Tuesday, August 25, 2009

Debt

Just watching the BBC announce the possible 9 trillion dollar budget deficit in the US. Saying how they find it hard to think about such a big number. Actually I think it is probably worse, but even so that was the sum of their analysis, not a sterling effort. The population census of 2006 put America over 300 million. Comparing that with nine trillion, if I don't have a missing set of zeros, it means that every man woman and child will be in personal debt to the round number of 30,000 dollars. Here in Japan we are already in debt to the tune of 77,000 dollars a head. Worth keeping track just in case someone comes knocking at the door.

Friday, August 14, 2009

Solar platform

By coincidence the cross bar on the blue shed was the same height as the spars across the roof for the panel mount, so I added another spar and I will now make a platform deck right across and then mount the panel on that. I remember sighting across to check that I could get the panel supports in, but I didn't notice my subconscious making this extended plan at the time.
The press has gotten a little covered in timber as I sorted through the stack of pallet hardwood to find likely candidates for the decking.

Wednesday, August 12, 2009

Rusty nuts

For some time I have been troubled by jingly metalic noises coming from the back of the truck, I assumed it was some neglected piece of scrap rattling around in the bed of the vehicle, but the other day I inspected the wagon and no such matter was present. It then occured to me that it might be other and I immediately found that the exhaust pipe had rusted right off the muffler and was hanging dejected from its little rubber mount. We have been busy with earthquakes and typhoons, so I haven't felt much like disabling the vehicle. Muffler manifolds can be a bugger to get off sometimes as they swathe themselves in rust. Today the weather was not too bad and I had a bit of a brainwave to give me an advantage in the battle of the rusty nuts. I wasn't keen on messing about with a blow torch up by the petrol tank. We have a kind of insect defense, which is an aerosol spray that just freezes the bug in question (We use it on the stinging centipedes -85 degreesC in an instant and they are still OK if they thaw out). So I went and got a can of that and froze the nuts for a moment to loosen the rust bonds.
The old adage as I remember it goes, if you can't get it on take it off, and if you can't get it off put it on.
If you don't play with your spanners much my interpretation is, don't struggle to get something on, get it off quick and find out why it is being such a bugger. If you can't get a nut off that is stuck, try tightening it a little first, if it shifts a bit that way go for the big pull to get it off.
I applied the later approach and was rewarded with success. The TIG welder is the bees eyebrows for this kind of welding job and there was just enough shape left in the rust to orient the parts correctly. I treated the thing to a neat new end on the pipe as well with a scrap of stainless pipe.
I had to drive the truck up the road to some flat ground to jack it up and get the more cumbersome welded unit back in place, it made a lot of noise with no muffler, but it all ended nicely and we should be OK for a few more years.

Monday, August 10, 2009

Practicalities

The practicalities of yesterday's circuit. A big black 12v battery, left. On the chopping mat at the top is my little LED sensor array with five LEDs oriented to poke out from a little stainless steel pyramid painted black. That is the one part I have finalized. There are two pairs of LEDs wired in series that feed into the big breadboard in the middle and each switch one half of a little quad opamp chip. The amp is only wired up on one side at present and has two wires running to the H-bridge that runs the little motor. That is the circuit from yesterday and it is wired up on the little bread board stuck to the shield on top of the Arduino board. I use the pen light at the bottom to shine on the individual LEDs and hopefully produce the appropriate response. The arduino will be wired up between the opamp and the H-bridge once it gets out of the quarantine phase it is in at present.
The other chip (right) on the big bread board is a smaller H-bridge that is only good up to 1amp. I used that to check the wiring on the op-amp during earlier missions into the land of electronica.
Incidentally the dual MOSFET arrangement did not like anything under 5v fed into the Gate pin, I had to get out the amp data sheet to make sure it was OK with 6volts, then the thing started to work.

Sunday, August 09, 2009

A little knowledge

I found out that a lot of the sites I had visited in relation to MOSFETs were overly complicating the issue. The diagram above from this site translated into a good solution for me, but at the head of the page it says that there is an error in the circuit. I must find out what that means before I start soldering.
The essential points are that all the MOSFETs are N channel and that the ground for both the Test Circuit in the blue box and the 12v area being switched must be common.
The pull down resistors are also vital, I am using ones valued at 100Kilo ohms, which work fine. I was getting bad results because I didn't have those essential pieces of information laid out in plain English and my knowledge is very limited. One exceedingly annoying thing is having positive and negative so jumbled by the world. Everything runs on electrons and they are negative, so all that time you think of positive as the power source you are wrong, the electrons are obviously coming out of the negative and flowing to the positive. This makes the labels on the MOSFET of Drain and Source perfectly reasonable, but why convention has us thinking of things the other way round I am not sure. Even here the arrow over the motor shows the flow in that direction, which is wrong. I guess it all keeps electricians up with their boat payments. This kind of deliberate loopiness in general knowledge can drive one nuts.
The positive field created in the MOSFET when it is connected to the switched 9v supply does not dissipate quickly, it will gradually fade, but having the resistor in place to give it somewhere to run to once the power is turned off makes it a predictable and spontaneous reaction. The MOSFETs I have will switch fine on 2-4volts, so my little operational amplifier circuit could get them running properly on its own, but I want to incorporate more functionality, so my next step is to replace the "switch" in the test circuit with output from the Arduino microcontroller.
The diagram shows the motor powered in one direction, reversing the switch will cause it to rotate in the other direction. With the black line lit up. I need to control two motors in this way, so I will use a relay to switch between them having them operate one at a time so that I don't need to send off for four more MOSFETs to make an independent H-bridge for each motor.

Friday, August 07, 2009

Lamps

The session on stained glass we had a few weeks ago started my daughter on making some little lamp shades. I promised her I would see about making up some kind of fitting to hold them when they get done, she is getting close to finishing one, so I set too at that yesterday and welded it up today. The light fitting stems and sockets were from a skip at a local shop. The brassy looking tube is one I haven't annealed yet, the three poking out of the blobby wall bracket base have been done and bent to a little shallower curve. Annealing steel is done by heating to a nice red color at least and then allowing the metal to cool slowly. It was fun beating the blob out of a flat sheet. I wasn't sure how much I could get it to go, but it worked out OK and it wasn't too hard to hit out a flat rim around the edge and apply some texture to it in the press. The leaves are blanks cut out of the same plate as the blob and then struck a few times in the little pattern I made up ages ago for door window decoration. They are then teased into better shape with pliers and whatnot while they are red hot. The flowers didn't grip too well struck from the thin plate, they really needed to be pressed out of a thicker blank heated in the press before striking, but I added a few little stamen things with the TIG welder and they look OK. Some encouragement for missy to finish the glass bits I hope.
Meanwhile the solar project hit a bit of a snag today as I hadn't realized one of my parts was cast iron. Not good for welds and one of the welds failed as I was testing the thing in the shop waggling it about under load. I will have to rethink the attachment. I am also struggling to comprehend the operation of N channel MOSFETs, the web is a lot of help, but my results on the floor so far are not showing the degree of repeatability one requires from a rude mechanical. I was delighted to find the operational amplifier relatively simple to comprehend, but the Mosfet is essential to handle the switching on and off of the motor as well as changing its rotation direction. Even the little window motors draw about 3 amps when they are working, so little hobby transistors or H-Bridges that only handle 1 amp are no good.
Given time I hope to prevail.

Wednesday, August 05, 2009

Ladies day

A lot of classes take a holiday for the summer, but the three ladies were keen to finish off their little shelves today. The second session was again around five hours, so they have about ten hours of work in these. Putting on the finish at the end cheered them up. The beach wood flooring I had pulled out of the local gymnasium before it got torn down was fairly dull until it started to soak up the stain, then some parts came to life. It turned out a good choice of wood for its tollerance too as some of the tenons took a while to fit. I think they are all intended as little odment shelves to go somewhere against a wall. Hopefully they will all find a happy place.