• A 20x20x15cm build area
• A heating print bed
• A heated build chamber (ambient to 100C) to possibly eliminate warping.
• A respectable print resolution and speed

Over 12 months ago I made my first serious attempt at satisfying these by designing and half building a rep-strap based on the mantis CNC design. However serious limitations in the design and a lack of free time have resulted in the ‘gunstrap’ collecting dust for over 12 months.  Among others, the problems with the gunstrap were:

• High rolling resistance due to metal on metal sleeve baring
• Slow speed due to needing to physically move the heavy print bed and print head. assembly.
• Poor resolution due to the design of the Z and Y axis.
• Extruder stepper located within the build chamber.

To overcome each of these limitations I have spent some time designing a replacement in SketcUp, as seen below.

What you see above is a fully enclosed build chamber that will be constructed from 12mm wood fibre board or similar. The blue transparent section is a double layer glass viewing window that is opened by the handle below it.

When opened, the two axis print head and print bed are accessible. To the right of the window will be a 16 character 2 line LCD display for temperature readouts and the like.

This design features the following:

• All electrical components and motors (ex end stops) are located outside of the heated build chamber.
• Print head weight has been reduced to as little as possible to increase print speed and resolution.
• Rolling resistance is lowered through the use of ball bearings.
• Scissor lift Z axis for increased stability

The scissor lift Z axis will be constructed by modifying a  lab jack similar to the one shown here. If the wooden shell, which also acts as the main structure of the printer, is removed then the workings become more clear. The modified lab scissor jack coupled to a stepper motor can be seen below (click image to enlarge).

Looking from the front top down on the two axis print head stage its seen that its composed of stainless steel shafts for guides like a Mendel, a PTFE sleeve bearing for the print head holder similar to a Ultimaker. Rather than an expensive belt I plan on sourcing some fine braided wire to use as a pulley which I have seen work quite well on older mechanical pen plotters for lab work. I plan on cutting box section aluminium from corner to comer to make L pieces to hold the guide bearings.

Feeding filament into a wades extruder on the side of the printer will be a mounted filament spool. The wades extruder will force the filament up a PTFE tube which enters the printer at a hole located at the top of the printer.

Finally, in a side compartment insulated by double thickness paralleling will be the electronics. This includes a RAMPS based stepper driver system, ATX powersupply and cooling fans.

As no low melting point plastic components or electrical equipment is contained within the build chamber I believe the high build chamber environment of 100C should be achievable. The heat will be provided by the heated print bed and print head only and will be actively circulated by a fan at the top of the chamber.

I am plananing on sourcing the bearings from smallparts.com.au and modifing them to include a flange. The stepper motors will come from robotgear.com.au for around $85 for 4,  including shipping within Australia. I have a month off before starting a PhD in 2012 so hopefully that allows enough time to get this all built and calibrated.

You can find a copy of the 3D model from here. Some parts of the model were sourced from Googles 3D warehouse including the Steep Reel, bolt, Arduino Mega, character display and ATX powersupply.

I would love to hear what people think of this design and so welcome all comments and criticisms. If you have any suggestions for improvements or alternative ideas please leave a comment!

Happy new year to all!

Hasn’t the DIY 3D printing scene been doing this for quite some time now with the use of spars internal structures for solid objects?

The printing, routing or embedding of wires to form a circuit board track is now common place and with pick and place of components well on its well this could become feasible sooner than thought.

There was already an an entry on the RepRap wiki regarding PCB motors but I cant find any evidence that they have been successfully used.

I’m told that its used to grind plastic granules down into a fine powder. What is particularly interesting though is that this will only work if liquid nitrogen is poured in while it is grinding or else it will quickly become jammed. It seems the liquid nitrogen cools to the plastic (−196 °C, 77 K , −321 °F) and so reduces its fracture toughness and allowing it to be broken into smaller pieces.

So is this a practical approach for those at home wanting to attempting to recycle their own plastic for use with a 3D printing? Obviously not all of us have access to liquid nitrogen or even dry ice, so would a home freezer provide much of an advantage?

I was unable to find any useful information on the deformation properties of HDPE, ABS or PLA at low temperatures in the limited time I have available and so its difficult to tell if a domestic freezer (-20°C?) would be cold enough to make a usable difference.

If anyone wants to put on their science hat and undertake a few home tensile tests on a short lengths of 3mm filament to produce a stress-strain curve at low temperatures and at room temperature it could provide interesting results.

A collection of the gears seen in this page can be downloaded from the 3d warehouse.

To begin with, download the involute gear plugin and copy it to your sketchup plugin directory. This plug-in was not produced by me and all credit goes to Cadalog Inc for writing this very useful tool. Then open sketchup and choose ‘Involute Gear’ from the draw menu. From the dialogue box you are presented with three options.

• Number of Teeth.

The number of teeth on your small gear (pinion gear)  relative to your large gear (gear wheel) will determine the gear ratio. If you desire a gear ratio of 2:1 then your gear wheel will have twice as many teeth as your pinion. However, your gear with more teeth would need need to be twice the size in order to have the same sized teeth.

By choosing an even number of teeth you have a range of different exact gear ratios. For example; 12 and 24 (1:2), 12 and 36 (1:3), 12 and 48 (1:4) ect. However, an even number of teeth will result in the same pair of teeth meeting over and over again. By having an odd number of teeth each tooth will meet a different counterpart on every rotation. This helps to distribute lubricant and reduce ware. For example, Wade’s geared extruder has 11 teeth on the pinion and 39 teeth on the gear wheel.

Try and select the highest number of teeth possible while still maintaining a printable resolution. Have no less than 12 teeth and try to avoid a gear ratio any higher than 1:6.

• Pressure Angle.

The pressure angle effects the geometry of the gear teeth. A low pressure angle (14.5) is normally used with high number of teeth (40+) and will give a greater contact area but lead to increased noise and backlash. Source. I would recomend sticking to a pressure angle of 20. What ever pressure angle you choose, make sure it is constant among all gears used together.

• Pitch Radius.

This is not the outermost radius of the gear (the root circle). The pitch radius is the distance from the centre of the gear to the ‘pitch point‘, the point of contact between the two gears.

As such, the outermost radius of the gear for a fixed pitch radius will change for different numbers of teeth. However, no matter the number of teeth, the ‘pitch point’ will remain constant for all gears.

The key point here is that if you want two gears to mesh well, then the pitch radius must be increase or decreased by the same ratio as your number of teeth. For example, lets say you have a small gear with 10 teeth and a pitch radius of 10mm. If you wanted a ratio of 1:3 then your large gear would need 30 teeth and pitch ratio of 30mm.

Once you are comfortable with these settings its quite simple to make a range of different gears. When you have finalised your gear design you can export the file as an STL.

Involute Gear

To create an involute gear, use the plugin as described above to first create the gear outline. The gear outline will appear at the point 0,0 so its good idea to mark the centre position before moving it. Add your centre shaft hole and then use the Push/Pull tool (hot key ‘P’)  to make the object 3d.

Helical Gear

Make sure the centre of your gear is still positioned at the point 0.0. To create a helical gear simply follow the same process as for the involute gear but with one extra step.

After you have ‘pulled’ your gear surface up to make it 3d, select only the top face of the gear and choose the rotate tool (hot key ‘Q’). With the top face still selected position the protractor over at the centre of the gear and right click. If you have an empty space for a shaft in the centre then just hold down shift while hovering the protractor over a horizontal area to lock it in that orientation. Then its just a matter of choosing your desired angle. For a single helical gear the angle used is normally 6,8,10,12,15,20 degrees. Be aware that a greater angle will result in a greater axial load (see below).

Why would you use a helical gear? They can reduce backlash, ware and noise as opposed to an involute gear. This is due to the gears meshing slowly over their length rather than all at once. However a single helical gear will result in a axial (in the direction of the shaft) load. This can be overcome by using a double helical (herringbone) gear as seen below.

Herringbone Gear (double Helical)

To make a herring bone gear follow the same processes as for the helical gear. Due to elimination of axial loading (see below) higher helical angles can be used. Eg 25,30,35,40 degrees. Select the top face of the helical gear and pull it up so that your gear is now double in height. Then use the rotate tool to rotate it back in the opposite direction as the lower half.

The advantage of the herringbone gear is that all axial forces are cancelled by the opposing helices. This allows for all the advantages of a helix gear with out any of the disadvantages. A herringbone gear will also self centre. Please note that if two herringbone gears are even slightly miss aligned and then fixed in place then they will damage each other. Its best to first fix one gear in place on its shaft while leaving the other free to move parallel to its own shaft. Then turn the fixed gear’s shaft and allow the second gear to ‘self align’  before finally fixing the second gear in place on its shaft.

More information about double helix (herringbone) gears can be found here and here.

Straight Bevel Gear

To produce a straight bevel make sure you add the shaft hole after making the bevel or else the shaft the wrong shape. Follow the same steps as for the involute gear then select the top face only. Choose the scale tool (hot key ‘S’) and while holding down Ctrl (to scale into the centre) select an outside corner and reduce the top face in size.  You will notice that as you do this you change the angle of the gear teeth.

The angle of the gear teeth is important. You can choose any angle so-long as the matching gear has a corresponding angle that together they make up 45 degrees. For example, if the teeth on one gear are at 10 degrees off the vertical then the second gear will need to have teeth that are 35 degrees off their original position.

By producing two bevel gears you can transfer mechanical power over 90 degrees.  Bevel gears can also be used to change gear ratios. The possible combinations of size and shape are endless so its best just to have a play around in sketchup to get used to it.

Internal and Planetary (Epicyclic) Gears

The involute gear plugin can also be used for creating internal gears and planetary gears. To create the external gear, first select the circle tool and starting from the origin drag it out to your desired size. This is your external gear.  Next, using the involute gear plugin produce a gear with the exact dimensions you desire for the inward facing teeth of your external gear. This will also appear at the origin. Using the pull tool, make the new gear you just created into a 3D object of any height.

The next step is the important one, select the gear ‘group’ (one click on the gear instead of two) and then right click and from the menu choose Intersect -> Intersect with model. There will be a short pause. Once done you can delete the gear group entirely and you will be left with your original circle with the outline of the gear ‘intersected’ into it. From this you can delete the centre face and pull up the outer face to give you your external gear. This can then be populated with other gears to make a planetary gear system or similar.

You can find a collecting of all these gears in one model here. If you have any suggestions, comments or corrections then by all means leave a comment.

One such tool is Google insight. Google insight is a free tool that’s used by company’s to asses the impact of their product marketing by observing how search interest changes with time and location. It can also give some interesting insight into how awareness of desktop 3D printing has changed over the past 12 months.

For example, the number of searches for the term ‘reprap‘ has continued its relentless growth over 2010 and shows no signs of abating.

The scale bar on the right (0 to 100) just represents the normalised values. In other words, the date with the most searches so far will have a value of 100 and all other points will be some value from 0 to 100 that’s relative to that peak value.

Focusing on only the last 12 months, it can be seen that 3D printing related search terms have continued to experience rapid growth on the preceding 12 months.

What I find interesting about this table is that ‘Reprap’ related search terms are growing much faster than Makerbot related serch terms. This is a little perplexing considering the amount of media coverage makerbot has had over the last year and the shear number of cupcake cnc’s that are out in the wild. Then again, its almost impossible to keep track of the number of Mendel’s in the wild and any media coverage surrounding them.

This is also seen when you search the term ‘reprap’ while excluding makerbot and serch the term ‘makerbot’ while excluding reprap. Keep in mind that this doesn’t reflect the total number of seaches for each term, but rather suggests that one is growing faster than the other.

Looking again at the search term ‘reprap’, this time by region shows another defining trend.

As you would expect, only wealthy first world countries are currently seeking out information about Reprap’s.  Its a little hard to see on this scale, but the Netherlands has shown the biggest increase. Conducting the same search again for the term ‘makerbot’ shows a slightly different picture.

If you look closely at Europe its clear there is less searches for the term ‘makerbot’ when compared to the term ‘reprap’ above. This may again be due to the mostly US based media coverage of Makerbot.

Something else to consider here is that all this information is gathered from google searches only. As google is far from the dominant web search engine in countries such as China (20% market share as opposed to Baidu’s 62%) , it hardly provides a clear picture. Also language barriers may play a part. There is a definite reprap presence in China, and even their own home grown Up! Personal 3D printer which isn’t reflected in the goggle results.

All up one thing is very clear from all this: Desktop 3D printing is growing at an astonishing rate.

—————————————————————————————————————————————–

Looking into the crystal ball I suspect it wont be long until well established fortune 500 company’s begin to take this emerging domestic 3D printer market seriously. Within a decade we may begin to see sub $1000USD desktop 3D printers from brands such as Epson, HP, Cannon, Brother or even Apple (The iprinter?).

No doubt each company will have a slightly different idea of how things should be done (filament size, plastic type ect) and chaos will ensure until a set of standards can be agreed upon.  Even if this means a Betamax vs VHS style war.. If your under 20, think Blu-ray vs HD DVD.

What worries me the most though isn’t the hardware standards, but the software standards. If each company tries to limit what you can print on ‘their” machine to only 3d models downloaded from their own market place then this will really hurt creativity. Worse still will be if they word cretin user agreements so that any model you upload instantly belongs to them or if they start to censor 3D models they don’t agree with. Imagine the horror if you discovered your brand new shop bought printer wouldn’t let you print a basic female form because it was deemed unsavoury.

Then again, isn’t new technology usually first utilised by the adult entertainment industry?.. Thankfully all these things,  for better or worse, are still years away.

For some interesting discussion on the future of 3d printing take a look at the pages  linked below.

This is an idea that has been in the back of my mind for quite some time.  Obviously you couldn’t use liquid water directly, but would need to freeze it to form ice. If this could be done successfully with a respectable resolution then you would find that you have a cheap, near unlimited and environmentally friendly feed stock. Not to mention the crazy coolness of being able to sip your next Martini out of a cocktail glass made of nothing but ice.

The RepRap IceRap?

Obviously, your newly printed creation wouldn’t last long at room temperature and so would limit the useful applications drastically. Although it could be an alternative for test test prints of half finished models with out wasting filament. Parts could also be stored in a freezer until needed.

After a quick search it appears this idea has been considered before, as seen on the Future Tool Ideas page over at the reprap wiki. To quote from it:

“Water ice extruder

For mass-produced ice sculpture, individualized ice lolly, etc.”

That’s as detailed as it gets and so isn’t much to go on.

So how exactly could you ‘print’ ice? This is one possible solution that I have been toying with:

• Step 1 – Create a sub zero (degrees Celsius) print environment.

Probably the most practical way of doing this would be to place the entire 3D printer into a walk in freezer or similar. Alternatively, an insulated build area such as Nopheads ‘wooden overcoat‘ or the Gunstrap’s insulated build chamber could be actively cooled. This could be achieved through the use of a conventional compressor based system scavenged from an old freezer, a thermoelectric cooler (Peltier) or through the use of dry ice/liquid nitrogen. The system used isn’t important provided the print area is well below freezing over the duration of the build.

• Step 2 – Use your 3D printer to deposit fine beads of liquid water.

The idea is that you deposit a very thin bead of water onto your build surface is the same way that would extrude plastic. Once a layer has been laid down there would be a delay of a few minutes while it freezes in the cold build chamber. Once frozen, the next layer is deposited and the process repeats. By keeping the beads of water only a few millimetres in size surface tension alone should be enough to keep them in place. I suspect that even limited overhangs could be achieved. Once printed, a quick once over with a hair dry or heat shrink gun would turn the object clear.

So how would you deposit very small beads or even individual droplets of liquid water in a controlled manner? Well thankfully this problem has, for the most part, already been solved with Adrian Bowyer’s Reprappable-inkjet print head. This was inspired by a previous design by Johnrpm’s Piezo Printhead.

Adrian's reprappable inkjet.

A video of the print in action can be found here. With a little tweaking it should be possible to further reduce the size of the output nozzle jet to increase the printable resolution. It would also be necessaries to insulate the inkjet head and feed pipe to stop it freezing as it operates in a cold environment. Similar precautions would also be needed for the electronics and stepper motors if directly exposed to the cold. Condensation would be the real killer hear.

• Step 3 – Get creative.

Add food dye for colourful creations! Include cordial in your water so you can print a tasty, lickable version of Andy’s face. Print your own spiral drink cooler that cools your drink as you pour it. Design an ice swan centrepiece for your next dinner party. The possibility’s are endless.

I have created a reprap Wiki page on the topic of 3d printing with ice which you can view or edit.

Thoughts?

Lids and bases of the ammunition boxes.

Its 16mm thick, easy to work with and best of all, free. Hence I have called this new repstrap the “Gunstrap”…

I have made a reprap wiki page with a few details of the build, including all the dimensions which I have been using to cut out the required parts.

An exploded view of the 'gunstrap'.

After a few hours with the jig saw I ended up with most of the required parts.

Plywood parts for the gunstrap.

You could use a hand saw in theory, but it would take considerably longer.

A few notes on the construction; by drilling through both sides that hold the stainless steel rods at once, you can ensure a perfect alignment. This, in combination with the gluing of parts onto the hollow sliders means even someone as care free with measurements as my self can end up with a smooth sliding repstrap. Both of these ideas I greatfuly borrowed from David Carr and his construction guide for the Mantis CNC.

Drilling through both parts at once ensures perfect alignment.

After another few hours of drilling, gluing and screwing together its about half done.

The Y axis (print bed) is mounted and ready for the belt to be attached. I had planed on imitating Nophead’s heated bed design which uses 240V AC. However I’m now having second thoughts due to safety concerns and so will instead look into the 12V computer power supply option.

One key lesson from all of this is that Sketchup is indeed very useful for designing new things. Its easy to learn, very easy to use and you have a large collection of parts to source designs from on the 3D warehouse.

Lastly, I’m still unsure if I should commit to the Peltier idea for the extruder stepper motor as mentioned in the previous post. A lot of people commented saying that 100 degrees C is too hot and that the commercial printers only run at around 70C. Even at this reduced temperature the stepper motor would still benefit from some form of cooling. However other options such as a Bowden extruder or ducted cooling (used by commercial printers!) for the extruder motor were suggested which could be possible options.

Before I move on to all that though I want to just start printing first. If my wades extruder arrives some time this week then then my first prints should be before the end of the year.

The design scheme for this new repstrap require that it:

• Has a print area of 200x200x100mm.
• Has a heated print bed.
• Has an precision of 0.1mm
• Has a manually quick swappable extruder/cutting tool heads.
• Does not use any bearings (extruder excluded)
• Is constructable with only basic tools.
• Only requires parts found in domestic printers and scanners.
• Is easy to access for cleaning and maintanance.
• Has a viewable window that can be opened.
• Has a entirely contained heated build chamber with a maximum temperature of 100 Degrees C to increase part quality.

Obviously it is the last dot point which will be the hardest to achieve. At 100C any thermoplastic components will be softened enough that creep will become a real problem. Also it will put real stress on the stepper motors, which are rated to only 40C.

In order to (hopefully) solve these problems the build will go against the reprap theme and not be constructed using low temp thermoplastic. Even the Wade’s extruder will be  a polyurethane resin cast, which is usable at 100c. To try and protect the stepper motors they will be contained outside of the build area and exposed to the ambient temperatures only. This is of course all stepper motors apart from the extruder stepper, which must be within the build chamber.

To protect the extruder stepper motor it will be insulated and then cooled with a peltier(wiki link). If you find the wikipedia entry too heavy going, then take a look at this less physics based description instead. On ebay there are currently Peltiers rated for 230C , however I cant find much detailed information on them. This is far from the most efficient design, but I think it should work maintain a safe temperature for the stepper motor. Though more research is required though before I commit to this idea.

This is the design so far (to scale) as compared to the current mantis inspired repstrap:

The old repstrap design on the left, the new one on the right.

The x, y and z stepper motors are located outside of the build area. Like the original repstrap, it will be constructed from plywood. The glass doors on the front will open by folding up on hinges until they lock into place on the top with felt covered magnets.

The new repstrap with the glass doors folded up.

If more access is needed to remove printed parts or swap out an extruder then the sides with handles will also fold down on hinges to open up the print area completely.

The new repstrap completely open.

However, even when closed up there will still be a 20mm wide gap where the Z extension bars move up through. My hope is that I can use bristles from a brush to stretch across the gap but still allow the rods to move freely from side to side.

Mind the gap..

The heated bed will be based on nophead’s 290W proven design using this aluminium clad resistor. My hope is that the combined heat from the extruder and the heated bed will be enough to heat the chamber alone.

Here is a close up of the extruder and z-axis. I havent drawn in the sliders in this image. The extruder was imported directly from the stl files and scaled so it should be the correct size and shape. The heat sink and fan is coupled directly onto the hot size of a Peltier and the cold side would be directly coupled to the stepper housing, which would be insulated.

The extruder and z-axis

I’m not sure if the fan cooling for the hot side of the peltier heat sink will survive for long at 100C+. If not I will need to build something custom.

If you want to take a closer look at the model you can download a copy from the google 3d warehouse.

Thoughts?