A simple solution is to just use a cheap multimeter and a stainless steel bolt.
I'm actually selling a book about homemade shunts so I don't want to give away all the good stuff ( http://www.PoorMansGuides.com ) but here is a pic to help you get started with a 800 amp  1mv/amp shunt.  So, set your meter to mv and hook to the shunt.  And 800 amps would read 800 milivolts.

Spelljammer

One more thing, if your requirement is 0.1 ohm shunt that can handle 20 amps, my shunt calculator says that you can use 63 feet of 12 gauge copper insulated wire as a shunt.

Spelljammer

http://jaycar.com.au

product codes
QP-5410
QP-5412
QP-5414
QP-5416

Yes, they do international mail order.

No, I'm not affiliated with them in any way whatsoever.
Just a regular customer.

Thanks for a prompt response.

Flux, yes I am floundering a little. I have heard of 'shunts' over the years, but never studied them in earnest.

Please visit the attached two web pages and in particular the second which contains a circuit diagram. The author places a 0.1 Ohm shunt on the negative side of the DC regulated output from a wind turbine. I am trying to replicate the entire arrangement.

http://www.thebackshed.com/Windmill/PicLog.asp

http://www.thebackshed.com/Windmill/images/piclog/PicLog-circuit-2.gif

Further comments would be of great interest.

Even at 58 years of age I still enjoy learning!

Measuring low resistances is quite hard (e.g. your .01 ohms that you measured; if you measured it with an ordinary DMM, I have strong doubts about the results).

But measuring milli-volts is easy. The trick is to take a piece of wire, put a known current through it (say, 4A, from a car headlight bulb?), measure the current and measure the voltage drop over the shunt-to-be. The resistance can then easily be calculated with Ohm's law (R=U/I). The piece of wire could be shortened or lengthened, depending on what resistance you want.

Myself, I built a 'deluxe' version of the above thing: a milli-ohm meter that puts a known current through a wire (1A) and measures the voltage drop. Readout is directly in milli-ohms. With the help of a valve voltmeter, it can read down to 1 micro-ohm...

http://www.anotherpower.com/gallery/dinges/milli_ohm_coil_under_test

http://www.anotherpower.com/gallery/dinges/milli_ohm_meter_schematic_resized

But the trick with measuring current and amps is basically the same and quicker to set up for a single test.

Do have certain characteristics and these are the thermal if the Shunt is going to be used for current measurements, as well as, low resistance value to minimize the dissipation and with an accurate resistance low value to guarantee current precision readings with an analog meter

The current Shunt resistors have an alloy construction that have the temperature coefficient close to Zero ppm.

Higher the current higher the dissipation and for such reason you may see them built with blades separated between themselves and mounted in a massive metallic block.

In the case of the Picaxe project, possibly the best solution is to use on of the Allegro Hall Effect current detectors that have by definition the elimination of the temperature behavior plus amplification for further easier processing.

I use the ACS family up to 150 amps capability.

Visit www.allegromicro.com and look for current sensors

I also use devices from :

Raztec (NZ) Ltd.; PO Box 30075; Christchurch; NEW ZEALAND

tel +64 (0)3 982 3490 mobile +64 (0)21 374 590 fax +64 (0) 3 982 3496

eric.ball@raztec.co.nz

Visit www.raztec.co.nz

Raztec has some models that can be use for wide current range and by wrapping turns on the Hall effect a more sensitive arrangement is formed.

SHUNT resistors were invented in long past times when the electronics was in its infancy and we did not have the means to read the current in a sensible way, but with the available electronic current sensors the job is easier and with much less power dissipation.

The OTHER Shunt is just a misnomer used to imply a low value load across the supply for fast heavy current discharge and this shunt is most of the time a resistive Nichrome alloy with some stable temperature resistance value with much higher resistance ranges in the high ohms.

However... The response of a rusty nail is probably not quite so linear - especially with heat variation (ambient, load etc). This setup was suprisingly linear over the 0-20A range of my meter and I scaled results from there.

You can use anything conductive as a shunt. One of your main considerations is the reliability/stability of the connections.

Using a shunt with the picaxe is a bit of a trade off with out going to some lengths to amplify the signal. The picaxe requires approx 0-5V swing on the analog input to achieve a full 10 bit resolution (1024 steps). A 5V drop across a shunt represents a significant loss in a 12V system! You can get around this using an opamp to boost the voltage from the shunt.

Your 0.1Ohm shunt will give you a 2V drop at your 20A rated load. This will give you a resolution of approx 0.05A That is the picaxe will be able to measure current in 0.05A increments. However if this is a 12V system at your rated 20A the shunt will be costing you close to 17% of your output! You can basically use any value shunt and adjust the picaxe code to compensate (i.e a calibration process) If your shunt is a higher resistance then you will get a higher resloution at the expense of a greater resistive loss and viceversa.