As promised in my last post, here are some photos of the weight plate collar and the new regulator flow knob. Also a photo of the old knob for comparison.
Dacor Fury small knob
Dacor Fury new larger knob
Weight plate collar (the yellow spot marks an original collar)
This is a set of base brackets I printed to hold my PiDP8 kit at a better viewing angle (design from Thingverse)
It’s time to make some fall glass. Today at noon I started up the glass furnace. It’s a slow process: it takes 8 hours to heat from 100F to 900F at 100F/hour, then 16 hours to get to 1300F at 25F/hour, and then a further 6 hours to reach holding temp of 1900F.
Once it’s at 1900F, I’ll start charging with raw glass batch. The batch I use is Spruce Pine, or SP87 which is the classic batch for glassblowers. I should have a fresh pot of clear glass early next week. Then it’s time to make some nice things.
I made a few things this past week, and designed (and made) a few others.
First, I didn’t like the stark white of the new Dacor Fury flow adjustment knob, so reprinted it in black. I also flipped it so the square hole was on top for a better print of that side. When done it fitted perfectly and has already been tested in the water. It works perfectly.
I loaded cave line onto the dive spool and also took it in the water. It held up perfectly but time will tell how it holds up.
I also designed and printed a replica of the weight plate ring for my weight bench. The ring is a complex plastic part that sits in a hole in the weight plates and keeps things from banging loudly when in use. Each plate requires three (on for the center bar and two for the side slide bars) and many were missing or broken when I bought the (used) weight bench. I tried contacting the manufacturer in eastern Canada, but the bench was made in 1995 and they had no records for it at all! I designed one in Tinkercad and it printed perfectly. It’s now on the 50lb plate so that it can get some everyday use to see how it holds up before I print more.
I also designed and printed some clips for the cats’ litterbox. It’s 10 years old and the originals broke. The printed ones fit perfectly and so far are holding up very well.
Finally over a week ago now I designed some small plastic ‘buttons’ to cover assembly holes in a set of drawers Linda bought for her sewing room. You know the ones – the small white plastic 1/2in buttons that go in the screw assembly holes in pre-pack particle board furniture. This one was missing all but one (of 8) so I used it as a template to make 9 new ones. Some had to be white glued in place as the holes had enlarged, but all are on now and look great.
I notice that my photos weren’t ready for uploading, so I’ll add them soon. (New reg knob, loaded dive spool, weight ring and litterbox clips).
Lastly, some news on filament. The white AMZ3D filament was working very well, with only occasional snags. Thinking about it, I realized the spool did get loose during the coathanger holder creation. I decided to open the AMZ3D black filament and be very careful loading it. As a result, I’ve been able to print many things, some with over an hour print time, without any snagging at all. It would seem that a) this AMZ3D filament is awesome stuff, and b) loading procedure is critical – you cannot get any loosening of the filament when loading or you will get snags. I really like the AMZ3D white, but I absolutely love the AMZ3D black. I even ordered another roll.
After printing the regulator flow knob, I was rather pleased with myself. Then the black filament ran out while printing a coaster, so I changed to the new AMZ3D filament I bought recently. I decided to try the white. The spools are smaller, both the hole and the width, so my current ABS pipe holder would not work.
I tried printing “from the box” as I’d done with the MG Chemicals filament, but it didn’t work with the smaller, side opening box. Instead it jerked the z axis at the start of a print and threw off the bed adjustment!
I managed to get that fixed, but took the opportunity to change the bed mount system to now use nylock nuts for greater vibration resistance. While I was at it, I again ran into the mess that is setting the z endstop.
So I decided to design a new adjustment system for the endstop. It uses the same new z endstop I printed a while ago and have been using with good results, but added a screw adjuster.
I designed the plastic to hold the screw using Tinkercad, then printed it. The first one wasn’t correct, but allowed me to see the problem and correct the design. The second was perfect. I mounted it on the non-motor x axis end that rides the vertical bar. One problem I discovered is the screw could not possibly touch the z endstop microswitch. I needed a lever, but turning the screw would spin the lever. I needed a round lever, a.k.a. a washer. Finding a suitable washer, I assembled the z endstop adjustment system and in the first test it worked perfectly. Better than that, it is now easy to make micro adjustments to the z endstop for quick adjustments of the extruder height. Photos attached.
Then I needed a new filament holder. During these test prints, I used a holder I made quickly from a bent metal coathanger (see photos) clamped to the table. It worked but looked like crap.
Today I designed a new spool holder, also out of a coathanger and a length of broomstick, but modified it based on what was wrong with the first one. It works great, and again photos attached.
All in all a great end of the week.
OK, I’m brilliant.
I have a SCUBA regulator that’s missing the flow adjustment knob. Years and years ago (2006?) I bought a primary second stage for my open circuit regulator set from a shop in Calgary. I got it for $100 even though it was brand new because someone unscrewed the flow adjustment knob and then lost it. I’ve been diving it that way ever since because you could not buy that knob for love nor money.
Last night before I went to sleep I thought… “why not 3d print a knob?”.
The knob fits a square shaft 7mm x 7mm x 2mm and is sized to match the regulator. It attaches with a 3mm stainless hex bolt. Today I designed and printed the knob, and with a bit of drilling to open the bolt hole, it fits and works perfectly.
The last many posts have been text only, without the benefit of photographs. To compensate, here are a set of photographs of the “keepers” – 3D printed objects that have turned out well.
Yesterday I spent quite a while printing with great success, until the last print of the evening. By now the printer had been laying down filament for a few hours, and about 3/4 through printing a coaster, I heard the dreadful clicking sound that indicates a fouled nozzle. I immediately shut things down and called it a night.
Later on, while considering why the filament would jam after all the corrective actions I’d taken, it came to me.
My printer has one fan, which came as part of the extruder assembly I had purchased. Seeing there was a fan connector on the RAMPS board, I connected the fan there and found the slicer software would control that fan. The two settings were “always on” and “auto”. I had chosen auto.
Reading about auto mode indicated the fan speed was a function of print time per layer, determined during the encoding process. I thought this a bit odd, but figured it would be fine.
I asked my friend which setting he used, and indicated that his extruder fan was wired directly to 12v, and that the slicer setting was for a print cooling fan, not the extruder fan. Further reading of the RAMPS manual verified this. The parameters affecting the auto setting now made more sense to me, as you would want more cooling on a print where the layers are taking longer to print.
Further reflection also suggests why I was having problems. First, all my jams occurred after much printing, when things had been hot for some time. With the extruder fan not running all the time (auto setting), the heat from the extruder hot end would gradually travel up the metal feed tube. Eventually, the filament in the tube would become warm further from the nozzle. When this happened, it would be harder to feed the warm filament and it could resist and jam, much like trying to push a piece of hard spaghetti vs. cooked spaghetti. The purpose of the fan is thus to keep the feed tube cool so the filament remains cool and feeds properly.
Later today (after my dive) I will rewire the extruder fan to the 12V supply so that it’s running whenever the printer is on.
I have been experimenting with print settings for my 3D printer, to overall good success, until things failed.
The printer uses 1.75mm filament, which is heated and then extruded through a 0.4mm nozzle. However, it is possible to set the layer height to any value from 0.1mm to 0.4mm through the slicer software. The default was 0.3, but lower values are supposed to create smoother and more detailed prints.
Last week I stared printing with 0.2mm layer height, and things were going well. The prints did seem more detailed, but then disaster. During one print, I started to hear a clicking sound from the extruder. This is not good as it indicates the extruder is jamming somewhere and not feeding properly.
Checking the print I could clearly see that it had stopped printing. I killed the print and had a closer look. Filament was coming out the extruder gears where it should be feeding the heater & nozzle. In the end I had to completely disassemble the extruder so I could access the stepper motor, gear, pulley and filament jam. The extruder stepper was also very hot, which is not normal.
It took an hour, but I was able to disassemble, clear, and then reassemble the extruder. Unfortunately while reinstalling it I broke one of the printed PLA clips. I was able to glue it, but would need to print a replacement in the future.
With the extruder cleared, I printed the 20mm box without problems, so then started to print an octopus model. Midway through the print, the extruder started clicking again. This time I killed the print before the jam was bad and was able to remove the jammed filament without difficulty.
However, it left me with a problem: why was the extruder now jamming when it had been working fine for weeks?
I searched the internet for “filament jam” and found several common causes for jams, but only one really seemed to apply to my situation.
I had noticed the base layer was thin, and wondered if the nozzle height was too low. This is one common cause of jams, as the low height prevents smooth flow of filament in the first layer. After testing I realized the nozzle needed to be raised by 0.1mm or more. This is not easy with my current z endstop, but there’s a parameter in the slicer software to include a “z offset”. I did this and the base layer was much better on some test prints.
However, this was not the case before the first jam. I had checked, and removing and reinstalling the extruder had changed the nozzle height just enough to cause the problem. But that did not explain the first jam.
I therefore reflected on the classic problem solving approach, “what changed?”. The one thing I had changed was lowering the layer height from 0.3mm to 0.2mm. As I considered this, it seemed to me a puzzle – why would you think you could get 0.2mm from a 0.4mm nozzle? Should not the nozzle diameter dictate the layer height, not the other way around? You do want adhesion between layers, so it seems logical to use a number like 0.3mm for a 0.4mm nozzle, just to squeeze the layers together a bit. I had been using 0.3mm for weeks without problems, so the logical next step was to return to 0.3mm and see if filament jams ended.
I did that a few days ago, and since then have printed more objects than I have in the weeks before. Every object has printed very well, with good quality, and NO JAMS. So 0.3mm it shall stay. I even printed the new extruder clips and have installed them.
I’ve even stared playing with “vase mode”, where prints are a single layer thick and a continuous spiral of filament. The vases are paper thin, but really fun to make.
So it would seem that I’ve solved my filament jam problem.
Last week I designed another gear item for my camera and underwater housing. The 60mm focus gear works perfectly, but I was curious if my 17-85mm zoom lens would work in the housing with the current port.
The problem with ‘just trying it out’ is that while I can put the lens on the camera and fit it in the housing & port, I know that at 17mm there will be vignetting as it’s not a dome port. What I really wanted to test was whether or not it could zoom enough to be useful. However, just trying to install the lens zoomed might not fit. What I needed was to be able to actuate the zoom with a proper zoom gear.
Which meant designing and printing a proper zoom gear. Based on my focus gear design, I created a new model based on the focus gear cylinders, but sized for the zoom lens which had quite a bit larger diameter. The zoom lens also has a noticeable bump where the autofocus switch is located. After a few test prints, I completed the final design and printed it. The first print was very good, and fit perfectly.
With the zoom gear complete, I was able to install the lens on the camera (with zoom gear) and then into the housing on wide angle. Once in, I could clearly see the vignetting, and then tested the zoom gear. Again, it worked perfectly, but the lens quickly bumped the end of the port. Sadly, it was nowhere near close-up, and still showed vignetting. In the end I will not be able to use the zoom lens in the underwater housing with the current macro port, but if I ever win a lottery and buy a dome port, it will work perfectly.
In other news, I’ve been watching a lot of youtube videos on “delta printers”. These are 3d printers that employ a triangle pillar system to control three stepper motors, which in turn move a trapeze supported on the three pillars. It’s a very fluid moving system, and most fun to watch. There is one new version that has one extruder but three filaments, and can mix colors on the fly!
Watching build videos it’s clear that some of these printers are extremely well designed for kit building. The only downside at the moment is there are very few inexpensive kits that offers both heated bed AND multiple extruders. The best kit I’ve found so far (for build quality) comes with only one extruder and no heated bed. While the heated bed is pretty easy to add, conversion to multiple extruders is a very advanced topic. The other downside is many come direct from China, and that’s always a bit of a crapshoot when it comes to the thing actually arriving at my door. So far I’m around 90% but still… to lose a $300-$400 kit would not be good.
It finally happened. After the past several years of apple trying their darnedest to lose me, they have.
I bought an iphone 4 when they came out and loved it. But I refused to pay $1000+ to upgrade to another, especially with all the nonsense about changing connectors and such. My iphone4 has officially been an orphan for a while now. Other than hard crashing every few weeks, it seems to be OK. So as long as it continues to ‘work’, I’ll keep it, but I can see cheap android in my future.
The ipad2 I have just seems to work worse and worse with every app update or ios upgrade, so now it barely able to find the network that a Raspberry Pi has no trouble with. Apps keep getting larger and large with each update. I imagine it won’t be long until it’s orphaned as well. I have no plans to replace it.
With all the orphaning, apple was also intent on making app creation a worse and worse experience. Objective-C was a horrid brute, but it was a language. When they junked it for swift, I refused. I’ll learn something open source, or widely useful, but I won’t waste my time on something so totally proprietary. Last year when the app developer renewal rolled around, I declined. Besides, even with app developer my one and only iphone/ipad app would no longer compile and run on my two devices, so what was the point?
That left my 2012 apple macbook pro (17in version), which was heavy but really nice to use. I had a decent hard drive, max memory (not much, but max) and used remote desktop to control it fairly well from my main PC with the big screen. It too ran slower with each OS upgrade, but at least it was based on freebsd (unix) and I liked it.
Then last week, the final nail in the coffin. I require a VPN to mark exams from my university, and I’d been using openconnect on the mac. It worked and was open source. Sometime between the last week in July and the second week in August, the university broke the link between their VPN (pulse) and the mac. After considerable diagnosis, I was forced to admit defeat and install the pulse client on my win7 box. It worked perfectly.
That meant the last reason to keep the mac running was gone. With the obvious disdain apple seems to have for all things desktop, including the horrible compromises that they’ve been offering as laptops, I had to conclude that apple does not want people like me buying or using their stuff. So I’m gone.
I doubt I’ll buy another apple product again as it truly appears the clock has come full circle and the sculley years have returned. Like the sculley years, they now have more crap and flash than actual product. And as SNL’s Gilda Radner used to say of the phone company, I feel apple saying to me “We don’t care. We don’t have to. We’re apple”.