Sunday 14 August 2011

Battlebots at Serial Space

Saturday 9 July 2011

Come watch real life combat robots duke it out at Serial Space, Sydney

Friday 8 July 2011

Replacing Halogen down lights with LED. Costs and photos

We just have 2 halogen down lights in our house, but I would like to use them more often that we do.

I bought a few and replaced them. They work well for lighting up the stairs. The light is more blue than the halogens but OK for here.

I tried some at a friends place with electronic transformers and they don't work. The transformers probably detect no load and shut down.

So, How much can I save?

The test voltage was 13.65v

LED Current 0.31A
Halogen Current 4.2A

At 240v with say 90% efficient transformer this equates to:

LED: 4.65W ~ 20mA @ 240V
Halogen: 63W ~ .263A @ 240V

At 30c/kWh this amounts to:

LED ~0.014 cents per hour
Halogen ~ 0.19 cents per hour

Difference is .176 cents per hour. At a cost of 9$, payback is in 5113 hours or 213 days.

This might not seem much, but the lighting energy is now 7% of what is was. I call this significant.


One of each:

After with LEDs.

Photos from Sydneys Traffic Management Center. Exclent Screens!

Tori made a nice woolly coat and hat for her toy dog

Coat and hat by Tori my little maker

Sunday 3 July 2011

eFXC TTXGP at Easter Creek Photos and Video

Linked below are the photos and videos of this weekends TTXGP / EFXC Electric super-bike round.

Flickr Set

eFXC / TTXGP Australia entrants #TTXGP

From left
Daniel Sailer - Ripperton Racing
Jon Eggenhuizen - Catavolt. Ridden by Jason Morris
Chris Jones

Friday 24 June 2011

Robot Wars come to Sydney

Support and entry details here.
Sunday 14/8/11

The location of Serial Space is
33 Wellington Street, Chippendale, Sydney, NSW 2008

Thursday 23 June 2011

Toy Soldiers

Toy Soldiers

Diplomats meet
And enter discourse
But can’t agree
On a Trojan Horse
While greedy capitalists
Flock like birds
And war with bombs
Instead of words
But do we know
The worlds for lease
And the contract expires
In hand with peace.

By Graham Boyles.

The Mitsubishi iMiev will be $50,000 AUD in August


Wednesday 22 June 2011

EV Efficiency and pollution

I wrote this piece in response to a post on the EEV Blog to clear up some points.


I wrote a blog post a while ago detailing CO2 output. In NSW inefficient EVs are on a par with petrol. Coal power stations are more efficient that you mention. (33%-53%)

I have not accounted in this equation for how much it costs $ or CO2 to extract oil, transport it, refine it, transport it again. Not have I accounted for extraction costs for coal, but I would bet coal is cheaper on both counts.

Nissan says it takes 7.5kWh just to refine a gallon of gasoline. see That doesn't count extraction and transport.

Or 2kWh per Litre. We can drive an additional 10km on that alone.

But enough on the CO2 and efficiency. There are other significant factors that come into play here, cost and environment aside.

We can produce our own electricity to power cars, and do not have to be dependant on foreign oil and the associated price vulnerabilities. The point of pollution is out of the cities. A Coal fired power station is more efficient than a petrol engine, and easier to scrub the output from one power station than 50,000 or whatever number of cars.

Mitsubishi iMiev review with Dave Jones from EEV Blog and myself.

EEV Blog Episode #179

Tuesday 21 June 2011

Vectrix Electric Motorcycle Battery Repair

I was contacted to see if I could assist with a battery fault in a Vectrix motorcycle.
A Vectrix owner has a bike that turned out to have a faulty cell causing it to have poor performance and stopping after just a few kilometres.

I arranged a visit the bike and determine the fault.

After much thought trying to figure out how to remove the covers and expose the battery pack, we finally opened the enclosure with 102 cells arrange in two blocks. The rear block is 9 cells long, 2 wide and 3 cells deep. The front block is 8 long, 2 wide and 3 cells deep. Each set of 8 / 9 is bound by a steel compression band to prevent the cells from expanding when heated during charging and discharging. Each layer comprises of two rows of 8 / 9 in a plastic frame. Each of the 3 layers is held together by long threaded rods extending between the rows and also by duct tape around the outside of the plastic frames.

The top of the rear block was removed and some evidence of heating and corrosion was visible on one of the cells.

Leaky cell and voltage - temperature sensors

The cell had overheated, the plastic frame the cell sits in was melted, and the pressure relief valve on the top of the cell had released causing some corrosion and crystallisation of electrolyte around the top of the cell. We verified the cell had failed electrically also. It measured only 32mV.

The others were consistently about 1.31V across them, we surmised that there was only one dead cell in the pack as the other strings of cells had a near perfect multiple of 1.31V across them.

The cells in the Vectrix are 30Ah Nickel Metal Hydride cells. From GP Battery. Replacements are available from the distributor in Melbourne. Cell data sheet:

Good Cell. Some corrosion evident on terminal. This seems quite common. Cell is not swollen nor hear damaged.

Bad Cell. Corroded where electrolyte and gas had escaped through relief valve, and also swollen and heat damaged.

I later returned to the bike to disassemble the entire pack and measure each cell voltage and check for evidence of more failing or faulty cells.

The first thing to do is to disconnect the large blue Anderson connector on top of the batteries. This breaks the battery pack up into a safer (still potentially lethal) lower voltage blocks.

The temperature and voltage sense boards were also disconnected at the connectors at the front and rear of the bike.

There are temperature and voltage sensing elements screwed to some of the battery terminals. These were numbered with a marker on each sense board and also on the cells they were removed from to ensure they were replaced in the correct locations. Each sense board was wrapped in insulation tape to prevent them accidently touching the battery terminals once removed.

All the other cells were of the correct voltage (~1.3V) and had no physical signs of failure. Disassembling the blocks was fiddley. The cells are assembled in plastic frames that are then joined with steel compression bands to prevent cell swelling.

One cell had over heated and the 2 adjacent cells had suffered heat damage. The plastic holders were also melted. The pack was left disassembled as 3 new cells, and plastic frame pieces were ordered from the Australian agent.

It is important that the new cells introduced to the pack are at the same charge state as the old cells. I used a digital Radio Controlled model charger to charge the 3 replacement cells. The charger also has a discharge function so I was able to discharge the cells down to the same voltage as the existing good cells. The charger can charge at 5A and discharge at 1A. The cell chemistry, charge current, end point sensitivity etc can all be programmed in the charger. NiMH cells self discharge quite quickly. A few 10s of milivolts each day. I found that after discharging the new cells to the required level, the new cells had to be discharged further to match. I got the final voltage of the new cells to within about 10mV of the existing cells. The first balance charge on the bike would balance them further. I proceed to replace the 3 cells in a string of 9. Each ½ layer of cells is bound in steel a binding strap to prevent the cells swelling during charging and discharging.
I have made a jig for compressing the packs in order to remove and replace the steel bands more easily. To make things more tricky, the brittle locating tabs

I proceed to replace the 3 cells in a string of 9. Each ½ layer of cells is bound in steel a binding strap to prevent the cells swelling during charging and discharging. I have made a jig for compressing the packs in order to remove and replace the steel bands more easily.

An interesting point is that each cell is electrically insulated from the next by plastic spacers.

Reassembling the pack consisted of reversing the disassembly process. Extra care was taken to double check visually and with a volt meter that each 9 / 8 cell bank was reassembled correctly and the temperature and battery voltage monitoring sensors were installed in the correct locations.

All interconnects and battery terminals that had been removed were tightened to a specified to torque of 10Nm.

After reassembly the owner took the bike for a ride and came back with a memorable grin. The bike performed better than it had for a long time.

The final task for the owner is to deep discharge the pack 5 times to condition the batteries and ensure longevity of the new batteries. This procedure is recommended by Vectrix to remove any ‘memory’ effect in the batteries in order that they can provide the maximum usable capacity.

Monday 2 May 2011


Not that I condone grafiti at all however I quite liked this stylised face drawn on the bus seat thismorning.

Wednesday 6 April 2011

The ute has gone to a new home,

I finally sold the electrolux. It was picked up to be take to Perth a couple of weeks ago.

Here it is leaving my place for the last time.

It is destined to have new life breathed into it with a new VFD and some 30Ah Vectrix electric motorcycle batteries.

My 'new' BMW E30 318i may become my new project :)

Monday 14 March 2011

BMW E30 1st Electric Conversion Post

Well I finally bought an E30  BMW 318i Coupe, with a manual gearbox.

Great fun to drive. It is a bit bogan, so like the ute, I will remove the tinting and put a chrome grill insert back into it.

If it does become an EV, then it will most certainly need to have some respectable times at the drags!

E30 Specifications

Here are some pictures to get started.

Wednesday 23 February 2011

555 Contest Entry

I was excited when the 555 contest was announced to recognize the most used Integrated Circuit ever.

After considering for a while what I could do for the competition, I started thinking about how best to celebrate the 40 years that the device has been around and the 10s of billions produced.

I have often admired home built computers that use discrete components so I though, I will make a 555!

This 555 project recreates an operational 555 timer using discrete transistors, resistors, diodes and capacitors.

It may not strictly conform to the competition rules, but I thought it would be fun and original anyway.

Initial Research.
I downloaded a few 555 datasheets from various manufactures only to find that the circuits often have subtle differences. Eventually after some Googling I located a scanned copy of the original Signetics datasheet from dapj Circuits, Including the Equivalent Circuit below.

After inspecting the circuit, I noticed Q19 is rather unusual and has 2 collectors. How can I implement this using discrete components? I decided to simulate the circuit using the brilliant and free Java app from to make sure it worked and to quickly experiment with different arrangements for Q19.

This is the simulation. You can run the simulation yourself by clicking the link .

Arranging the resistor on the base of Q19 as in the simulation works perfectly. If I used 2 transistors to replicate the original Q19 design, the circuit did not oscillate.

In order to simply construct the circuit to show its operation and at the same time preserve the original authenticity, I photocopied and enlarged the Philips schematic, glued it to some card and inserted the components through the card and used point to point wiring on the rear of the card.

Component side of card.

Rear of card using soldered point to point wiring. Drawing pins are used to secure the input and output pins. Much ice cream was consumed during the construction of this project.

A pin was used to poke holes in the card to insert the components. This was a surprisingly quick method of building the circuit. For prototypes of simple circuits I may use this technique again.

the only thing I might do differently is to paste a reverse image of the circuit onto the wiring side of the card in order to avoid wiring errors.

In order to test it works, I attached an LED to the output and suitable timing components, Ra, Rb and C to give an approximate 1Hz flash rate.

the final testing can be seen in the this video.

Making it presentable
The final thing to do is to make the project presentable. I decided the inside of an old data book would be appropriate. I found a copy of a suitable book on the 'free' table of my local ham radio club, the Manly Warringah Radio Society.

This was the first time I have used a book as an enclosure. It works well if you are not in a hurry, which I wasn't. I followed the guide by Bre Pettis from Make Magazine. (Cool guy who I had the pleasure of meeting at a Maker Faire in Austin, Tx.)

One half of the book was cut to accommodate the circuit, while the other half was to house a battery, the 555 timer data sheet and memorabilia.

Right hand side of book contains the circuit

The left hand side of the book contains A5 printed copies of the Philips data sheet, The 555 Timer Wikipedia page and finally an interview with the 555 inventor Hans R. Camenzind.

The finished project closes nicely into the book for safe keeping.

Finished project in operation

Friday 11 February 2011

WWF - Living Planet Report 2010

WWF - Living Planet Report 2010

The report, a leading statement of the planet's health, found that people are spending beyond the Earth's ecological means and dipping far into the Earth's natural capital. The ecological deficit caused by consumption of energy, water and materials at rates 50 per cent beyond supply will have serious repercussions for wildlife and ecosystems, as well as for future generations of people, all of whom depend on nature's ecological services.

Australia is #8 and New Zealand has the 32nd most heaviest footprint.