Milliohm meter

[Wizpic]

I'm working on a milliohm meter project, I got the idea from here (Kerry Wong)
http://www.kerrywong.com/2011/08/14/accurate-milliohm-measurement/

I'm using an arduino,ADS1115(ADC) with the added feature of auto zeroing by pressing a button all this is working very well and accurate, I've got 23meters of 35MM cable and the manufacture states if good should read 12.65 milliohms and it reads 12-13milliohms on my meter and exactly the same as a purpose brought meter £300.00, Once I'm happy I will post the project here  in case any one wants to have ago or needs one,
At the moment it's running off 2 X 1.5V AA batteries using an LT1173 set up to boost to 5V the whole unit including the LCD draw's about 45Ma (without the backlight). If this goes well I plan to make about 30 of them ,But now I want to add some sort of auto power off to get the most life out of the batteries in case it gets left turned on. The idea is to have a push button to start the test,turn on take the reading and display it then say turn off after 30 seconds, But also to press another button(if needed) to set zero if extra length of wire is required plus remove the offset voltage.  I've had a browse a round on the web found plenty of ideas (sort of) so this is where you guys could help to find which would be the best method or how would you do it ??, The way I look at it you can do it one way and someone else can give a better method.
I have thought of putting it to sleep but there is still the other parts of the circuit that would be powered up which I don't want.
I've attached photo of the proto-type (plan to draw/layout pcb once design is finished)

[Checksum8]

This is a circuit I found a while back and saved. I don't remember the source
but it looked worth while.

[lou]

Is that 8 miliohms ?  Or .000008 ohms ?

[optikon]

Simple display units are incorrect. The Caddock resistor is 0.008 Ohms so his display should read "8 MilliOhms."

Hey Wizpic, do you have a source / link to those 4-wire clip probes you are using?

[Wizpic]

Is that 8 miliohms ?  Or .000008 ohms ?

It's 8 Milliohms, I now display open circuit/Ohms/milliohms on the LCD.

Checksum8 that looks good and simple enough will have a play with it over the coming weekend.

optikon
I got the probes from here
http://uk.rs-online.com/web/p/kelvin-clips/7148263/
I did look at these but a bit costly if I'm going to make 30 of them
http://uk.rs-online.com/web/p/kelvin-clips/3353240/

[dirtydigits]

what kind of circuit needs small precise resistances like this. I'm guessing high voltage. Most of what I'm into is audio related but this does seem interesting.

[Wizpic]

what kind of circuit needs small precise resistances like this. I'm guessing high voltage. Most of what I'm into is audio related but this does seem interesting.

It's for checking restistance of Earth cables, if the cable starts to rub through the outer sheaving and the cores start to break the resistance of the cable will increase as the cable is a set length

[crunx]

I have worked in a couple of projects with milliohm-class resistivity measurements, and learned about another problem associated with measurement of the very low voltages involved: Thermoelectric voltages, which occur at junctions of dissimilar conductive materials (e.g. different metals and platings). That kind of voltages depend on material and temperature differences, and are generally difficult to manage in DC measurement systems, measuring micro-volts.

Therefore the professional systems usually use AC, for example 1kHz, as the DC error introduced by the thermoelectrics can be eliminated.

If you want to use a DC measurement it is very important to use two equal 4-wire connections/probes so that the thermoelectric is in better control, and also be aware of importance of the temperature differences between the two measurement points and other parts of wiring to the micro-volt-reading amplifier. For demonstration or test of the phenomena, just warm up one measurement point or wire joint to higher temperature than the other, and the reading will likely change.

[Wizpic]

The original one (purchased) unit produces a DC 5 volts at 5ma signal from a 9v battery using the good old icl7106 chip to display the reading costing £275 plus vat,One of the cables is 50mm 23 meters long and manufacturers say it should read 13milliohms which the bought unit does and the one I've built displays the exact same reading pleased with it so far still in design stages and testing. Always looking to make it better also considered to reverse engineering the bought one just for the crack  

[crunx]

There seems to be other similar projects to learn from. One quite interesting construction is in

http://www.pittnerovi.com/jiri/hobby/electronics/milliohmmeter/index.html

All good milli/micro-ohmmeters use 4-wire techniques: One pair supplies the excitation current, and another pair measures the resulting voltage. Even with a four-wire arrangement there are several disturbances. One is the thermoelectric phenomena I mentioned earlier, which is a nuisance on the DC method. Also measuring amplifier offset voltage has a similar effect, but it can be usually nulled out.

Another type of disturbances is the induced noise, from mains AC and other electric environment. That bothers also/especially the AC measurement systems. However, by having the excitation current and measurement in synchronized, the noise can be cancelled pretty well: The synchronous measurement is essentially averaging out everything, which is not exactly on the measurement frequency (or some unlucky harmonics of it)

[Gallymimu]

It's for checking restistance of Earth cables, if the cable starts to rub through the outer sheaving and the cores start to break the resistance of the cable will increase as the cable is a set length

I use small miliohm measurements for a few things:  Rdson for power MOSFETS, current sense resistors in switching power supplies.  High current connectors.

Generally just run a big current and measure the voltage drop...

[Wizpic]

I'm still working on this project when time allows. I've also been looking through the links given here for ideas. I know the best method is 4 wire method but what I'm trying to figure out is that it uses common ground
. The other thing I noticed when it runs from 2 1.5v batteries boosted to 5v the unit reads accurate comparing the data given by manufacturers data but when it runs from the Usb it reads a lot higher unless the power pack is connected which is strange which I'm looking into it, it draws the same current 10ma on both supplies
I never thought about using it for mosfets did think about using it for making shunts and calculations

[Gallymimu]

I'm still working on this project when time allows. I've also been looking through the links given here for ideas. I know the best method is 4 wire method but what I'm trying to figure out is that it uses common ground
. The other thing I noticed when it runs from 2 1.5v batteries boosted to 5v the unit reads accurate comparing the data given by manufacturers data but when it runs from the Usb it reads a lot higher unless the power pack is connected which is strange which I'm looking into it, it draws the same current 10ma on both supplies
I never thought about using it for mosfets did think about using it for making shunts and calculations

USB power causes problems with a lot of precision instruments.  Remeber USB power is subject to ground loops through back through the computer.  Also the 5V on usb is SUPER SUPER noisy.