|
Post by scotttucker on Apr 22, 2022 0:04:14 GMT -7
There is an incredibly impressive 3D model of a Zeiss projector on a website called digitalcentrality. Sadly it looks like the owner is reluctant to share the file (I may be wrong). In any case, I was inspired to start learning 3D CAD to make planetarium related components and perhaps a model or two too. After messing around with Tinkercad which is a great introductory tool (but really limited) someone recommended Autodesk Fusion 360. Amateurs can download free license/copy for Fusion 360 for non-commercial projects - I think it is limited to 10 editable projects at a time. The package is incredibly powerful and is a real professional level product. With that comes a very steep learning curve. There are plenty of youtube tutorials but I am old fashioned enough to have bought an introductory book. I will freely admit that I am at the very slow clunky clumsy stage of using it. Perseverance is a pre-requisite but as with all computer packages the basics functions are not too difficult to learn but the refinements take weeks to get to grips with. Playing with it seems to be the way to learn. I attach a screenshot of my first attempts at the beginnings of a projector model - it is clunky and not accurate but it will gradually get there. When it is in a more advanced state I will share the printable file with folks here. As with all my projects you really should not expect speed!
|
|
|
Post by Ron Walker on Apr 22, 2022 10:41:54 GMT -7
I will try and find the posts that the fellow did here who made the beautiful model. Can anyone remember under what topic that was posted???
|
|
|
Post by scotttucker on May 6, 2022 13:39:38 GMT -7
Progress is slow, but I am gradually learning the techniques. One thing that has been surprisingly useful- do you remember the thread on the original OC about co-ordinates of centre points of the truncated icosahedron? Well you can incorporate those co-ordinates into the model - it makes spacing out the starplates in the correct pattern over the surface of the sphere much easier. I have adjusted the model since the original post.
|
|
|
Post by scotttucker on Jun 10, 2022 1:45:07 GMT -7
aaargh - 3d printing is so frustrating. I only seem to be able to manage about 24 hours of printing before I get a blockage and the print fails. So far I have been eliminating the following problems:
1. Rubbish filament x 2 - the filament I was using just has too much friction to feed freely through the bowden tube (the teflon tube that guides it to the extruder) and the other cheap one is dimensionally poor 2. Poor quality bowden tube and fittings - replaced 3. Poor nozzle replacement practice (did not realise you had to tighten the nozzle by one full turn after the bowden tube is abutted up to it.
I have ordered a direct extruder mount to try to eliminate all my problems with the bowden tube and a good quality filament.
|
|
|
Post by Ron Walker on Jun 10, 2022 10:58:41 GMT -7
My son in law is having very similar problems. A print will get so far and then it appears the next layer of filament does not melt into what has been laid out before. He has replaced many of the things (if not all) that you list above.
|
|
|
Post by Ron Walker on Jun 10, 2022 11:04:49 GMT -7
aaargh - 3d printing is so frustrating. I only seem to be able to manage about 24 hours of printing before I get a blockage and the print fails. So far I have been eliminating the following problems: 1. Rubbish filament x 2 - the filament I was using just has too much friction to feed freely through the bowden tube (the teflon tube that guides it to the extruder) and the other cheap one is dimensionally poor 2. Poor quality bowden tube and fittings - replaced 3. Poor nozzle replacement practice (did not realise you had to tighten the nozzle by one full turn after the bowden tube is abutted up to it. I have ordered a direct extruder mount to try to eliminate all my problems with the bowden tube and a good quality filament. Could you post some close up pictures of what you are talking about above. Especially the extra tightening of the nozzle as well as the direct extruder mount and how it works differently then the original.
|
|
|
Post by scotttucker on Jun 12, 2022 0:37:48 GMT -7
Hi Ron The direct extruder conversion is pretty simple and really just designed to eliminate most of the bowden tube and the problems associated with it. The standard description is that the bowden tube is the equivalent of pushing a rope down a hose pipe - plenty of opportunity for it to not go well so if you mount the filament driver directly over the hot end it should be less prone to problems (diagram should show principle). The longer the bowden tube, the more error there is when the printer momentarily retracts the filament when the nozzle transits to a new position. The problem is made worse when the little pneumatic fittings starts to wear and loosen their grip and the whole tube starts moving slightly. In practice I bought a conversion kit (lots of them on e-bay or Amazon) which is just a replacement y axis carriage with a bracket positioned so you can unbolt the existing extruder drive (red on my printer) and bolt it directly over the hot end. The kit also contains extension leads for the stepper motor because it will be physically further away. So far it is pretty straight forward, I hope to finish it later this morning. The matt filament I was trying to use was just getting stuck in the bowden tube. As soon as it got a bit sticky, the driver wheels would squash/ deform the filament a bit as it worked harder to push it through which just led to more sticking.
|
|
|
Post by scotttucker on Jun 12, 2022 0:49:00 GMT -7
As for the nozzle gap / nozzle tightening procedure - a really good explanation can be found at : community.cookiecad.com/t/how-to-fix-under-extrusion-caused-by-bowden-gap/209In a nutshell, if there is a gap between the upper surface of the nozzle and the end of the bowden tube inside the hot end it gets all clogged up with charred filament. A way of minimising the problem is to unscrew a clean nozzle by one full turn, install the bowden tube, then fully tighten the nozzle creating a little pressure between the upper face of the nozzle and the tube to make a good seal. This was definitely happening in my printer and probably the cause of most of my failed prints.
|
|
|
Post by scotttucker on Jun 12, 2022 2:43:51 GMT -7
Just finished bolting everything up and running a test print as we speak. Will let you know how it goes. It looks quite promising and there are none of those horrible clicking noises you get when the extruder skips a beat when the friction gets to much. I can't do any of the big prints for another week as I am away for work next week and could not risk leaving it for so long.
|
|
|
Post by Ron Walker on Jun 12, 2022 17:23:50 GMT -7
Thank's for the info. I will be looking forward to your experiments.
|
|
|
Post by scotttucker on Jul 15, 2022 13:29:22 GMT -7
I am very slowly building the model. I realised that actually, it has forced me to really look closely and examine photos and drawings of the projector to understand how it is put together. I think previously I had mentally seen the mid section of the projector as an impressive jumble of protrusions and parts - I have found it really interesting to need to tease out the dimensions of each of the parts and have more of a feel for it now. The model is not perfect but it is getting to the passable state.
|
|
|
Post by scotttucker on Jul 15, 2022 13:34:24 GMT -7
I am too impatient to wait to get a 'perfected' model before printing it. I am currently using Hatchbox filament which in my limited experience is much better quality than the more standard brands and is the only one that manages the 36 hour starball print. As outlined above, I have also swapped to direct extrusion. The component count is rising!
|
|
|
Post by Ron Walker on Jul 15, 2022 19:34:00 GMT -7
Very nice work.
One of the problems we are having is that after the print has been going for a while, the filament will not stick to the build so far. Once a filament does not stick, the project is history. Talking with Michael he has learned to stay away from PLA and has been having great luck with ASA. He also just told me that his big problem with the filament not sticking was that he was making several things at once and when it went from one to the other it would often not stick to the existing print. He is now getting much better prints.
|
|
|
Post by scotttucker on Jul 15, 2022 22:32:34 GMT -7
Agree - I stopped trying to print multiple small parts at once a while back. Somewhere on youtube there is a video with a few suggestions on how to recover the situation when you return to the printer and find the nozzle waving aimlessly in mid air and a half completed print below.
You can remove the failed part from the bed, measure the height of the piece so far with vernier callipers then use the modelling software to split the model in two at that point and simply print the remaining part of the model and glue them together. Not very elegant but it did work for me on on occasion.
A more elegant approach is to leave the failed print in place, measure the height so you can figure out what z axis height you have reached then go into the g code file and delete everything between the initiation lines and the correct z axis height. Then start the print again - takes a bit of g-code bravery and deletion of many, many lines of code but seems like a better solution than binning it and starting again.
|
|
|
Post by Ron Walker on Jul 16, 2022 12:33:05 GMT -7
That is like g-code Greek to me. Just like the codes for the plastic filament.
|
|