First Microchip PIC16F628 programmed...

This ought to keep everyone laughing at my expense for a week. Have fun and welcome to it! :-D

I built the JDM PIC chip programming board more than a month ago, but just got around to trying to program a 16F628 chip on it this weekend. This is an important thing to be able to do because the 16F628 is the core of all the controller boards for the Godzilla prototyping machine. Simon, Vik and Adrian have been moving towards a final design for all the boards for the past month and it is looking like I'll be able to start building them in a week or two.

Anyway, the first bit of drama was finding the proper software to drive the board and program the 16F628's. As usual it was right in front of my nose and Simon kindly showed me the link. He then told me the settings to use to get it to work.

I plugged the board into my PC, fired up IC-Prog and had a go... nothing. For the next several hours I played chimpanzee on a keyboard trying things to see if I could get it rolling... nothing.

Simon began to suspect that my ancient skills with soldering iron and PC board had maybe slipped a bit. Given that it had been 25 years since I'd last personally built one it sounded reasonable but annoying all the same. I considered just buying a premade, cheap programmer board.

Simon finally ran me through the troubleshooting protocol on the wiki site. The voltage numbers that I got made absolutely no sense whatsoever. I was going blind with fatigue when I finally crawled under my desk and had another look at the back of my PC. 25 pin cannon female connector for Com1, 9 pin cannon male connector for Com2... I'd rigged mine to use the 25 pin connector, just like the old days.

Wait a minute. Where was the printer connector? In the old days we used a Centronics connector, but those had fallen by the wayside some years back and NOW we used a... 25 pin cannon connector... female. It was still confusing because when I checked mode I had 2 serial ports. If that 25 pin cannon female connector was a printer cable where the hell was the other serial coms port? I've been using USB ports for peripherals for some years now so this was all remembering how to read ancient Egyptian for me.

I finally decided that I didn't know where the other coms port was but that the one I had had to be a 9 pin male cannon connector. This morning I went down to Radio Shack and bought a 9 pin female connector. I hooked it up ran IC-prog with the settings Simon talked about yesterday.

It wrote to the 16F628 and verified perfectly the first time out. I read it back into buffer 2 and compared buffer 1 and buffer 2 and it gave me a thumbs up.

Anyway, have a laugh. I deserve it. :-p

Making Version 2

Brett has kindly agreed to knock out a few kits for version 2 of the filament extruder and made some very valuable contributions towards its design. He has access to metal scrap the quality of which I can only dream about.

Here are the first two extruder barrels that he's fabricated. They're made of a good quality stainless steel so we should be in better shape with the colouring of the filament if my guesses about how they're getting coloured is right.

First off, version 2 will have a 1" diameter extruder barrel. The reason for this was my serendepitous discovery of a high quality, extremely cheap barrel heater used by the plastics industry in conventional extruders.

Its a sleeve heater designed to fit over cylindrical bits of plastic extruders to keep molten polymer inside molten. This bad boy prices in at about US$6-7 for a 300 watt heater. It can be bought for either 110 v or 230 v AC mains power. The cartridge heater it replaces costs closer to US$10-15.

The only drawback to this kind of heater is that it makes the use of a bimetallic thermostat as we were able to do with version 1 impossible for the simple reason that there is no place on an extruder barrel covered by this heater to mount one. That is not a big problem in that the controller for the Mk II extruder that is being used on the RepRap itself already makes use of a thermistor temperature sensor for control. I am sending along a kit of pieces for version 2 to Simon who has kindly agreed to morph the Mk II extruder controller board design into one that can handle version 2.

The polymer pump will be made, according to Brett, from 1.25"x1.25" stainless steel bar stock.

Overall, version 2 will weigh about half of what version 1 did. This is especially important in the extruder barrel in that it will have a much smaller thermal intertia than the version 1 did. The sleeve heater should also make for a much more evenly distributed thermal input to the polymer.

I am expecting great things from version 2. :-)

Results of a complete breadown of the quarter-inch extruder...

After having removed the auger and cleaned it I left the extruder barrel heater on at 120 Celsius to cook out the rest of the polymer out of the barrel. Afterwards, I disassembled the system completely.

The first thing I noticed was that the half-inch PTFE thermal barrier between the polymer pump barrel and the extruder barrel showed signs of swelling. Whether this swelling is an artifact of the heat that has been applied to the extruder barrel or the pressure that has been developed in the upper reaches of the extruder pump, or both is not known.

What was more interesting was the observation that polymer had not only coated the hot end of the auger but also formed a nice tube between the auger and the inside the PTFE thermal barrier.

Not only had it done that, but it had also migrated another 24 mm up the polymer pump barrel even though molten polymer had not previously been observed on the auger that far back from the extruder barrel.

Given that the system had not been broken completely down for cleaning since it began to be operated it is not clear whether this polymer tube was formed all from the get-go or later on.

Two interesting observations came from looking at this CAPA tube. First, the swelling of the PTFE barrier comes from the inside and is expressed in the tube wall thickness. Second, and even more interestingly, the tube is a clear, clean white all the way down to the end of where it exits from the PTFE barrier into the extruder barrel. Extruded CAPA, on the other hand, has been a light silver-grey. Further, there are no flakes in the tube whereas you will occasionally see a flake of dark matter in the filament.

This argues that the CAPA goes from white to silver-grey in the extruder barrel.

I'm drawing two initial conclusions.

• we are developing pressures in the extruder barrel sufficient to cause a bulging of the PTFE thermal barrier
  
• we are seeing either a purely thermal transformation of the CAPA colour or a thermally driven interaction between the CAPA and the mild steel walls of the extruder barrel that causes the colour change

More on polymer fouling of the auger...

I removed the auger from the extruder again prior to setting up for a new extrusion test, this time with a heated quenching bath. Polymer melt was again found to have formed as far back as 50 mm from the end of the auger. This means that the meltdown is confined to the PTFE thermal barrier zone and does not extend into the zone of the steel polymer pump barrel.Just as a side note, the German masonry auger is double flighted and is supplied by Artu USA.

It is made of chrome vanadium steel and appears to be titanium coated if the colour is any guide.

More post extrusion filament treatment...

Solvay lists the melting point of CAPA at 58-60 Celsius. I created a water bath of that temperature on my stove and dumped in the filament produced with yesterday's experiment.

The filament instantly became soft and pliable. Straightening it was no problem at all, though it did break at one of the weak points where the extrusion rate had slowed down when it was being extruded.

One error in yesterday's observations was apparently in reporting that CAPA floats. Today it sank immediately, which is in keeping with it's listed density. I am wondering now whether the "floating" behaviour that it exhibited yesterday had to do with its almost neutral buoyancy coupled with the fact that it was still attached to the extruder tip.

In terms of post extrusion "polishing" of the filament, a heated quenching bath seems to be a good way to go for now.

Post extrusion treatment of filament...

Reviewing the on-line literature about producing monofilament it would appear that for CAPA and polymorph at least we could use a heated quenching bath that would keep the filament very near the melting point of the polymer. If we regulated the temperature of the bath with enough care we might well be able to straighten and stretch the polymer to a proper diameter in one go. That would certainly reduce the complexity of the system.

Closing on producing usuable filament...

I realised late this afternoon that I need not build a framework as such to hold the quarter-inch extruder vertical. Ten minutes with a few pieces of scrap lumber and C-clamps gave me a solid platform for operating the extruder vertically. A water bath was created with a fruit juice jug and a piece of construction paper formed a makeshift polymer bin.

The water bath was situated about 2-3 mm below the extruder tip. I heated the extruder barrel to 120 Celsius and started the system.

This was the result.

The filament emerging from the extruder was instantly quenched in the water bath. The diameter of the resulting filament varied between 3.2 and 3.9 mm. The kinks that you see occurred when I had to manually poke the polymer feed bin with a thin bamboo stick to keep it filled.

Interestingly, the filament floated. This might indicate that we are entraining air in the filament since the density of the CAPA is marginally higher than that of water. No obvious bubbles were visible, however.

The practicality of Brett's advice that we need to think about designing a polishing station to straighten and regularise the dimensions of the raw filament now becomes obvious. That should be a perfect RepRapable product.

The extruder produced about 125 mm of filament in approximately 90 seconds. For these settings that is an output level of approximately 50 cm^3/hr. I had designed the system to produce about 15 times what a Mk II could consume per unit of time. We're producing about 18 times which is very close to our design goal. I currently have no way of knowing what the duty cycle of the bimetallic thermostat was, but given the setting chosen even if the cartridge heater had been on full-time it would have had a power demand of no more than 150 watts.

What we have is a prototype that proves the concept for a small polymer extruder. We need to take the concept through several generations now to achieve a useful system. I see several developments as being needed to make this a start-and-forget system.

• an improved polymer feed hopper that makes sure that the auger receives a steady supply of resin.
• roll godets
• heating oven
• takeup reel

It would be nice if this system could be self-threading. We obviously need to design a hands-off control strategy now that we have a way of making filament. Given the work that has already been done by the team with controlling steppers and the Mk II I have no doubt that we have the talent already on board to do this.

Here is the system that we are replicating.

It costs US$85,000 used. We're going to do it for under $150 tops as best as I can see.

A note on polymer fouling of the auger...

I disassembled the polymer extruder for the first time since it was assembled and tests began. As usual the auger, whose extreme end resides in the heated extruder barrel was frozen. Turning on the heater for the extruder barrel I was able to get the auger to rotate freely after a few minutes. Extracting it from the device proved, however, to be surprisingly hard.

When I recovered the auger I discovered that molten polymer had migrated through between two and three flights of the auger back towards the polymer pump of the mechanism. This melting was restricted to the PTFE thermal barrier part of the mechanism. Cleaning the auger was a trivial exercise. No polymer remained in the PTFE thermal barrier barrel.

This mechanism may explain why the extrusion of filament has been somewhat irregular during the tests carried out on the extruder. During the tests I've been rather cavalier in leaving the extruder barrel heated with the polymer pump turned off. I suspect that this is when the molten polymer migration occurred.

It would appear to me that extracting and cleaning the auger before making filament would be considered a good practice rule for operating this device.

Wednesday morning tests...

07:45 PDT AKA 16:45 ZULU TIME 8 March...

This morning I did another test run of the extruder, this time with the barrel temperature set to 110. I taped down a lot of things and installed a funnel so that the feed of polymer into the pump would require less attention.

Before I powered up I noticed that the auger bit was frozen from having its tip embedded in solidified polymer in the extruder barrel. I turned on the cartridge heater as usual but this time with the extruder barrel full instead of empty. After a few minutes thermal expansion of the melting polymer forced a bit of filament out of the extruder tip before the electric screwdriver was turned on. I tapped the on-button for the screwdriver and discovered that it rotated freely now.

At 110 Celsius the extrusion rate was considerably slower, but the filament tended to hold its shape better. The axial thrust on the auger bit seemed to be considerably stronger, too. The emerging filament still tends to foul on the extruder tip and thread as the length of the filament increases. Beveling the extruder tip like us done on the MK II design would be a great idea. I'll pursue that this weekend when I can get out the the drill press at my sister's house.

I took note of Adrian's experience with the filament swelling after leaving the extruder tip. The tip diameter to filament diameter proportion implied in Adrian's report was about 0.79. I worked backwards from this and determined that drilling a 3/32nds diameter hole in the extruder would result in a 3 mm filament. In practice that works out fairly close from the measurements I took this morning.

It seems obvious at this point that if we are going to extrude CAPA in this temperature range we are going to have to cool and support the filament almost immediately after it leaves the extruder tip. Otherwise it is going to draw into a thin thread because of the weight of the cooled filament end acting on the emerging polymer which, because it is molten, has a negligible tensile strength. Perhaps I can fabricate a little PTFE trough to provide that support.

08:45 PDT AKA 16:45 ZULU TIME 8 March...

It occurred to me that I had a little PTFE trough already made in the reject misdrilled bits of PTFE cylinder that I had made. I sawed one of those in half and taped it into place. You can see the result. (hmmm... pics don't seem to be uploading right now). The filament extruded into the trough all right but promptly adhered to the PTFE surface IN the trough. It seems that drilled PTFE surfaces are quite rough, a circumstance which probably explains how we can pump polymer through the PTFE thermal barrier in spite of the fact that the friction coefficient of PTFE is so low.

I also tried dripping water onto the filament as it emerged from the extruder tip. That worked to an extent but I had to be careful not to drip it on the extruder tip which chilled it and caused flow problems.

We appear to have another successful filament extruder concept...

20:45 PDT AKA 04:45 ZULU TIME 8 March...

Not to put too fine a point on it, the upgraded Mk II concept filament extruder... extrudes filament.

I decided to do the first run hot so that I could avoid problems with jamming. I set the thermostat on the heated barrel to 150 Celsius and let the system heat up for fifteen minutes keeping track of the temperature. When it hit 150 I began priming the polymer pump. Within 5 minutes molten polymer began to be extruded. I caught the output on a ceramic plate. The filament was too hot, as I expected and although it emerged at with a diameter of approximately 3 mm gravity quickly stretched it thin.

The product landing on the plate had no coherence in diameter.

Because I was working alone and was standing in for a thrust bearing for the test rig and spooning polymer into the feed chamber as the experiment was underway I was unable to photograph the extrusion as it took place. Here is a photo of the extrusion tip after I shut the system down. You can see a thread of filament remaining on the cool tip.

The first few feet of filament was dirty with grit and grease from the machining of the barrels. That soon cleared, however, and the last filament extruded was clear. CAPA turns translucent when it cools. It feels rather rubbery and is quite strong even in thin cross sections.

In the next days I will endeavour to configure the rig so that I can operate it with hands off. I will also have to figure out a way to connect the electric screwdriver to the extruder with something a little less informal than duct tape. :-)

Interestingly, the electric screwdriver provided more than enough torque to operate the system. There was no jamming of the system for any reason. As well, I was able to manually resist the axial thrust against the screw pump. Apparently, the larger diameter extrusion orfice means that the operating pressure in the heated extruder barrel is considerably less than I expected. That is wonderful! :-D

I think as well that it would be good to slightly incline the extruder so that the filament doesn't try to hang onto the extruder tip. Having a water tray directly under it would also be a good idea. Godet stations and a takeup reel system are something that somebody might want to think about reprapping before too long.

I will also do a series of test runs at lower temperatures to find a more optimal operating regimen. I'd like to find a less informal gear motor to integrate into the system. I don't get the impression that the plastic gear motor that we use on the Mk II is going to be up to the job. I think that we will need just a little more torque than that motor can deliver. On the other hand, I get the impression that this system will produce filament at quite a good clip.

It would appear that the only reason for going to a larger diameter auger would be to be able to handle polymer in larger granules.

Using the masonry bit appears to have made all the difference at the last.

I would like to thank everybody on the team for their advice and observations. Thanks goes to Dr. Bowyer who also ran with the concept using several different assumptions and whose example prodded me to work a lot harder than I would have on my own.

Finally, especial thanks go to Vik and Brett, whose timely last minute suggestions and encouragement yesterday got me past the jamming problems that I encountered. I was very discouraged at that point. I don't think I would have attempted to run the auger through the thermal barrier and into the extruder barrel had Brett and Vik not suggested it.

Thanks guys. This has been and continues to be a LOT of fun.

Tests on the 6.35 mm filament extruder

After reviewing Vik and Brett's comments, I decided to see if I could pump polymer through a PTFE barrel with an auger bit. It turns out that contrary to what I'd read, viz, that the wall's friction coefficient had to be higher than the auger's isn't so. CAPA pumped through a drilled length of PTFE with no trouble at all. That means that the whole pump and thermal break assembly could be made out of a single piece of PTFE.

There are two practical problems with doing that, though. First, PTFE is expensive at US$0.15/cm^3 (US$2.50/in^3). That wouldn't be so bad if I weren't worried that you'd be replacing it regularly because of wear between the auger and the barrel walls.

Vik makes another, more intriguing suggestion though. What if we applied a water jacket to the pumping section. My mind immediately went back to the old water cooled Browning heavy machine guns my grandfather used in WWI. I seem to remember that they worked by allowing the water to boil off. That would create a liming problem, but those are relatively easy to sort out.

Hmmm.... Adrian! How is that prototype of yours going?

20:22 PDT AKA 04:22 ZULU TIME 7 March...

Hit the hardware store again and had to choose between two German bits. One was half again longer than I needed and was made of some mad alloy of titanium and vanadium. I figured that I'd never be able to cut the carbide tip off of it, never mind mill the shank end down a bit to accomodate a thrust bushing.

I settled for a more modest German masonry bit that will require that I shorten the thermal barrier just a touch so that the tip end of it will extend into the melt chamber. I sawed off the end of that one and tested it with the pump barrel housing and thermal barrier and it happily pumps CAPA and even pumps it in the PTFE barrel. I suspect that it would have pushed the polymer completely out of the extra 25 mm of PTFE barrel beyond the end of the bit and out of the existing PTFE thermal barrier except that I was having to resist the thrust of the drill bit manually.

While I was at the hardware stockist I got the bits of power cord and taps necessary to seat the thermostat on the extruder barrel and get the melt going. I also bought a vise so that I can try drilling extruder tips here rather than having to trek out to my sister and brother-in-laws every time I need to put something in a vise.

22:35 PDT AKA 06:35 ZULU TIME 7 March...

Successfully pumped polymer through the polymer pump, through the PTFE connecting barrel and into the heated extruder chamber with the new, longer masonry bit. Set up to drill a better aligned PTFE connecting barrel in situ and marked it. Drilling awaits a more godly hour.

0623 PDT AKA 14:23 ZULU TIME 7 March...

Managed to get a true hole through a PTFE cylinder on the second try. Double checking the polymer pump test when the batteries on the electric screwdriver ran down. Time to make some breakfast. :-P

07:05 PDT AKA 15:05 ZULU TIME 7 March...

Repeated polymer pump test with recharged batteries. Interestingly, this time the pump filled the entire heated extruder barrel, 89 mm long and .89 cm^3 volume, before jamming. The additional flights provided by the masonry bit apparently makes for a more efficient pump for the amount of torque that we have. Will be drilling an extruder tip and wiring the heater barrel after breakfast. Hope to run a full test this morning.

09:43 PDT AKA 17:43 ZULU TIME 7 March...

Thermostat and power cables rigged correctly. No fires, electrocutions or explosions so far. Have my fire extinguisher on hand just in case. :-o

Presently running tests to get a rough idea where the thermostat setpoints are. This takes time because of the thermal mass of the heater block. Taking temperature readings in the extruder mouths with the IR non-contact thermometer. The mass of the steel will predominate with this system so I don't have to run it loaded with polymer since the heater barrel weighs a couple of pounds and a full charge of polymer in the barrel less than a gram. :-)

Temperature in the barrel is at about 60 Celsius just now and the thermostat setting is at 12:00.

10:16 PDT AKA 18:16 ZULU TIME 7 March...

Thermostat setting of 12:00 stabilises at 127 Celsius. Shifted thermostat to 09:00.

10:30 PDT AKA 18:30 ZULU TIME 7 March...

Thermostat setting of 09:00 stabilises at 94 Celsius. Shifted thermostat to 11:00. The indicator knob on that bad boy gets HOT!

10:58 PDT AKA 18:58 ZULU TIME 7 March...

Thermostat setting of 11:00 stabilises at 116 Celsius. It looks like the thermostat is pretty linear, which is what I'd hoped for. Now, I wonder what the deadband looks like.

Adrian got a bit of extrusion going at 120 Celsius, but Solvay took their melt flow index at 160, so I'd better check to see what higher settings on the thermostat gets me. The data sheets that Jeff's people in the UK so kindly sent along with the powder samples indicates that CAPA decomposes at 200 Celsius. Fortunately, there are no halogens or nitrogen in the formula for CAPA so if I screw up and burn some it's unlikely to kill me immediately.

Oh well, now to get some food and drill a extruder tip. :-)

13:30 PDT AKA 21:30 ZULU TIME 7 March...

Thermostat setting of 15:00 stabilises at 132 Celsius. I'm going to crank the thermostat up to max, about 17:00 and see what happens.

By the way, the deadband is about 5 degrees. Not bad. :-)

13:50 PDT AKA 21:50 ZULU TIME 7 March...

Thermostat setting of 17:00 (max setting) stabilises at 160 Celsius. How convenient! :-D

Looks like I'm going to have to develop a graph for the thermostat performance. That certainly isn't linear. I also need to go back and check 15:00 again.

14:15 PDT AKA 22:15 ZULU TIME 7 March...

Thermostat setting of 15:00 (max setting) stabilises at 137 Celsius. Looks like last time I caught it at the bottom of the dead band. Ok, fair enough. :-)

So far so good! :-)

Monday tests on the 6.35 mm filament extruder

I decided to run a few preliminary tests on the 6.35 mm filament extruder before I did further work on it.

I used semolina as a proxy for polymer resin to check to see if the assembly would pump. There was no problem with that. I then inserted the PTFE thermal barrier into the assembly to see if I could pump semolina through that.

The semolina immediately jammed in the thermal barrier passage. It formed a 15 mm plug almost exactly like that I encountered when using the same auger bit in 1/4 inch steel pipe except the steel pipe only made a 5 mm plug before jamming.

The plug was easy enough to clear once I detached the thermal barrier from the polymer pump assembly.

Think that the problem might have been specific to the semolina I then used a few grams of the CAPA resin in powder form that I recently acquired as samples. The same thing happened.

Any ideas?

Building the 6.35 mm filament extruder

The weather held this weekend so I was able to almost finish the 6.35 mm filament extruder prototype. I am using Adrian's size convention in setting the calipers open to 20 mm for scale.

You can see that it bears a close resemblance to what my design study conception looked like.

I oversized the PTFE sleeve in the thermal barrier so the design sags a bit right now. Following Brett's suggestion I will put a third support plate between the polymer pump and the PTFE thermal barrier to counteract any tendency towards buckling and more importantly, to provide a bit more support to the heavy steel extruder barrel to at the right hand side. That plate will support the retaining rods via PTFE sleeves so that heat moving down the retaining rods won't be transferred into the front of the polymer pump.

I had originally thought to use a 1 x 1 x 2 inch PTFE thermal barrier with two PTFE cylindrical plugs to connect it to the polymer pump at the left and the heated extruder barrel at the right. While I was building the extruder I realised that that would make four potential gaps in the polymer flow path between the pump and the extruder tip. I brought that down to two by drilling out a 1/2 inch cylinder of PTFE and then drilling a 1/2 inch hole in the 1 x 1 x 2 inch PTFE bar so that it acted as a sleeve for the cylinder. That worked quite nicely.

Here is a detail of the polymer pump. I got too tired to finish the thrust bushing for the auger bit so the auger is seen lying in the foreground of the pump. I am using an ordinary plastic funnel for a polymer feed hopper.

Finally, here is a detail of the extruder block and a mockup (3/8 -24 1/2 inch bolt) of the extruder tip. Again, I was too tired at the end of the day to attempt to drill out an extruder tip. I've done trial runs on both the extruder tip and the thrust bushing, both made from 3/8 inch bolts and found that to be very finicky work. I've got a handful of the bolts so I can afford a lot of errors. :-)

You can see the wires from the cartridge heater in the photo and also the thermostat sitting atop the heated extrusion barrel.

I have to tap in the thermostat and lock the extruder down on a block. I also have to mate the shank of the auger bit with the electric screwdriver you've seen in the pictures. That all will take another morning, I expect.

It's going to be another two-tylenol night. Right now I am for some supper, a shot of MacAllan 12 year old, a movie, a hot bubble bath and a good sleep. It's supposed to begin to storm in about an hour. We've got another pineapple express storm that's come up from Hawai'i this evening.

Design study for the 6.35 mm filament extruder

I put in a few hours using AoI to see if I could make all the pieces for the quarter inch filament extruder work together. I'd like you to walk through the design with me and comment if you can see problems with it that I've missed. I put in a few hours using AoI to see if I could make all the pieces for the quarter inch filament extruder work together. I'd like you to walk through the design with me and comment if you can see problems with it that I've missed. First, I'm using a quarter-inch diameter wood auger bit that is just over seven inches long. It has eight flights in the screw for a length of four inches, two inches of bushing behind that and a shank that extends for an inch and a quarter beyond that. I didn't take time to develop the screw portion of the bit on the left on AoI and placed a thrust bearing where the bushing meets the shank on the right. The thrust bearing is made of a drilled out 3/8ths or 1/2 inch bolt.

From my previous experiments I determined that the auger should extend slightly beyond the end of the pumping section to preclude polymer packing in the pump barrel. You can see the end of the auger extend just beyond the end of the connector plug that I've made from another one of those drilled bolts from which I've sawn off the head.

Now I include the body of the polymer pump. You can also see the polymer feed into the top of the pump and the feed chamber where it intersects with the auger. Next we screw the PTFE thermal barrier onto the connector plug. I've shown the PTFE bar, which has been drilled to a quarter inch to accomodate ploymer flow and tapped at each end to accomodate the connector plugs, as slightly smaller in cross section than it is so that you can see where the steel polymer pump section and the PTFE thermal barrier join.

We can now screw the heated extruder barrel onto the second connector plug seated in the end of the PTFE thermal barrier. You can see that the heated barrel is drilled not only to allow passage of polymer as it is heated and is pumped towards the extruder tip but also to accomodate a 300 watt cartridge heater.

Now let us insert the cartridge heater. There is a quite inexpensive anti-lock compound that will prevent the cartridge heater from freezing up in its housing after a number of heating cycles. You will have to imagine the wires for the cartridge heater extending out the back of its housing towards the rear of the extruder. These cartridge heaters use regular lines electricity.You can also quite easily acquire bimetallic thermostats which can be used to control the temperature of the system. Like the cartridge heaters, these thermostats are very old technology and are also quite inexpensive. From this point we can apply the front and back retainer plates.

These are held in place by screwing the extruder tip into the heated extruder barrel on the front side and screwing the thrust bearing into the pump barrel on the back side of the assembly.

From there you secure the whole device with retaining rods made from threaded studding which are secured with lock nuts on both sides of the extruder.

This arrangement hopefully will preclude the extruder from coming apart from the internal pressure of some 40 atmospheres created by the pumping action of the polymer pump. I've calculated that this pressure will create an axial thrust of about thirty pounds.

Once that is done you simply add the feed funnel for directing polymer powder resin into the polymer pump feed chamber and you should be good to go.

I am planning on taking the newly conservative approach that Adrian has taken of using an electric screwdriver to drive the system. I am guessing that since Adrian has already demonstrated that an electric screw driver will drive his 13.5 mm extruder it should be more than enough for my quarter inch device.

I was going to use the gear motor for the Mk II, but decided that there was too much drama in accomodating such a small piece of equipment into what is a rather large (about 1 foot long and weighing 5-10 lbs) piece of equipment. Using an electric screwdriver will, I think, make the question of securing the system to a mounting block considerably easier.

I should also be able to run it without having to resort to complicated drive and control schemes.

As for operating the system, I already own an infrared thermometer which will allow me to measure the surface temperature of the heated extruder barrel rather well. Initially, I plan to let the system heat up slowly whilst empty by adjusting the thermostat (which I will tap mount into the chin of the extruder barrel, not shown) until I get the temperature I want. At that point I will start the electric screwdriver and start feeding polymer into the system. The thermal capacity of the system is very much larger than the that of the polymer flow, so I doubt that the system will much notice the polymer being extruded from it energetically.

This design largely emerged as an effort to salvage some of the materials bought for the old Gingery extruder but not used. It was also designed with an eye towards being easy to break down and clean in case of jams and allows for the performance of different polymers and also different diameter extruder tips to be surveyed. The cartridge heater is also magnitudes more robust than the hair-think nichrome wire that we are currently using in the Mk II. Adrian noted that dipping his extruder head in hot water to get polymer out wouldn't be good for his device. I can demount the extruder barrel by loosening the retainer rods, slide the heater cartridge heater out, remove the retaining plate and extruder tip and immerse the extruder barrel into boiling water with no danger whatsoever. :-)

Now... objections... observations?