Sunday, December 28, 2008
Some new High Temperature (Ceramic) Materials for Thermoplastic Extruders
This is my first post to the builders' blog, so I'll keep it short. I've had some success with a new (to me at least) ceramic cement. I was able to reverse engineer (i.e. RTFM) the components of a commercial high-temperature cement and buy them in quantity, for a lower price. The binder is sodium silicate (which Viktor had already experimented with), the thermally-conductive filler is zirconium silicate powder. Both of these are available from ceramics/glaze suppliers, at lower cost than buying from a chemical supply business. More detail is available in my new blog.
Photographs are (for now at least) online in a picassa album
(I'm not having any luck embedding the photos in my blog; please bear with me as a beginning blogger. Please let me know if either of the above links give you problems.)
-- Larry Pfeffer
ursine @t gmail d0t c0m
Wednesday, December 24, 2008
Extruder Madness Part Deux!
The goal was an extruder with no specialty parts. Everything can be obtained from your local hardware store which is exactly what I did. The only tools I used was a cordless drill, a vice, a bench grinder, a small bastard file, and a 10-32 tap. Keep in mind that this does not include the heater barrel which is a separate problem that I'm addressing.
So, here's the extruder. Yes, it extrudes quite well. Put a couple feet of 3mm ABS through it with no problem. Hadn't even bothered sharpening the threads yet but that would certainly help also.
This is basically a take off of the lasercut extruder design, the 608 bearings and such, but without any need for the specialty parts. The design isn't as heavy as it looks though it does have a bit more weight than the Mk II. For my machine it wouldn't be a problem though ymmv.
Here's how I put it together. Very little text from now on.
Cut head of of 5/16-16 x 5" screw.
Assembled the 608 bearings (Drive Bearings 1 and 2) on the drive screw.
Took a 1/2" aluminum cable crimp, drilled and tapped it 10/32 for two set screws and drilled one end to fit the extruder drive screw and the other to fit the GM3 gearmotor shaft.
Next I took three small L-brackets (maybe one inch on a side) and bent them as follows. I then drilled out their holes to fit an M6 shoulder bolt. This along with a shoulder bolt and some washers makes up the Pressure Bearing assembly.
I then assembled the frame to which everything mounts from two 3" L-brackets.
Then came the attachment clamps for the drive screw Drive Bearings 1 and 2. Drive Bearing 1 is held by plumber's tape and a piece of steal and Drive Bearing 2 is held by a pipe clamp. I would only make the plumber's tape ones in the future. It held better.
And that's it, your done!
Keep in mind this is not a finished extruder design but simply a starting point for better designs and more detailed builds. I test this by clamping it to the desk and hooking up the GM3 gear motor. Worked like a champ. Total cost for all parts was something like $20 US. Total time to build once all the holes were drilled and the pieces were bent was about one hour.
Adjustments concerning the height of the 608 bearing form the drive screw--the one that presses the filament into the threads which I call the Pressure Bearing--can be made by "shimming" the 608 bearing held by the pipe clamp--Drive Bearing 2. I did this by inserting small pieces of scrap tin between the frame and Drive Bearing 2 so that it pushed the drive screw incrementally closer to the Pressure Bearing. This works very well and is very quick.
Please, if you have a good idea for improvements, fixes, etc., then please let me know.
The photo shows one of the last prints from the machine. The object is straight off the table with no post finishing.
The machine is a Bits from Bytes Laser cut machine running one of their extruder's. I have done little to modify the machine, just make sure everything is straight and true. The extruder has two gear motors driving the spindle and a few detail changes to the hot end.
Control is PIC32 running a G-Code Text file straight from an SD card.
Temperature as measured on the nozzle 210DegC
The head is moved at 8mm/sec and this is matched to the filament extrusion rate.
This setup is made with a 2" flat washer sandwiched between two 3/8-16 square bolts on the heater barrel. In the flat washer are two M3 holes which will be used to hold the heater a specified distance away from the extruder body. Basically, the air has become our thermal break thus changing the slow thermal gradient to a very fast thermal gradient.
I then threaded the ABS filament down through the pump body and into the heater barrel assembly. I was sorta doubtful about this setup, it seemed too easy, and I was right.
Upon initially heating up the barrel, the filament had a tendency to pull to the side once the plastic got warm enough to deform if there was any lateral forces cause by misalignment or by a curvature in the filament itself. This isn't a show stopper, however, because a guide can be constructed to hang down from the extruder to align the filament a few millimeters form the heater opening.
Once the heater got up to temp--which took longer and required more power due to the huge increase in surface area presented by the nuts and washer--I turned on the extruder and this is what happened:
Friday, December 19, 2008
Wednesday, December 17, 2008
Developing HDPE milling know-how on Tommelise 2.0
There is nothing like actually milling things to learn how to mill things... do you want to read more?
Tuesday, December 16, 2008
Frank back up and running...again...almost
I've detailed Frank's vital statistics in the Builder's Wiki for everyone to see here. Frank is a McWire 'strap running Gen 1 electronics (full stepping firmware 0.8 host). My biggest problem with Frank has been the extruder--like most of us. For a long time I let Frank rest in the wings as I was working on the OIT RepStrap and didn't have time for him. Recently I got around to working on him again and wanted to clear up the rats nest of wires around the boards and tiding things up before tackling the Extruder issues again.
This turned out to be a great idea that was long overdue. Here is a picture of Frank pre-facelift:
And here is FrankNuvo!
There are several key changes in FrankNuvo. First of all there is the re-arranging of the boards. They are all lined up now and all th e wires run underneath them. Everything has plugs attached whereas before some things were hardwired in. The boards do not go directely out to the steppers or the opto-endstop but instead go to a row of screw terminals. This allows for easy test points--very handy--and also allows me to dismount the entire electronics section easily if I want to test just it or if I want to upgrade it later.
Another key addition are the fans above the electronics. These were salvaged off of a very nice Antec power supply that went belly up. They keep everything VERY cool on the boards and I don't really have much problem with their air currents cooling off the print in progress.
Another change--even though you can't see it--is the way the wires on and around the frame were routed. I simply cleaned things up, used lots of zip-ties and ran things in straight lines. Makes for a cleaner finish. The small loops of wire sticking up just before the screw terminals were to facilitate easy mounting since having too much tension in those wires makes it difficult to plug them in.
The next thing I did to Frank for his face lift wasn't so much astetically driven--not at all actually--but functional. There was simply way to much slip, wobble, and play in the extruder. So I clamped it down...seriously.
This turned out to be a key issue. After this, it wasn't a matter of how well Frank would extrude, it was getting him to extrude slow enough! However, after this was done and I had fired frank up to print a couple of things:
That first image was an attempt at a mini mug with .35 mm tip but I didn't have the height above the table set right at first so it detached. The second was another go at a heart for my girlfriend. The problem started on the heart.
You'll notice the singed ends and the discontinuous lines...my pump started loosing pressure during the build. This has happened before but not for this reason...I'd blown a hole in my PTFE wall. Take a look at where and what was oozing out:
That discolored spot on the PTFE about 1/3 up from the bottom is the hole. The discoloration is the hardened ABS in the hole. The little ram's horns to the right are what was oozing out of that and around the pipe clips holding the thermal barrel in. I bought the PTFE piece from the RRRF so I'm pretty sure it wasn't me messing it up. Apparently the wall was thin enough right there that it managed to tunnel through and leak out. Anyway, this caused all sorts of havoc with the heater barrel and eventually managed to dislodge it from the PTFE--I left it running because I saw no fix for it.
So, that's where FrankNuvo sits right now. I'm ordering a PTFE and heater Barrel from the fellow making the machined extruder parts--though not this week as I had thought, sorry guy--and will get FrankNuvo up and running again, again!
Monday, December 15, 2008
1800 Mechanical Movements, Devices and Appliances
This book covers a far more diverse set of mechanical devices than 507 - from gears and linear motion mechanisms to sails, steam engines, concrete mixers and mining devices.
This is a 302.9Mb DjVu file scanned at 400 dpi. I am working to get seeders up as quickly as possible for the torrent, as well as regular http hosting for any of us that are bittorrent-challenged.
Here it is: 1800 Mechanical Movements, Devices and Appliances
Edit: A regular http download is available now, courtesy of Includipedia here: 1800 Mechanical Movements, Devices and Appliances (wiki page)
Sunday, December 14, 2008
Tommelise 2.0 operational again
In which your narrator finds what may be a very clever way of attaching toolheads to the z-axis worktable... do you want to read more?
Saturday, December 13, 2008
Upgrading T2's z-axis
Tommelise 2.0 gains a more rigid z-axis... do you want to read more?
Tuesday, December 02, 2008
To introduce myself, I'm Jon, I'm English, I live in Germany, I'm a Java programmer by day and after following RepRap for a while, I've now been building myself a McWire in my spare time. I now have things up to the point where the machine is moving in three axes and (somewhat fitfully) extruding material. The electronics are generation 2 from RRRF, the extruder is the laser-cut one from Bits from Bytes.
A summary of some things I've encountered and the solutions I had for them:
- I was unable to find the normal steel/iron pipes that the original McWire build uses in local DIY stores. I suspect I was just being blind, since I've since now found them in the nearest two stores. Either way, I ended up using pipes and fittings made by Tubeclamps, for example RS part number 286-3459. (I ordered pretty much everything from RS. If anyone wants a complete list of the RS part numbers I ordered, mail jon at siliconcircus dot com.) These are big, thick, heavy pipes and fittings. They also have a 64mm hole diameter on the flange (as opposed to 70mm). Zach at RRRF kindly helped out here by laser-cutting the acrylic with a 64mm diameter. The pipes are a real pain to cut (you'll want a good pipe cutter and a spare blade) but other than that, they work fine and you end up with a very stable (and very heavy...) machine.
- Possibly due to the saddle bands I was using (RS part 283-2360), I couldn't get the captive nut for the Z stage to stay in place properly. My solution to this: solder the nut into the band using a whole lot of solder. While I realise that this is hardly the proper engineering solution, it works nicely.
- I managed to lose the extruder tip (the aluminium part with the 0.5mm hole) from the Bits from Bytes extruder kit. BfB were very helpful and sent me a new one at a reasonable price. Nonetheless, it cost me about a week. Don't do this :-)
- Once I got the extruder built, I found that the two cogs (the one on the motor and the one driving the M8 bolt) were slipping past each other. It's something of a makeshift solution, but I wrapped clear tape (Tesa Film/Sellotape to name two brands) around the motor cog. This provides sufficient pressure that they're now working nicely enough.
- Having done this, I got the extruder driving plastic forward and extruding. Unfortunately, I also got a leak at the barrel/PTFE insulator join. I'm using ABS. My first attempt at a solution was to wrap PTFE tape around the join and tighten a hose clamp over that. This seemed to work well for a couple of days, but I've just turned the heater on this evening and noticed a small leak again.
- Overall, the extruded filament didn't seem to want to stick to anything and the extruder didn't seem able to drive it sufficiently well. My conclusion was that the extruder tip (which isn't directly heated in the BfB design) wasn't getting hot enough for my ABS. I've now bought myself a Dremel, a 0.5mm drill bit and a brass M6 acorn nut. I drilled the hole and mounted this yesterday evening and results now seem better, though still not perfect. Certainly, when heating up to 235C (as registered by the thermistor in the barrel), I now see a little bit of extrudate leaking out of the hole, which definitely wasn't happening before. I've now applied some glass wool around the extruder tip and will see if that helps.
- It seems like the actual drive mechanism is having trouble driving the ABS. I've seen mentions elsewhere about sharpening the thread. If this is still a problem with the glass wool (and the consequently hotter and therefore presumably less resistive tip), then I plan to try this.