MENDELMAX - Quick-fit X and Quick-fit Extruder/s

Back in October I started on a number of MendelMax builds. Time passes I got distracted by Christmas (and the excellent Slic3r)   New Slic3r 0.7.1 out NOW!

It has been a really amazing period from October-Feb with so many new machines being released into the community, my own needs and desires for the 'perfect' machine are now starting to materialise.

MendelMax isn't 100% perfect, but it's the best machine I have built to date - and I  really like it.

MendelMax.

I have now been using a 'finished' MAX for the last 6 weeks, It's really great, and easy to hack with, I have already upgraded quite a few parts and built a whole new X Carriage concept based on using the MAX for a while.
MendelMax Created by kludgineer - the files and many derivatives can be found on Thingiveres Here - 
I have added links to my changes below -

I built up a frame concept back in 2010 out of 20mm Extrusions, but it never got past this picture -

It would have been an interesting machine. When I spotted MendelMax I just had to build one (or more)

I'm not going into too much details about building up MendelMax, most of it is straightforward, I'll focus more on where I have deviated from the 'standard' design.
MendelMax home page is here and the MendelMax GoogleGroup is here 

So firstly all the 4 different machine designs are +80mm or more longer on the X axis, so I can fit a Dual+ extruder (and my new X carriage, see below).

I decided to get different sets of Aluminium extrusions cut for four different sized machines. - I need better names but -

1) - The normal model is just 80mm bigger on X - this would be more than adequate for most situations and is very similar to a Prusa Mendel in size and shape.

2) - The mini model is more like a Huxley, it's going to be a single extruder (Bowden Fed) and designed to be very strong and transportable (150mm x 150mm bed size)

3) - The Big model is what I have built up first, this is going to be for my hack and test machine, with the ability to print upto 305mm x 305mm bed size.

Big Max and Mini Max

4) - The Massive one is a little silly really, and I'm not sure I'll build it up, and certainly not with 8MM smooth bars.

The Frame around the Mini Max above is the lower section of the Massive model, build area is about 450mm x 450mm, not sure what the Z would be, around 320mm+ I expect.

I'm using Bosch-Rexroth 20mm Aluminium Extrusions from KJN here in the UK - They are a really excellent company to deal with, and offer great pricing and cut to length service as standard!  - Do take a look.

This was my stack of Extrusions - 2.7M long package on a wood base to keep it all straight.

You will need to Tap the ends of almost all the frame sections, and with Bosch Rexroth you need an M6 Tap - the Mitsumi Extrusions listed on the MendelMax home page use M5 - so check before you buy the bolts!

This joint is the most tricky to join and get right, it's also a little tricky to print, check Thinginverse for a really nice derivative by jib also easier to print.

Assembly is very easy, you can use official T-slot Nuts or normal Half height M5 nuts (much cheaper)

 

Or, if your feeling adventurous print your own plastic T-nuts, I made a dual version and plate here

I have used a few of these on my first Max and they are holding up perfectly, I will try building the next one completely out of these.

 

I'm using some metal reinforcing brackets on the bigger models, it also means you can leave off the plastic bracing pieces from the top of the machine, looks a little nicer in my opinion.

It was about 5 degrees C outside when my Bling machine printed these.

I don't recommend printing in the cold! - Ambient needs to be over 13 degrees C for successfully printing in PLA, it's a nightmare below that temperature.

Y Axis - I'm now using the V2 Motor mount, not that I had an issues with the first version shown above.

Also two Idler mounts and two bearings, this allows the belt to drift a little and so does not rub against washers etc. - Well worth using two bearings on both X and Y axis.

NOTE : - Just spotted this really nice Y idler with built in belt tensioning by DanFineArt - very cool.

As this is a bigger machine, I'm using wider belts, about 8mm (cut-down from 10mm) with a 2.5mm pitch.

Really nice LM8UU Holder for the Y axis - by jridley 

Simple MDF Clamp for the Y axis belt, I'll fit an adjustable one at some point, or change to this Y idler 

I contemplated a 20mm Alu frame for the Y axis, opted for MDF just to get it running. 

Heated build Bed.

I considered designing a bigger PCB heated bed, and also an aluminium one, but as this was a test to see if I would need, and actually also use a bigger build area. I firstly decided to use a standard V2 Prusa PCB and a standard copper clad PCB to spread the heat over a wider area.

I'm using the 'spare' hole in the middle of the Prusa V2 PCB to clamp using a countersunk M3 screw into my copper clad PCB board.

Simply wired up and clamped with M3 screws in the corners and middle, thermistor is on the copper clad board at the side. The middle screw is also acting as a cable strain relief.

I'm not a believer in measuring the temperature in the middle of a head bed, I always measure at the edge or corner.

Z Axis - 

NOTE : - I'm still using M8 Stainless smooth rods on all the MendelMax machines, but if you are thinking about making a MAX (or any big machine) with a bigger build area than 300mm x 300mm then I really would recommend going to 10mm / 12mm, you will find 12mm costs less for the Bearings but the rods are quite a bit heavier, 16mm is overkill.

LM10UU and LM12UU versions of MendelMax parts are starting to appear on Thingiverse

Back to the old style Z design (Clamp, rather than push-fit) - slight modifications including an adjustable Z- height screw/spring. Slightly, shorter, fatter and chunkier all around.

I always add Kapton tape to the X bar going into the Z ends, helps them grip in the plastic, I do this even if using push-fit Z ends.

I'm using Aluminium Flexible couplers on Max - these are really great if your Z axis is supported. They should not really be used on designs where the Z axis is hanging (Prusa and Huxley) or compressing (Printrbot and Wallace) without using some sort of support for the Z threaded rods.

I have started seeing a lot of people fitting these to designs that really need a solid coupling, for that I would always use Nopheads Motor coupler

On MendelMax the support for the Z rods are provided by Thrust bearings (shown above).

I'm also using M6 threaded rod for the Z axis over the more common M8, I find it gives better results and does not fight with the M8 Smooth rods for control over the X carriage placement in horizontal alignment when moving the X carriage vertically.

The original Z supports looked like this, and still used M8 Thrust bearings with a washer on top to allow for the smaller M6 threaded rod, quite a noisy arrangement, so they got changed to the below versions after first testing of the machine. 

This is based on the Z axis mount design by AlephObjects my version of changes the Thrust bearing size so M6 Rods can be used and also fills up some holes in the AlephObjects version along with making it a little easier to print (mainly the vertical curve on the back - it now has a sharp bridge point).

My version for the M6 threaded rods can be found on Thingiverse here.

New X Carriage and Quick-fit Extruder's  - 

Even with just two M4 bolts holding on my Greg's accessible extruder I still find it a pain to swap-out extruders so I decided to re-design the X carriage to allow a quick extruder change and also make it big enough for Dual extrusion and mounting other things -

I'm not yet experienced with OpenSCAD so I used Sketchup for this design, if anyone apart from me finds the idea useful it would be great to do it as a parametric OpenSCAD model, I can already see a need for a LM10UU version.

I'm getting to grips with Sketchup now, you can do quite a lot with it.

It's up on Thingiverse Now -  if you want to print one.

My aims for this design  -
Horizontal X design compatible spacing with Mendel,  Prusa Mendel and MendelMAX
Use Linear bearings (LM8UU) and make them very easy to fit.
Quick-fit and removal of Extruders (And other devices... more on that shortly)
Maximum space for extruder/hot-ends/heat-sinks between the smooth X bars
Allow dual or more extruder's to be fitted at once.
Allow for an easy extension of  the X carriage (for future upgrades) - that needs some explanation, see below.
Printable on a Prusa, Huxley or TOM without support material.

This is my design for a 'Quick-Fit' clamp and expandable X-carriage - 

It's printed vertically in two parts, without the need for support material - bolted together.

A key feature about the design is that you can extend it, even when it's fitted on the machine, you would unbolt the two halves and insert a printed spacer each side then bolt back together, allowing triple extruder's to fit or something bigger...

4 linear bearings LM8UU are used and provide the widest space for extruder's whilst keeping compatible with Mendel, Prusa, and MendelMax X bar spacing. 

I know it looks big, but actually is only uses a little more length and when you have a normal Greg/Wade extruder fitted to any other X-carriage it usually overhangs by almost the same amount.

It uses a small Mendel style Bearing for the Quick-fit latch.

The latch is quite tall in this picture, it's a little smaller now.

I have updated Greg's Extruder to be Quick-Fit and defined the spacing for the standard carriage adapter plate to be 100mm wide.

It has an adjustable belt clamp.

This is an example adapter for a normal Wade/Greg extruder to turn it into a Quick-fit version.

Lower profile Catch 

This shows a Dual extruder setup on the Quick-Fit adapter - using 2 x NEMA 14 (Huxley style) extruder's side-by-side.

Here is a Short video showing fitting and removal of the Quick-Fit Extruder on a Quick-Fit X carriage. 

And click here if you want to watch it on Youtube

Extruder - 

So this is what my Quick-Fit extruder looks like, you can fit and remove it in seconds.

Quick extruder connection is now Via a 9way D-type and 2way for heater.

Similar to how Nophead has done it on Mendel90. 

 

My setup is for Dual extruder's using the RAMPS 1.3/1.4 Electronics. 

 

*EDIT* - this Amazing Gearset is by triffid_hunter 

It's Thing 10707 Here - I have used 4 different types of Herringbone gears over the last year, and this is the best in my opinion.

I printed this with a normal 0.5mm nozzle, I'm sure it would be even better with a smaller nozzle if you have one.

Herringbone gears have a long history now and there are lots on Thingiverse, do also checkout the Awesome Herringbone Gear set from Rhys and Sublime and I also really like the look of this one  by RevK but that's one I have not had time to try yet.

I have also switched over to using Greg's new 'Guidler' - 

My version (shown above and here on Thingiverse) is slightly modified with a lever, more support for the guides and an M3 nut trap on both sides.

Hot end - 
My current hot-ends of choice are J-Heads - I could do a whole post on how good they are. I do make a few changes before using them - Here is what I do to a normal MK-IV hot-end -

All of this is done for mounting, cooling, support of the PEEK at it's weakest point and to stop as much heat from rising up the PEEK as possible.

Take a standard M12 Linear Rail mount (£1.50) and drill out to 16mm or slightly under.

Widen the mounting holes (more than in this photo) and de-burr edges.

You then have the minimum amount of Aluminium for support and maximum cooling.

You could start with a M16mm rail mount, but they are much bigger overall and will be heavy.

You can drill more holes to allow better cooling.

Fit and tighten around the PEEK/Nozzle end this way around.

You can then fit 12/16mm M4 Nickel coated brass PCB stand-off's to the extruder body 

Screw in Stainless steel M4 screws and washers, clamping the PEEK into the extruder nice and tight.

Optionally also fit a nut on the thread, to lock it on the stand-off.

After this fit your heating resistor (6R8 or 5R6) and thermistor and do remember to insulate the brass part with Ceramic rope/tape, glass rope/tape and/or Kapton.

For Printing in PLA I use a small 40mm directional fan blowing across the Aluminium rail block and across the fins in the PEEK, for Printing in ABS, don't use a fan. 

New Heated Bed surface - 
When building the big Max, I needed a glass bed, but wanting to keep down the weight I went looking for some 2mm glass, didn't have much luck and out of desperation I decided to try using a mirror, it was a good test size and easily came out of the nasty plastic frame. And it was only £1 from my trusty home shop.

If you try it, let me know, my results using mirror-glass have been really positive and interesting - a few observations -

• The glass surface heats up fast and spreads heat very evenly over the surface, cool-down is quick but seems slower than just plain glass alone.
• I hear less 'cracking' of the parts as they cool down.
• For me PLA parts never seem to completely lift-off when cooled, don't ask me why this is, maybe thermal expansion, retraction is acting differently due the the reflective surface, but they stay a little stuck down - still very easy to remove with fingers, but all other glass surfaces I have used to date make the parts lift completely when cooled.
• I needed to reduce my bed temperature from 60 degrees down to 45 degrees for PLA when using the mirror surface, if you don't do this it's like building on jelly, the PLA stays so very soft.
• I think you could go lower than 45 degrees, I'll investigate at some point.
• You need your first layer height set perfectly, (can be higher than normal) if not it seems to make more of a mess as the PLA is seems much softer for at least the first 3mm of building.

The mirror is also very handy for monitoring first layer bonding and looking at the underside of the X carriage.

I have just also started using mirror-glass with ABS (with PET tape on the surface) - I can reduce the bed Temperature to 85 degrees C and get no warp even on very large parts (120mm+ wide) I have tried with three different ABS types at the moment, but they were my worst for warping (Pink,Yellow and Blue).

2mm Mirror Glass with PET tape for printing ABS

I would be highly interested in any feedback on this topic - try it out if you can.

UPDATE - While I was typing this bit up, a post appeared from Theodleif on the RepRap Forum suggesting using a solution of sugar water painted onto a hot mirror-tile for printing ABS, sounds interesting and If we can ditch the PET tape it could be Great. I have not tried this yet, but will give it a go soon.

Electronics  - 

I used a scrap piece of Acrylic for the Electronics and PSU mounting, I wanted to make it all very super easy to access and tinker with on this machine.

The Acrylic plate slides into the vertical 20mm Slot sections that are also being used as reel mount for plastic filament reels.

A 240W 12V Power supply was mounted on the left.

With nice and easy Ring terminals for the connections.

One the left we have Mains 240V

Then the main +12V and GND connection to the RAMPS board

Heated bed source and Various other +12v signals for Lighting and fans.

RAMPS is simply mounted with M3screws and an ultra quiet, slow moving 80mm Fan for cooling the stepper motor drivers and Mosfets.

Using Stepsticks again and small heatsinks.

When you connect to screw-terminal connections (as on RAMPS) you should crimp the wires into wire ferules before inserting into the screw-terminals.

The smaller wires also crimped in are for an LED indicator on the outputs of Heaters 1 and 2 and also the Heated bed, quite handy to see what's energised. I need to make a little printed box for the LED indicators.

That's my first MendelMax - I'm really pleased with it.

A short video of the Big and Mini MendelMax machines

Watch it in full HD on my Youtube channel here -

More News - 

I have popped on to IRC a few times (as RichRap) It's really great, and opens up a whole new level of community spirit and help that the RepRap Forum does not convey. BUT it's a real time zapper, I would love to be on there more, but I'm going to have to finish some more projects first or they will never get done.

And I finally decided to get on Twitter, I going to blame Raspberry Pi :) for that, but I did eventually end up getting a RPi on order, so it was worth the early wake up, and Twitter looks another very interesting community for 3D printing.
I'm on Twitter as RichRap3D You can find me Here -

Mendel Bling Sold - 

Many thanks to everyone that contacted me before, during and after the Bling eBay Auction, I spent the following week getting messages back to everyone,  It really generated a lot of interest.

The Winner and new owner of Mendel Bling is James. Even though James had a long way to come, we had a great evening discussing 3D printing and I must have talked for at least a solid 3 hours straight. If I do that again, I'll video it as an intro tutorial.
James is planning to use his 3D printer to assist in one of his new businesses exploiting further work and research he did in his P.H.D.
James had some really good upgrades planned for Bling, I'm looking forward to see what he does with it and what it prints next. Almost everything on my Blog was printed on Bling, It did such a fantastic job for me printing most days for the last 7 months I'm a little sad to see it go.

James on the Left above and me on the Right 

Thanks James, it was really great meeting you, I hope you enjoy using your new 3D printer ! 

Don't forget the South West RepRap meet up is this weekend (24th March 2012) - check out the forum messages and keep in touch - A big thanks to Hamish (hairykiwi) for organising this one. I hope to see some of you there.

Not a day goes by without something new going on in the 3D printing community, it's already a very interesting start to the year, I can't wait to see what comes next.

I'm really excited about my next Blog update, It's something I have been working on for a while now, finally it's fit for release, I'm sure some of you are really gong to like it.

Thanks for reading, let me know if you make a MendelMax or try out the Quick-Fit X-extruder concept, and I have many more extruder's and tools planned for this X carriage, I hope you find it useful too.

Rich.

February update of RepRapping adventures

Here's a quick update, it's been a busy month and I'm still in the middle of writing up some longer blog posts, so here's a taster of what's to come.

I have finished building up MendelMax and have been printing with it all month, it's performing stunningly well, I'll do a blog post on it soon.

I have discovered a new Print bed material, still testing - more details soon, but it's working so well for PLA I don't think I'll ever use anything else.

I have designed a new Extruder - and a few other machine parts - more on all that soon.

Also been messing with Printrbots, they are very neat little printers.

While I remember -  RepRap Pro are running another Masterclass build weekend. in April - These are really great fun and you have a 3D printer at the end of it! - The one last summer was one of my RepRap highlights of last year. If it's an open house like last time I will try to come along and say hello and help out.

The South West / West of England RepRap group are planning another meet up in March, I would really recommend coming along, they are a lot of fun and you will always learn something new. If you can get there check out the forum messages and keep in touch - Hamish (hairykiwi) is organising this one, if you want to ask any questions or send him a message.

I had a small job to do for a friend recently, and I was quite surprised how well and quickly I managed it, it just reminded me how useful 3D printers can be for all sorts of things. This accessible technology really is changing the world of design and manufacturing one printer at a time.

My friend needed a small adapter for a Underwater digital camera housing to trigger a slave flash unit via an optical cable.

You can buy an official cable for this but it costs over £45 + VAT and delivery, he already had an older cable that fitted on the flash end, but with the connection for a previous camera housing, so we cut that off.

Having a full machine shop he tried to make one, and failed as the fitting is small with a friction fit and tapers slightly with a small section to lock orientation.

But It's perfect for 3D printing.

This is what I made up -

Less than 10 minutes doing the design in Sketchup.

I sliced it and printed three of them in 19 minutes on MendelMax.

It fitted perfectly very first go - a nice tight push fit.

One drop of super-glue and the optic cable is fitted in the hole, and we have a removable slave flash cable.

Printed in Faberdashery Fire truck Red it's almost a perfect match with the housing.

I can now be very confident that the things I design, come out and fit exactly how they are intended, I think I'm at that point where I now think I can 'make anything' >> A little over confidence for you there :)

* For Sale *

And finally a small sales pitch - I'm selling my Mendel Bling machine, mainly because I need the room for different / further printer experiments and it seems such a shame to break it down into bits when someone can love and care for it.
If you are interested it's up on eBay now - and I'm offering to show the winner how to set it up, install software and print some things with it when it's collected.

Mendel Bling up for sale on eBay - grab a bargain now.

It's a beautiful machine, built with Love

Thanks for looking.

Rich.

Slic3r is Nicer - Part 3 - How low can you go?

This post is about printing small layers in Slic3r. It's Part 3 of my Slic3r Overview, I hope it's useful.

In Part 1 we setup some Slic3r settings and performed the important Extruder Setup
Part 2 looked at print speeds, filament dimensions, heated beds and did some basic printing.

After Neil Underwood's recent Blog post about ultra-small layers I decided I really needed to experiment, but I wanted to see and compare for myself at what layer height it was pointless to go lower. I didn't fancy messing about with getting my bet 'that' level so I made and sliced a simple object at many different layer heights and combined the Gcode so I could have one compare print.
This was printed at a constant speed of 50mm/Sec, it's just a double wall (1mm) triangle column, I could have changed speed for the different layers, but as a starting test I wanted to see what happened at a constant speed and ever lowering layers.

It all sounded so easy.... ultra low level printing is not at all as easy as it first seems.

Fine Layers Test 1 -

Slic3r Makes it easy to play with different layer heights, it's a good way to experiment and test if both your calibration and machine setup is really good.

Before I start a multi-hour print of something I wanted to make sure low layers would work and look good on my machine and be worth the time and trouble -

Here is a solid wall multi-layer simple triangular cylinder where the Gcode changes down in height every 10mm starting at 0.3mm (300 microns - my normal print height) going down to 0.04mm (40 microns) after that I stopped the print as something was clearly upsetting the surface of the print and making it lower quality.
I started to see 'bobbles' on the surface at the 0.05mm height, I first thought the temperature was too high, so I experimented with lowering it, that seemed to make the problem no better, maybe even worse as you can see below at 170/175 Deg C.

This was my very first very-low-layer print and I obviously it had some issues...

So I started thinking that because the hot nozzle is constantly going over the same plastic and causing constant re-melting of the surface maybe causing the effect seen below - for the want of a better way to describe it - 'heat-friction-bobbles' on the sides of the object. The white flecks are quite fine and many can be rubbed off.

I have increased brightness and contrast on this picture so it shows every flaw and defect.

For reference the print material in all these tests is Greenery Green PLA from Faberdashery, it's highly dimensionally accurate and is also very nice to photograph.

For reference - The test object printed at 0.3mm all the way up.

I decided to take a closer look at every layer height change under the microscope  - 

0.3mm flat face and corner.

 

At the 0.3mm to 0.2mm layer change point, interestingly Slic3r changed build direction of travel when the layer height change to 0.2mm, in the picture you can see the effect above.

0.1mm

0.05mm

0.04mm

Clearly things are getting worse as layer height decreases.

I had produced Gcode all the way down to 0.01mm, but I didn't print any lower for this test.

Could it be a calibration issue?, too much plastic? or maybe a mechanical limit?

The Z axis on this machine is set to 8x micro-stepping, using an M8 rod this gives 1280 motor steps per mm - 0.05mm is still 64 motor steps, so Z resolution should be good enough. 

Even though I'm also only using 8x micro-stepping on the extruder it's a Wade style geared design so also has plenty of steps to ensure plastic is flowing at a constant rate even with very small levels of plastic extrusion.

I was also concerned about my choice of test object, being a triangle with solid walls the point requires much less volume of plastic, and I started to think the corners were getting over-stuffed - I'm not sure how Slic3r calculates this compared to Skeinforge, but maybe performing this test again with a Cylinder or Octagon will shape object would help see if this is why I'm seeing this issue...

I was very tempted to publish the Blog post at this point and ask for help and advice, but I wanted to know why my machine setup was not looking able to do layer heights as low and as easily as Neil had achieved with Slic3r, so on with the testing.

I have never been very happy with using Springs on my print bed, recently I changed to very stiff ones and now with a glass bed I expected no more issues, but it was starting to worry me that the spring-bed could be causing my issues with low layer printing. - I tightened everything ad hard as I could get it (Still using springs) and tried again, the print was better but not significant enough.

Second thoughts - Could it be just the speed of extrusion? maybe the tiny amount of plastic does not have time to form a reliable bond to the layer below - the corner of the object (slower point) is notably more 'solid' and shiny at 0.04mm layer point. - It was time to change something and try again.

Fine Layers Test 2 -

Different shape object - Octagon test - 

In an attempt to rule out issues with my first Triangular test object I performed a similar test on an Octagon shape, this time starting at 0.2mm, then 0.1mm and 0.05mm and finally 0.025.

Here is how it came out - I have almost exactly the same problem again.

Close-up shows a similar result to the first Triangle tests and as I was printing this slightly faster (60mm/sec) than the triangle (50mm/sec) I was now getting more convinced that speed was the problem with really low layer heights. - Temperature could also be playing a part, but not quite as significant as speed.

With too much temperature (190) the very low layers (0.05 / 0.025) were practically printing by just extruder ooze, when you reduce the temperature to 175 the ooze is minimised and so the print quality also improves.

Niel's red pink panther lady did look a little 'furry' in the photo and with no compare or close-ups I was not sure if he had seen the same results as I had. Niel was using 1.75mm ABS, I'm using 3mm PLA, so I didn't expect to get to 0.01mm but I expected better results at 0.05mm and 0.025mm - time to investigate speed and temperature 'v' layer height.

After a few more test prints reducing the temperature to around 175 degrees is about optimal for layers lower than 0.1mm but you still need to reduce speed, see below. 

Better with reduction in temperature, but still much too fast for such a low layer height (0.025mm)

This was a final test to see if I could beat my extruder ooze by speeding up the print, I concluded that slowing down was actually the only solution for a clean print at layer heights below 0.05mm.

After reducing the speed to 30mm/sec finally the very low layer prints look good, I think that had finally cracked it - time to print something more interesting.

Fine Layers Test 3, Magic Fish -

Many of the very low layer prints I have seen (lots of stunning ones from Ultimaker users) are hollow sculptures or usually more vertical prints, I was very keen to try a low but highly domed print as these can be some of the hardest to print nicely and finer layers should make a big difference to the overall quality.

My selected object for the next test was the Magic Fish (Tovervisje) by bpijls this is a really nice part, a little tricky to print well, so a good test of lower layer printing.

7 test fish later I had a good set of settings to help print a low layers.

My Reference print @ 0.3mm / 65mm/sec

A few close-ups on the tail of the fish as this is an extreme area for both print bridge and overhang. 

(All these fish are printed with 25% infill)

Reference print -  standard 0.3mm layer height - very quick reasonable print.

0.2mm layers, 2 perimeters and only 2 solid layers (2 x 0.2 is not enough)

190 Degrees C and 60mm/sec Perimeter and 80mm/sec infill

At 0.1mm you need to reduce print speed, reduce print temperature and increase number of solid fill layers...

If you don't reduce speed and increase layers it will end up printing something like this above.

Internal infill still builds ok at 60mm/sec (see above) when at 0.1mm layer height, but at 40mm/sec solid layers also print well on top of loose infill at 0.1mm so again slower is better at low layer heights (below 0.1mm)

You need to experiment to get the number of solid layers and perimeters correct for the type of print you are trying to do, this is so much more critical at very low layer heights than with 'normal' printing.

0.1mm layers, 3 perimeters and 4 solid layers.

0.1mm Layers and 40mm/sec - Finally a nice looking fish, the only issue is the eye.

As these are the last few layers to print, Slic3r is not yet slowing down so dumps layers too fast - Slic3r will soon have the ability to limit minimum layer time so this issue should be eliminated.

I was just about getting sick of so many fish prints, but just one more - this time at 0.03mm layer height, I didn't note the build time, but it was well over 1 hour to print this little fish-  (30mm/sec perimeter and infill)

The first three layers of Gcode are done at 0.2mm to help get the build surface level, then it's two layers of 0.1mm and then all the rest at 30 microns (0.03mm layers).

At this layer height I could hardly see the extruder gear moving at all, you need to stare for many seconds to see it's actually extruding. - Higher gearing and 1.75mm filament should help and may also further increase quality for ultra fine layer printing.

I'm still not sure it's worth all the effort going below 0.1mm and I expect I will not go below 0.05mm very much if at all in the future. At least now I know to Slow down and reduce my temperature if I want to go under 0.1mm layers. - I'm really interested in your experiences with low layer printing, do leave a comment below.

Enough of the fish, I hope you give lower layers a try, it will certainly teach you a lot about your printer and among many other things test the quality of your printing material.

Future things to try - 

• ABS with very low layers
• 1.75mm Material
• A bigger Extruder gear ratio
• Low layers with Bowden extruders.
• Compare with Skinning in Skeinforge - I have only done a few so far to test the option.
• Test the new Version of Slic3r as soon as it's Released ! :)

Until next time, I hope you liked this Slic3r series, thanks for reading.

Rich.