The Serviceman

cataract surgery

This time the servicemen were serviced! Kali and James faced cataract surgery with different results. For such a common operation it seems to be very unreliable. Several people in the Flowtrack orbit have had bad outcomes.

James is suffering bruising and pain and bad focus, but it is rapidly improving. Kali had instantly perfect vision after eye #1 and no pain or bruising. (She might have been a bit biassed after four months legally blind) But it all went south when it got dark. The reason is obviously that light is getting under the lens. Big pupil, small lens, placement too high are all good explanations, and all denied by the medicos. Apparently the brain is supposed to adapt. By not opening the pupil in the dark apparently. But Pilocarpine is a good kludge, even with the headache, the jerky vision and the unreliable cutin and wear off times.

P.S. Kali went to another eye doctor complaining about the night vision issue and he said the Johnson & Johnson lens in eye #1 has a silly lens rim that reduces the functional optic diameter. He put a SN60WF in the second eye and took care to get it placed accurately. Voilla! No night vision problem. Eye #1 night vision was fixed with a contact lens that shades part of the pupil.

Night driving with PCB00 lens

valve timing ...

No good deed will go unpunished! James replaced the engine in a friends car and it subsequently overheated and the oil filter fell off (#??). She claimed it was dead and blamed the mechanic. To finalise nastiness Kali bought the car, thinking it would be a good site vehicle for Flowtrack - 4WD Subaru.

A quick overhaul of the original engine showed that it had probably died of corroded out welsh plugs which let water into the oil and thus oil into the air cleaner which hopefully had stopped the engine without damage.

But the reassembled and installed engine would not run and much time was wasted with all sorts of theories checked. Eventually a U tube video confirmed that the cam shafts had to be timed with the flywheel set to the insignificant mark that had no relation to TDC. Why would you do that!

disk images ...

Trying to immortalise failing DOS computers we tried to make bootable CDs that could launch a machine into an environment that would run venerable programs like SCHEDIT, Lotus123, SYSTAT, DSOA, etc.

There are of course many utilities that purport to make an image of a boot floppy. But when you try to put them on a bootable CD they dont work because they have inserted headers and wrappers.
One particularly irritating utility was found to put a sprurious 13 byte header on the image that spoilt the size constraint and completely upset K3B for example.
This utility apparently expected to be the program that got to recreate the image somewhere. So James had to write a program to trim the image, suitably called "del13"


Video Formats!!!

Our indirect benefactor Keith McLaughlin recently died. He was a technical enthusiast, starting life in 1916 in the days of kerosene and horses, and sending his last emails as only days before he died. As he would have expected we compiled a lot of data about his life and ran a good funeral. The problems of file formats became extreme. Even showing a simple slideshow of his life required a creditable "red-eye" from Taree camera shop proprietor Alan Small. After the event it became worse as the house recording from the crematorium was damaged and difficult to edit.

Trying to produce an archive of Keith's life became a nightmare of formats. There were docx files, corrupt VOBs, mpg wrappers, mov videos, etc. Doing a google found a surprisingly emotional WIKI entry with suppurating angst about the state of proprietory video formats and attempts to monetise them. Keith organising his MA thesis on 
HG Wells

Clock chip tomography

After a month where lots of devices stopped working the archival SCU1 computer stopped booting with CMOS flat warnings. This was not new, and we had replaced the clock chip (DALLAS ds12b887) previously after importing some from China, but now all the spare chips appeared to also be flat!

James did a web search on this problem and was pleased to find some hero had posted photos of the dissection of this chip. The spirit of the internet lives! With this guidance it proved easy to grind off layer by layer the top of the chip until the ends of the battery emerged and could have wires soldered onto them. A couple of AAA batteries completed the job and the computer started booting perfectly.

Tail Light Warranty?

Doing the right thing Kali bought an electric fence energiser from an Australian company, Thunderbird.

After working well for a year or two it was somehow tipped crooked in the paddock and it filled with water. Disassembly revealed parts of the surface mount board corroded away and the flexible PCB connecting the switches etched back to mylar.
Thinking it was ruined and as usual being refused a schematic we sent the box back to the manufacturer who replaced it under warranty!!!

Bad memories :-(
After removing a lot of old papers from the office there lay on the floor a delivery docket from 26/10/2004 for a 4Kw induction motor from C J Pearce. The delivery was from 490 Geelong Rd Footscray to the pier kiosk at Stony Point, to go on the ferry to French Island. Kali had burned out a wind turbine generator by motoring it to find out why the new wind turbine would not start. Blade angle had to be increased.
The Pearce business plan was to supply motors fast to minimise machinery down time. And they moved very efficiently when rung. But later that day Kali received one of the most abusive phone calls in our business history. The manager had apparently got back from lunch and did not believe anybody on French Island would be using a big three phase motor and decided the whole order was a shakedown. Eventually the owner of the French Island Lodge was able to settle him.
When the motor arrived it was not good. Different flange hole pattern and other details, but worse, it did not work well. Its nameplate said WE, but whether it came from the Perth plant of Western Electric is arguable. Its remnant magnetism was a lot different from the ones we normally used and self excitation was poor.

Egg pileup!

An old friend has built a wonderful free range chook farm,
But the machine that sorts the 1000 dozen eggs into the different weights for the different boxes stopped working; eggs on their cute little conveyor belt sailed right on past the "kickers" that were meant to categorise them and ended up in egg jams at the end of the grading table!
The kickers were going off prematurely and ineffectually. Inspection led to the conclusion that the problem was the conveyor belt link sensor that was shooting off spurious sinals. Testing the sensor proved nothing. Then the distributor of the machine, Lesco, came through for us sending a circuit schematic from the U.S. which amazingly turned out to have been drawn in SCHEDIT back in 1997. The path of the sensor data was complex, but a bit of wire wriggling and power supply bypassing got the machine back into service.

Open source inverters?

After the great success of open source and open format software, surely it is time for open source inverters?
We have piles of inverters rotting in the scrap heap for want of a circuit diagram. Is this a form of compulsory consumption, or just the effect of lawyers and accountants who want to "own" IP?
PYE valve TV sets came with a circuit diagram stuck on the masonite back. Perhaps some inverter manufacturer will realise they could get an instant market advantage by supplying the schematic. They could even keep their secrets in the microprocessor rather than grinding off chip names the way some paranoid manufacturers did!

Power Electronics Revisited

After 17 years a Francis turbine controller built by Rainbow Power Company was hit by lightning. After a difficult repair the customer wanted a spare so they did not have another weekend of fifty people trying to enjoy a country retreat near a 5kW generator!
As the controller has been out of manufacture for some time we had to source parts and do a bit of redesign. For example the wafer multipole switch that changes excitation capacitors is now rated far lower than previously. They did often fail!
So Kali resorted to one of her pet concepts - a binary weighted row of toggle switches. More speed settings with fewer capacitors! Except that bigger transitions happen than in the original successive addition arrangement - and this made voltage shocks that took out the oscillator board. Another rally drive for nothing.

False Alarm

Last week the venerable 38 Y.O. Flowtrack Toyota broke down on the way to a conference. It seemed to be a dragging rear brake when lots of smoke escaped. But in the way car faults conspire to prevent diagnosis it was only the effect and not the cause of the problem. So by the time we got to Grafton the wheel bearing was obviously not going to go around much longer.

Things always seem to stop when their smoke escapes ...

Thanks to the admirable capabilities of South Grafton garages ( plug, plug), and our running around the place with the offending axle we made the conference. The locals were amused. Not much was happening so we got comments like "Lost the rest of the car?"..."You look safe!".

Then just yesterday while towing a trailer of bricks over the "road" from Wingham to Landsdown a similar sort of noise started in the right front of the car. Investigation however revealed it was only the horn being rattled on and off by vibration from the road after the potholes had shaken the switch out!

P.S. The new Japanese bearing later failed to hold oil in the diff which got all over the brakes. After much searching a solution was found; not only a bearing for a Borg Warner diff, but one from 1968.
The guy in the spare parts shop was incredulous that AMI had assembled a Toyota in Australia with a five year old imperial diff!
Moral: dont trust vehicle parts lookup programs. Take a trusty vernier calliper to the parts shop. (and remember that seizing bearings sound very different to worn bearings)


Roger was playing with a customers XP laptop that locked up when you tried to log on as admin.

He found that you could log on as one of the lesser users (who had not configured a password) and from there become admin using the correct password. James then conceived the completely illogical plan of changing the admin password to itself.

Success! Presumeably what was broken was not the password setting but its authentication process.

Kali used the same trick with DFAT to solve a bureaucratic snafu of 8 months duration preventing passport renewal. She changed her name to the same name, using the new processes of Birth Deaths and Marriages. Success at last!


We used to use a hp7475A plotter to draw circuit schematics. It was fun and multicolour etc.

Then it died :-(

Kali replaced serial driver chip and tried running it at 300 baud and eventually it became unuseable. So when our eBay expert Kevin Masterson found one for sale we bought another and after it sat for weeks in the flooded TNT depot we eventually got the thing.

The documentation on the internet was deemed inconsistent by James who did a full breakout box analysis of the interface. He determined that the thing was a DTE with some weird strapping and came up with a correct cable construction. It worked! The old one is likely repairable now that we know that the "clear to send" was strapped to ground with likely damage to the comms chip or its 12 v supply.

the MG 4512

This is an insanely cheap Chinese 200W wind turbine that everybody we know has tried to redesign. We bought one out of curiosity and a customer is actually operating it.

We put it up on a 10 metre guyed tower of 75mm square tube with Barret truss in the middle. Then we threw away the controller and replaced it with a three phase bridge.

More to get the controller off the desk than anything else Kali dismantled it. There was a littany of construction errors well demonstrated.

* Weak case had allowed several board connections to fatigue off

* Several nuts, eg dump load location, were rattled loose

* the soldering of the LM358 and several of its resistors was poor to the extent of non existent.

* Little protection from dirt, animals, corrosion.

In spite of this Kali thought it had some merit and assuming nobody would bother with copyrite of such a circuit we reproduce it here!

Circuit Diagram

the MAZATROL stops again

This time the spindle motor stopped after drawing vast current for no apparent reason. There was a phase reversal issue in the factory and other misdirections which made us tear the SCR section apart and find nothing wrong.

So eventually we tore off the motor cowling and removed the brush inspection cover and ran up the speed. Suddenly, crackle-fizz and we had a burning ring of fire around the commutator, no doubt consuming the vast buildup of brush debris not removed by the supposedly impeccable Ford maintenaince guys who last owned the lathe.

Portable Mobile Broadband

After weeks with no net connection during repair missions we have been studying mobile broadband options. They seemed to be very expensive and dependant on the evil monolpoly.

After some false starts with Vodafone we came back to the old bird Dodo. They offered a 1 Gb plan at a good price and the hardware we got worked, in spite of their advice to the contrary.

The HUAWEI E220 HSDPA USB MODEM proved completely satisfactory, and talked to the eeePc running Xandros once we put in the right silly conection point fields; connect, blank, blank. It is even possible to update this web page if the -p switch is used after the ftp command. Thunderbird refuses to send email still.

So now we have a backpackable internet facility - less than 1Kg!

And TV too....

Backpack TV is also there on the eeePc with the "My Cinema-U3000 Mini USB receiver". It does not say so on the box, but it works in Linux.

See how in kalitv.txt in the download on this page.


This 6 tonne Mitsubishi NC lathe had stopped working shortly after being purchased after the Ford factory shutdown. Its 27 year old computer arrived here with desperate entreaties to fix it. The 17 big chips with taped over windows looked really suspicious and the assumption was that they had lost their program.

After changing IO boards and poking around James demanded to see the thing malfunctioning and after checking wires back through the relay panel and reading the instruction book, an extra wire was discovered going over to the other tool turret that indicated lock.

Our suspected MOS problem turned out to be a hydraulic problem caused by some grot getting under a solenoid and leaving a bit of torque on a hydraulic motor so the turret could not settle.

Fake Batteries :-(

There in the scrap lead pile was a pallet of demonstration batteries from Battery Energy. Rainbow Power Company were always throwing out good things and I assumed these were old dry charged units they could not risk selling. So I bought vast quantities of acid and rinsed dust out of them with distilled water and attempted a charge. Nothing but gassing! After more failure someone pulled one apart and found the problem - no paste.

I thought they felt a bit light ...


Well it is a type of windpower ..... after making a new barrel for a "Rampone" clarinet (Milan ~ 1880) we got more ambitious. James has designs on a Serpent, and Kali has actually landed a Rackett - aka Wurst Faggott. It is a ridiculous instrument that proceeds to impossibly low notes in a series of woody farts. The main problem from an engineering point of view was that some of the 12 holes were unreachable and needed levers and pads. Some also needed elevating from the surface of the instrument on litle tubes.

As this computer (Slackware 10) at last is running ALSA you might get a multimedia experience next edition with both a picture and a recording of this "Sausage Bassoon".

iPOD :-(

For ages this iPOD hung around. First it was the wrong cable, then the wrong operating system, and then at last a net connected XP machine appeared and we tried to put some music on the iPod. After much irritation that worked, but the earphone was so microsopic and overpacked we did not see it. Grabbing one of the myriad of computer speakers around we connected the iPod but instead of music we got that brown smell ....

The computer speaker had the wrong plugpack with the wrong label in the wrong hole, and there was a nasty little direct bypass switch on the speaker to conduct the 12vac into the poor SOT output. Expensive. And the only thing you can do is rant about the idiocy of the audio plug "standards". What bright spark started using 3 mm signal plugs for modem ac supply ??


This week has seen the failure of two computers and an oscilloscope due to resistor failure. Carbon film resistors after a certain lifespan simply stop conducting with no external signs. It seems unrelated to temperature, but does seem to be a matter of voltage and possibly humidity.

Finding the culprit requires perseverance and reasoning, and often the failure has been slow and erratic, damaging other related components. It is infuriating seeing a $1000 bit of gear brought down by a component worth 1/10 cent.

The remedy is to use metal film resistors, but this is probably precluded by the "race to the bottom" situation in globalised electronics manufacturing. Also Flowtrack boycotts metal film resistors because the blue body makes the colour stripes unreadable. (why would anybody...!). The only solution seems to be to use the expensive metal oxide film resistors which are magnificent. They still work red hot ...but yes, the colour stripes become unreadable!


A recent wind repair was caused by a lift vehicle driver who did a great job of getting the turbine up but then sort of wanted to get it more than up and kept driving!

The back stay gave under the strain and the machine followed the pull vehicle at rapidly increasing speed until a final resting place in the paddock :-(


An essential part of remote area electronics is of course getting there, and the old workhorse I drive has over half a million Km on it. It is a petrol/LPG 1600cc Toyota: cheap to run and easy to repair. It had been working perfectly on gas but randomly going badly on petrol - nothing below half throttle. Everybody said "put a new carbie kit in it!" MISDIRECTION 1: The petrol pump had failed last year and an attempt to get home by gluing it had resulted in muck clogging up the final filter in the carburettor. MISDIRECTION 2: The vacuum line to the distributor was clogged with dirt. After disassembling the carburettor and cleaning it (twice) I at last worked out that it could be an air leak, not a fuel blockage - no vacuum means no petrol. The car could still run well on gas because the idle fuel is delivered by the gas system with regard to nothing except continued ignition. A loose manifold flange, and a dodgy pipe from the crank case ventilation looked the culprit. But it was not :-( Finally the car stopped idling on gas and I drew the right conclusion: .....valve seat recession. The inlet valves were not closing perfectly. The engine was not missing on a cylinder or two as you would expect, but there was the manifold leak - back from the combustion chamber! We all tell stories of how ignition faults can mimick fuel problems, but here was a valve problem mimicking a fuel problem.


There was an open circuit on the Malapiki Microgrid. Nice solid 50 sqmm PVC coated twin set off up the hill with 100 volts DC on it, but the other side of the sea of lantana there was no voltage :-(

After crashing down the hill through the lantana it became apparent that the crossing of Tuntable Creek had been pulled down near the ground by the clambering bastard weed. That should not have caused an open, so the scratchy exploration continued. there was the fault: - it was about four metres across and a metre tall - a Bush Turkey nest!

The day was about 40 degrees, and the Turkeys nest was made from damp hoop pine needles and was so hot as to be a danger to walk on. Yes, the voltage was disappearing under the nest and after some excavation it was obvious that the cable insulation had melted and the metal disappeared into the compost.

So, Turkeys are not only able to crap, scratch, knock bottles off shelves , honk, "fly" around inside houses, upturn garbage bins ....... they can melt cables!

Another remarkable shot from left field

..... the serviceman

Flowtrack: Back to Main Page

ABN 68 079 207 168
0266 891431 (a/h)
0266 890408 (b/h)
OFFICE:The Brown House,roads end, Upper Tuntable
Falls Road, NIMBIN