Coffee Machine

The office coffee machine broke. I need coffee. It's a simple drip filter machine that only cost about 15 quid but there a principle at stake here. It was obviously blocked,.Descaling didn't help, and nor did taking the feed tube off the top, so it's got to come apart. First problem is the bottom is held on with triangular recess screws. An old screwdriver plus a bench grinder solved that problem. After that getting at the inlet was easy enough. There is a small valve in the inlet spigot. Sucking and blowing on the tube showed it was not working. After a little trial and error, it turns out to be easy to push out from the reservoir side with a piece of wire. At that point, a lump of paper towel fell out of the hole. Job done.

Rubics cube

I finally decided to get a new tractor. Well, I say new, 6 years newer than the old one. A Ford 4610 with what is colloquially known as the "Rubics cube" gear shift. Having tried randomly wiggling the lever about, searching the web gave no clue about how it worked. I finally resorted to buying a manual. Out turns out it is fairly straightforward, although completely unintuitive. There are three neutral positions. The one in the middle of where the starter interlock is. The ones at the ends allow the lever to be lifted, from where a secondary movement forward or back selects a gear.

All this is the right hand lever. The left hand reverse/ low/ high one is much more obvious.

Walking Wheel

The Lady of the House wanted a great wheel or walking wheel, a type of spinning wheel that predates the modern flyer-type wheels. Actually I think a great wheel is slightly different from a walking wheel, but the idea is much the same.
Basically it's just a big wheel driving a small spindle, so surely that must be easy enough to make from a old bike frame...
This idea was to use the rear triangle and bottom bracket. The spinner has to turn the wheel by hand, and keep in contact with it, so half the rear triangle had to go. To get the wheel to align reasonably with the spindle pulley (or whorl as spinners seem to say) I had to crank the chainstay over considerably, which needed some hot forging. More hot forging to draw out a spindle from a bottom bracket axle. This comes to a fine point (it's what Sleeping Beauty is supposed to have scagged herself on). A 3D printed pulley finished that part off.
Now, it needs to stand on something. It needs to be at a convenient height to use standing up, so my first idea of using the headset and seat tube was out. I cut the front tubes off, and used the top tube to extend the seat tube to the right height. A flat handlebar forged into a smaller angle makes a reasonable tripod base, although not as robust as it might be. Still, it works.

And here it is...

And a detail of the spindle

Forge extraction

The forge is now housed in its own shed.

Lovely, isn't it. It also has a better blower. The heater box really wasn't doing it, so I bought a bouncy castle blower off ebay. That does it - maybe too well.  The only trouble is, if you look carefully you can see that the fumes aren't actually going up the chimney. When there's a lot of smoke coming from kindling it is unbearable. The first chimney was 1m of 5” flue. I tried upgrading to 1.5m of 6", and raised the hearth closer to the cowl; no better. Then I noticed that it was drawing OK before I turned the fan on. As soon as the fan came on, there was a massive downdraft pulling the smoke back out of the flue. Using a bouncy castle blower there is a lot of spare draft, and with the vent closed it all comes out of the bottom of the tuyere, through the ash plate.

So, we need to put that waste air in the right direction. A T piece in the air duct was easily fitted, allowing me to fit a length of dead exhaust pipe ducting up into the cowl. This didn't really help at all. It caused a lot of turbulence at the entry to the flue, and I could see the smoke rising up then turning around and coming back down. Extending the duct right into the flue did it, now there seems to be a nice laminar flow right up the chimney. I can now light the forge and breath at the same time. Here's the "after" view, with added lights as well.

 

Forge v0.2

The forge worked OK, but some improvements were needed. The hot zone was quite small, and I thought the relatively small duct was not helping. Off came the tube elbow, replaced with a 2" tube (once the housing of a Landrover shock absorber). I joined that to the blower with some 65mm aluminium duct (actually bought for the purpose) with a tomato can as a joiner (normal service is resumed). To allow more air through, i put more holes in the tuyere plate.

I managed to get hold of some proper forge coke. Lighting it was an absolute nightmare, it took about 6 attempts. I think this is because the coke was damp. I blocked up some of the new air holes with fire cement, in case the problem was lower speed of the blast. The air still comes out over a wider area than the original version, which does seem to have made a larger hot spot. Maybe I'll open them up again one day to compare. Anyway, it seems to be working, eventually.

Success!

Finally, we have a print! Still needs some refining, but it is getting there.

One of the problems was a tendency for the extruder to jam. It looks like there is an insulator missing, which doesn't seem to be causing any thermal problem with the Arcol hot end, but does leave a large void at the bottom of the Wade's extruder where the feedstock can block up...

RepRap Heated Build Platform

I have been trying to get a RepRap working, on and off, for a few months. It was bought second hand, supposedly complete except for a power supply. Obligatory tat item: I have a spare ATX PSU, which ran the main RAMPS electronics and extruder fine, but cut out when I tried to power up the heated bed. This is apparently a known problem; the solution is to put a dummy load on the 5V rail. I tried 4 off 3.3ohm 10W resistors in parallel, it helped, but it still cut out whenthe heater cut in. The resistors got toasty though. Which made me think:

Instead of wasting the power going through the ballast resistors, use that to provide a base heater load and then have a lower power - which can be handled - on the 12V supply to give control. It needed nichrome wire in 18SWG and 20SWG. The original build platform was a piece of unetched circuit board, which I re-used, sticking the wires onto the non-copper side as shown in the diagram using Kapton tape. I think the copper should give better heat distribution, but I have no proof of that. I put Kapton over the copper and pinned it all down with bulldog clips, and it works just fine. The shorter (green) track is 18SWG, attached to the 5V supply. The longer   (red) track is 20SWG, and is connected to the D8 terminal on the RAMPS board.