of credibility with just about everyone. You have to look at it in this way. Your unique selling point for the UM2 is the quality/speed combination. There are printers that can print as quick, and there are printers that can print as well, but not both. Certainly not for the same kind of money. Backing all this is your openness and commitment to behaving in the spirit of open source; i.e. none of the corporate behaviour we have come to love and know from the likes of Apple, Google and Microsoft. So, you dropped the ball on this one in a big way. Now you have people like me who still really like the UM2 but feel quite badly done by. Not because I paid for something that you have not delivered - as I said previously I would have quite happily paid extra for the upgrade - but because you failed to deliver on a promise. This means that your trustworthiness has taken a huge dive. Of course you should strive for a quality experience, and if you cannot deliver on the UM2 that's fine. What upsets me is that you sold me a product with a promise that you have now violated. This does not detract from the UM2 itself as it is. It is still a wonderful machine, and delivers superb quality. But the promise - this should have never been made. And I bought your machine on the basis of that promise. Had you never mentioned dual-head printing, I might have gone for a different machine such as the Taz 4. To be honest, I don't really know, because that option was never on the books while the promise of a future upgrade lured me in. Can you make this up to me? I fail to see how. On the one hand, you never charged me for this upgrade, so you owe me nothing. On the other hand, this was a big factor in my choice. So part of the reason I invested in your company (only part, because by and large I really like the way you guys and gals are going) has now been dashed. This has no tangible value. All in all, right now I feel quite upset and let down. I trusted you guys, and you failed me very badly. You desperately need to re-think your strategy.
  16. I'm printing on a UM2 in Manchester, and I keep the printer in the same room I work in so it's at ambient room temperature most of the time. Overnight the room temperature drops to 10C - 15C but overnight prints are fine as well. Mind you, all of this is with PLA. You might want to try using UHU glue stick on heated glass, I find that the prints stick really well but are tricky to get off after the print! The fridge trick seems to work most times though.
  17. What I have found is that the filaments I use vary wildly in their performance, even from the same manufacturer. So, some PLA is quite happy at 200C if I'm printing at 0.1mm, and will have to go to 215C when printing at 0.3mm (both of which I do fairly frequently); all this at between 30mm/sec and 60mm/sec. Some require way more heat to print at 40mm/sec at 0.3mm layer height, as much as 245C. When printing something really big that I need in a hurry, I run the filament as fast and as hot as possible until the smell changes from popcorn to burning plastic, then I dial down the heat a tad until the feeder stops clicking. Essentially you are pushing molten plastic through a small hole; so to get more through (printing faster/thicker layers) you have to increase the pressure. For a given temperature there is only so much plastic that will go through. So, the only other thing you can do is increase the temperature. Even if the plastic is rated for 220C, when printing thick layers quickly, the plastic will never hit the indicated temperature. That is to say that the nozzle might be 250C, but the plastic is going to be at a lower temperature. So don't be afraid to push things a little on the temperature front. Also, with a blockage, I usually dial up the temperature really high for a few minutes then hand-feed some material through, drop the temperature and feed some more. Or reverse hand-feed it from below, which is messy but convenient! The upshot is that every spool will be different to the next, and it's a matter of perhaps trying things really slowly and cool at first, and then pushing things as you get more used to it. A trick I use for tight spools is to wind it onto a bigger spool (oh so boring and time consuming, but we ll worth the effort) and then I leave it on a hot radiator for several hours for the plastic to anneal. I buy branded and garbage, and in the process of rewinding the garbage usually starts going white where it is stressed, but the comes nicely in the wash :-). It probably makes more sense to anneal the plastic first, then wind it onto the bigger spool and anneal again.
  18. I use Blender and FreeCAD. Blender is very mature, and since everything that will be 3D printed is going to be composed of polygons anyway after export to STL and OBJ, you gain little in the end product if using a solid modeller like FreeCAD or SolidWorks. Blender also has a vast collection of tools, and using modifiers can be used to successfully create both organic and engineering models in a parametric fashion. Although it's not a solid modeller, it's absolutely fine for engineering projects if you like to work that way (which I do). For complex, high-poly projects you will need a powerful computer, a powerful card and lots of patience. Blender has a steep learning curve. It's well worth the effort, though, since it can create things you would tear out your hair trying to produce in SolidWorks. FreeCAD is maturing. It has many useful tools, is much easier to learn than Blender, is a solid modeller and a great CAD tool. Under the surface, if you are handy with Python, you can do almost anything you want including assemblies and what have you. When do I choose between one or the other? Well, if the project is purely technical or engineering then FreeCAD is my fist choice. If there is an element of design then I go to Blender first because I "think" in Blender better than I do in FreeCAD, and Blender can do things that FreeCAD can't. The sort of projects I have used Blender for are things like a model submarine, jewellery (mainly rings), a smart watch case (very high precision), a couple of current projects one which is quite large and complex (all parts are printed in scale) and one which is really small but very precise (25 microns roughly). With FreeCAD I have designed and printed a barreling machine, various simple bits and bobs, and a really small and very precise IP67 electronics housing. So if you want free tools, they do exist, and they are perfectly usable for most if not every thing you will print on an Ultimaker. What's important to bear in mind is that you choose the tool according to your skills and the job at hand, not the other way around.
  19. I have found that sometimes the filament oozes out sideways in the head, and gets jammed there. So even if you try to yank it out of the feeder it will simply not budge. So what I do is to remove the bowden tube from on top of the print head, and then slowly wind the filament backwards using the "move material" menu option, and with some care snip off the end of the filament that will probably have a little shoulder. Then back goes the bowden tube. What you have to bear in mind is that the motors are not all that powerful, so if the filament gets stuck and the motor won't pull it out, chances are that the filament is stuck somewhere. Pulling on the too hard on the filament is probably going to cause problems. So generally speaking, you should not need a great deal of force to extract filament manually. If you do, then pulling hard is not going to resolve the problem.
  20. I pointed this out to them, and for some unfathomable reason they felt that the standard deviation for 21 reviews came out almost identical to 140 reviews. I am not a statistician of their calibre clearly, but this seems a little implausible.
  21. I'm no expert, but I have bought no-name PLA of ebay at a quarter of the price of the colorfabb filament I previously purchased. Compared to the stuff I got with the Ultimaker (translucent blue) and the colorfabb filamanent, the ebay stuff either performs equally well or outperforms both of those by a huge margin (luck of the draw, basically). With one roll I bought I can print waaay faster than with either branded filaments (the table was shaking too much, so I had to slow down the Ultimaker) and issues like stringing are far fewer and the cheap stuff outperforms the brands for bridging by a wide margin. Mechanical performance seems pretty similar; i.e. the cheap stuff seems to do the same thing as the brands. So, YMMV. I'd say, take the plunge and try out one reel. They are a third to a *less than a quarter* the price of the brands if you know where to look, and you might just be surprised by the results.
  22. In which case, why not call the new version "Cura Xterminator 2014 Skynet Edition"? Sort of a hint to the clued-up types who have already figured out the plans from the name "Ultimaker".
  23. The regular feedback and openness reinforces my decision to have purchased an Ultimaker 2 (aside from the insane speeds I manage to squeeze out of it and still get quality prints). Well done to you, and to Ultimaker, for not taking the old-economy route of trying to cover up everything.
  24. Even though almost every point regarding open source has been made already, I will add my tuppence worth. Open source means the customer can trust the company. If the company is sharing their source, they are confident that they have got the product right. I bought a UM2 since I knew that all the plans are available, so should they go bust I can still repair my machine. So the customer gets that degree of comfort as well. It means that I can modify my machine based on the UM plans, even if it's just to build a camera arm/holder. Way more effective than reverse-engineering. Open sourcing your product on an open source platform provides you with an ethically clean look. Even if you plan on taking over the world and subjugating humanity to your fiendish schemes, looking ethical will get you there sooner. Lastly open source means getting free R&D. The best R&D is done by the users actually using the product and discussing the options. See the https://www.youmagine.com/designs/alternative-um2-feeder-version-two.
×
×
  • Create New...