My Sumpod 3d printer now has several upgrades. I've upgraded the software to Marlin (Thanks to Stohn and Erik for the firmware.) This gives noticeable improvements in print quality and allows the printer to be driven a bit quicker.

I've upgraded the hardware adding:

• an SD card reader (for printing without a computer),

• an rotary encoder (knob) to control the printer and navigate the SD card, and

• a speaker for feedback from the rotary encoder. (I'm planning to add beeps for error events in the firmware so it is more obvious when something goes wrong or when the printer is preventing me doing something stupid - like extruding when it is too cold.)

This morning I printed these coins for my children:

They are very happy with them. I uploaded the designs to Thingiverse as thing 23369 and thing 23374 respectively.

Most of my other prints have been things for the printer itself, such as a small mount for the SD card reader and a spool holder.

Now my Sumpod 3d printer is working, I've been taking requests for things to print. Tracy designed these with inkscape and I did the rest:

Caleb asked for a blue snake. I did my best to design one:

He loves it which is what matters but I think it needs more work - both the design and the printing. I did rather like watching it print especially the hex in-fill:

and it was good to see that the printer coped with the slight overhang created by building up and out on the lowest layers.

I initially used spheres of blutack to level the bed on my Sumpod 3d printer. This worked surprisingly well and was good enough to start testing but wasn't very robust.

For something more robust, I followed Stohn's method. The components I used were:

• M5 Stainless Steel Countersunk Allen Screws from ebay,

• M5 captive pronged tee nuts from ebay,

• 9.5x15.9mm springs from a set on Amazon, and

• washers from a set on Amazon.

For each corner in turn, I used a screw with a large washer (to spread the force) through the platform without the bed in place to drive a captive pronged nuts into the platform and then removed the screw again.

I then put the springs and bed in place and put the screws back in (using smaller washers). I screwed it together leaving a bit of space to tighten it during levelling.

I don't know how others are doing the actually levelling, but I used replicatorg with X axis and Y axis homed then moved the Z axis until it was less than .5mm from the bed. I then attempted to slide a business card under the nozzle; and tightened the closest screw until the card wouldn't go under and then loosened it very slightly until the business card just went under. I then moved slowly towards another corner. Every 10mm (or sooner if it looked too close), I tried the business card and, if necessary, adjusted the screw (of the corner I was heading towards).

I'd welcome comments about better processes for bed levelling.

I've never been so pleased to see a pink alligator!

Today I did exactly what I said I'd do in the previous blog post. First, I took apart the hot end of my Sumpod 3d printer to clean it. I'm not sure if Tracy approved but found a Brulee torch (like this one) made a perfect tool for cleaning the nozzle. I used it to burn off all the PLA from an earlier leak. I kept at it until the smoke stopped. Although the smell wasn't as nice as burning sugar, then end result was almost as satisfying as a Creme Brulee. I definitely recommend this to others who've got a dirty nozzle.

Next, I put everything back together with a fresh piece of PTFE tube (from ebay). This presented a small problem as my hot end has a very slight misalignment between the PEEK/insulator and the nozzle. (This is probably why I initially failed to insert the PTFE fully and thus caused a leak.) I'd got around this last time I reassembled the hot end by shaving a very tiny amount of the tube with a pencil sharpener. This seemed to work but didn't lead to a successful print so I thought I'd try something else this time.

So, I threaded the push fitting, the MDF holder, and the PEEK onto the PTFE tube. Then I pushed the PTFE into the nozzle making sure it went in completely. I then screwed the nozzle and PEEK together - this way the PTFE is pushed in really tight. Finally (with some care to push the fitting down) I screwed the PEEK and push fitting together. This seemed to work better than any of my previous attempts where I screwed everything together first and pushed the tube in last. It seems a more reliable procedure to ensure success than the obvious/natural (?) way I'd been doing it before.

I said that I was planning to clean the filament as it entered the extruder. I did this but it was rather half-hearted effort and I don't think it made much difference.

I printed the same 20mm x 20mm x 3mm block with solid in-fill that I tried yesterday and it came out much better:

One corner is slightly less neat than the rest - the corner where the layers start:

This was quite a short print less than 30 minutes, but I was optimistic enough that I'd beaten the jamming problems that I called Richard Sum to thank him for this help and let him know I'd made good progress.

This evening I tried a more complicated (and more useful) print. This time I used the default skeinforge hexagonal in-fill and the print took a bit less than an hour. I watched very nervously as it took shape:

I was very relieved when it finished (especially as the pink filament was close to running out):

and I'm pleased with the result:

Now I have to actually think about what I'm going to use this printer for. I'm sure my lovely boys will have some ideas.

Today, I spent a bit of time cleaning out the hot end of my Sumpod 3d printer. I put everything back together making sure I was really pushing home the PTFE tube. Unfortunately prints are still messed up by intermittent jamming.

I wondered if perhaps the jamming was corresponding to the natural temperature oscillations of the heating process. I did several prints at different temperatures (195'C, 200'C and 205'C) and videoed the process with the temperature reading visible on the LCD. The good news is that the best print occurred with the set point at 200'C which was the temperature I was using. The bad news is that the intermittent jamming occur-ed with all prints and the video showed that the jamming wasn't directly related to the temperature variations.

The 200'C print started very well:

but then on the third layer there is a trough where the filament jammed temporarily:

The intermittent jamming continued right up to the top layer but there were still periods of good extrusion:

While cleaning the hot end, I noticed that the Bowden tube was warped:

I need to think about what caused this.

My next steps are:

1) replacing the Bowden tube with a fresh piece,

2) cleaning the hot end again to make sure there is no old filament around the interface between the nozzle and PTFE tube, and

3) Add a bit of lint-free cloth before the extruder to clean the filament.

On a more positive note, I attempted something like Stohn's bed levelling mechanism which worked very well and was significantly easier to control than the squashed blutack I was previously using.