Daily Monday 3rd March 2014 - Thursday 6th + thoughts

Monday 3rd March 2014
'Her Indoors' likes those little garden sun powered lights, especially the £1 ones that color change red / blue and green. Nice week-end, grand children helping her in garden, one of the plastic ends that allow the light to be driven into the soil got broke. Knocked up a new one in OpenSCAD, failed to print.

Tuesday 4th March 2014
Re-designed, printed 2 (1 for spare), job done.
We have B&Q fans mounted in all bedrooms and computer room, they are great, large ceiling fan makes it cool in hot weather, light and fan are remote control, off/on and dimmer, nice kit.
Some time back one of the glass bulb covers got smashed. Can not find replacement anywhere. Been naked bulb in son's room for over a year, but Son now moved into his own house, we take over main bedroom, so light needs fixing.
Got clear ABS from night lights.
Used OpenSCAD to create replacement design. 1st print not so good, too sloppy into fitting, too fragile.

Wednesday 5th March 2014
Re-designed, printed, fits nicely, can't actually tell difference between original glass and printed. Quite chuffed, job done.

Thursday 6th March 2014
No pending small jobs, washing line snapped yesterday, need to climb tree and re-fix, bulb gone in automatic light outside kitchen door, but nothing that can be done in evening. Decided to create this blog.

Other thoughts / possible projects
When I said I built my printer, it cost only £250 because I created the parts myself.
Hot end is actually key to 3D printer, it is the main part, at the time I could have bought a 'J-End' for around £60, but I wanted to understand. I did not have a lathe at the time (bought since). So I created the main part on my pillar drill from a 6 mm steel bolt, created the PEEK mounting using pillar drill. The heating element is nichrome wire. I've had a few minor problems with it, but on the whole it works well and it's nice and small (less than 30 mm long).
I'd like to create a new brass hot end, I have 6 mm brass bolts, I now have a lathe, attempted once, but 0.35 drill snapped off drilling hole (being too clever, should have done using hand PCB drill).
So, that project is New Brass Hot End

I would quite like to try moving Wade Extruder off X-Axis using Bowden tube. I have got the PTFE tube, but that's as far as I've got so far, not taken a serious hit at it yet, garage not pleasant in winter. Moving Extruder off X-Axis should make it lighter and faster. So, that project is Bowden Extruder

I used to create PCBs using photo etch, buy photo resistant covered PCB material, I could expose using wife's forgotten sun bed (yes, a 6 foot by 4 foot UV light box!!!!!!), but sun bed is long unused packed into garage, so I bought an old monochrome laser printer off Ebay (£12), I tried direct transfer using normal paper. I produced my 1st attempt RF Id reader using it. Looked really good during etch, it wasn't until fully assembled that I looked closely under a magnifier and realized that there were micro fractures all over the damned thing. Now using laser photo transfers from Maplin, but I would like to revisit direct transfer.
So, that project is Direct Transfer PCB

Along the way to current PCB method, I tried ink jet printing direct to PCB. I bought a cheap printer from Amazon for £30, but it was the wrong one, not flat bed and the recommended ink is all pigment based, which does NOT go through Epson jets designed for dye based inks. But that does mean I have a supply of pigment based printer inks. PLA and ABS in natural state are cream coloured. You can get clear ABS where they remove the natural impurities, but you can not currently get clear PLA. But I digress.
All coloured ABS and PLA is coloured using pigments. I don't know if pigments in ink are the same as pigments used in plastic manufacture, but I do know they are both pigments. That leads to the thought that maybe I could alter the colour during printing. I am thinking that I have 4 syringes, use threaded rod to drive plunger, use a small stepper motor per syringe, same colours as standard printer (cyan, yellow, magenta and black), maybe get some 7/02 PTFE wire, remove wire and leave PTFE thin tube. 4 colours, 4 syringes, 4 thin tubes. Tubes end at some kind of very small washer shaped metal arrangement that can be attached to business end of Hot end. Plastic extrusion gets coloured, mixing may be a problem, but extruded plastic thread could get coloured around outside (middle doesn't matter, not seen).
Modify firmware, drive overall colour feed so that colour feed is matched to extrusion speed, each colour varied to make up overall feed, quantity of each colour specified in G-Code. I noticed recently that a commercial printer has just been published as the 'worlds first full colour, multi material 3D printer', nice, and a snip at only $80,000. How cool would a Full Colour Reprap be?
So, that project is Full Colour Raprap. Would definitely mean a lot of experimentation, but doable, might not work, but really cool if it did.
Hey presto, we have Reprap full colour 3D printer. But we have another major problem. Current 3D printing is based upon STL files, STL does NOT support colour, that means creating new color STL (CTL?) then modifying OpenSCAD and Slic3r, both are open source, but quite a daunting task.
So, that project is Full Colour STL. Probably too much for one man to do, I have the skills, but it would take a very long time.
Now, I did consider a stop gap, a simple bit of software, take in a standard JPEG, that gives colours and X/Y co-ordinates. Allow user to drag out areas and/or pixels to required height, that gives X, Y and Z co-ordinates. Turn a photo of a loved one into a 3D mask. Probably not too difficult to go from X/Y/Z co-ords straight to G-Codes. I did play with C# and a few very interesting 3D libraries, but that is as far as I got, really need to prove we can print in colour first or this is pointless. But project would be Jpeg to Colour G-Codes.

I buy interesting bits on Ebay:
I recently bought a Real Time Clock module, nice little PCB, full battery backed, includes battery all for 99p!!! Like to play with it (not arrived from China yet).
Daughter would like a new night light, 2 shades, say moon and sun shape, probably white moon on always, yellow sun on demand. Needs display and a few buttons to set time like Alarm clock. Idea is that Sun comes on at a set time in the morning, show young children when they are allowed to get up. Our grand children do seem to like to wake at various times between 05:00 and 08:30. The Real Time Clock Module would be invaluable here. That is Social Night Light.
I also recently bought a pair of modules, an RF transmitter module and matching RF receiver module, both have GND and VCC pins, transmitter has 1 DATA pin, receiver has 2 DATA pins, curious, need to play with them, but both really small and cool and both cost 99p!! No idea how to use these yet.
Previously I got hold of a load of very small motors, 11 mm long, about 2 mm diameter, very fast, designed as mobile phone vibrator drivers, about £1 each, I'd like to create a fully printed quad copter using these, maybe with camera, could be really cool, not really a project yet, not really useful enough.
My daughter has some 'dodgy' neighbours, shouting in the streets, swearing, hammering at their door, she would really like to keep an eye on them. Tenvis cameras are really cool, but a bit pricey.
I have a Rasperry Pi with Camera module, got that running last night whilst Fan Light cover was printing. I have a Wifi dongle hanging about. I have a camera mounted to the front of my house, old school, B&Q special bought years ago, it feeds into an old VHS video recorder (encodes to standard PAL signal), that is fed to house TV system, picture to small TV above main TV, we call it the monitor. In the past, I rigged up a second old VHS video recorder (for tuner) and a frame capture device and fed that into PC. I created a program called "Monitor", it fed frames into video files in 1 hour chunks, it also added date/time and monitored each frame, difference between 'areas of interest' in frames caused it to output a jpeg. It worked rather well, you ended up with a directory for each day, within that directory was a various number of jpegs of 'action' scenes and 24 one hour video files. Browse the jpegs. anything intersting, use time on jpeg to look in correct place at video. So, can I re-use my old code (windows based) with Rasperry Pi (Linux based), camera and Wifi dongle to produce a small, neat Wifi stand alone controllable camera (like the Tenvis, but using already available bits)?