How to measure Milli-Ohms

[rkosiorek]

lately i am finding it more necessary to measure milliOhm values to check motor windings and controller shunt resistors. so i have decided to build an adapter to allow me to use my regular Multimeter to measure and resolve 0.001 ohm.

the oldest form of ohmmeter would put a known Current through the unknown resistance and measure the voltage dropped by that resistance. Ohms law dictated the resistor value as R=E/I

now if i use a 1A constant current source and i measure the voltage drop the resistance would be R=E/1A or R=E. As long as my meter can resolve 1 milliVolt i can measure 1milliOhm.

one of the common IC Regulators can be used as a precision current source. the regulator works by adjusting itself so that there is always a 1.25V difference between the output and a reference pin. if i put a 1.25Ohm resistor there the current would be 1.25V/1.25R = 1A. but then my meter would only be as accurate as my resistors. which could be not accurate at all.

so i made my resistor variable so i could calibrate the value. that is why i have a resistor network making up the 1.25R resistor. i picked 1.3R because it was the next size larger than 1.25R. by calculation i established that i would need a 32R resistor in parallel with the 1.3r to make 1.25R. ths closest variable resistor would be 50R. i also wanted to make sure that i could not short out the output to the reference pin and toast the regulator. so i put a 7.5R in series with the variable resistor. i picked 7.5R becasue if i adjusted the 50R variable to the center it would be 25R add that to the 7.5R and i would have the 32R that i would need for the total resistance to add up to 1.25R.

i also added the 2R at R3 for the same reason. i did not want to stress anything by preventing a dead short.

i got a 5V 1A power supply for $2.00 from the Salvation Army thrift shop. the other parts i got from my junk drawer except for the 50R potentiometer. i had to buy that and as long as i was buying that i bought a plastic box and some binding posts to put it in.

Here is the schematic:



so to use it you plug in the wall adapter, and hook up the meter. switch it to read amps. calibrate the meter by adjusting the variable resistor until the meter just reads 1A. now switch the meter to it's lowest DC Voltage range, usually 200mV.

measure your unknown resistor by hooking it up to TP1 and TP2. read your value in milliOhms directly off of the meter. it will be as acurate as your meter is.

rick

 

[safe]

I "splurged" and bought the $30 multimeter that reads ohms down to that level.

 

[rkosiorek]

Please what make and model of meter is that? i could not find a meter that could resolve better than 0.1ohm (100mR) for less than about $250.00. here we are talking about 0.001ohm(1mR).

if i know the make and model i would also "Splurge" the $30.00

rick

 

[Ypedal]

Ya know.. one of these days... i'll understand all this... i will.. :x

 

[Mathurin]

+1 to "Where can I get that multimeter/milliohm meter for 30$"

Also here's another adaptor to work milliohms out of a multimeter:
http://users.tpg.com.au/pschamb/lom.html

 

[rkosiorek]

basically the same idea. even uses the same IC Regulator. i just chose to use 1A instead of 100mA as the measuring current for an extra digit of accuracy. i also used a variable resistor so i could zero out the meter. by using 1A the digit behind the decimal point becomes reliable.

you could also add a calibration to that meter by replacing the 300R resistor with a 500R pot in series with a 75R resistor.

rick

 

[fechter]

If you want to measure something like the shunt in a controller, you will probably need an even higher current, like 10A to make the measurement since the resistance is around 1 milliohm.

There's also no reason you have to use a multiple of 10 other than it makes the math easier. You could use 5.7A, for example, and use a calculator to find the resistance.

 

[safe]

My mistake.... it only goes down to 0.1 Ohms. It's got a setting for the opposite extreme "M Ohms" which would be bigger than "K Ohms". There's a big difference between "M" and "m".

I hadn't used it for that yet and had only used the voltage measurement which does go down to mV.

 

[Miles]

I found this one, in UK: http://www.farnell.com/datasheets/63069.pdf

I'll try to find out the price...........

edit: £215-00

 

[rkosiorek]

If you want to measure something like the shunt in a controller, you will probably need an even higher current, like 10A to make the measurement since the resistance is around 1 milliohm.

There's also no reason you have to use a multiple of 10 other than it makes the math easier. You could use 5.7A, for example, and use a calculator to find the resistance.

i wanted it to be direct reading with at most shifting a decimal point. no math required that is why i chose the decimal value. the circuit works well as it is. i built one over the weekend. it works on the motors perfectly. it also gives believable values8 for the shunts in the controllers that i have. but it really does need that extra digit to nail it down.

now i really don't want to go up to 10Aso i was thinking of using a low drift opamp with a gain of 10 and a trimmer to cancel out the input offset. like maybe a TLV272?

what do you think?

rick

 

[Link]

I believe my old Radio Shack 44-range (I think) one went down into the mili-ohm range. Shame I fried it before it really saw much use. Checked capacitance and tested transistors, too. Cost me like $70.

 

[fechter]

now i really don't want to go up to 10Aso i was thinking of using a low drift opamp with a gain of 10 and a trimmer to cancel out the input offset. like maybe a TLV272?

what do you think?

rick

That should work fine. I think that's what the expensive ones do.
That would be much cheaper than a voltmeter that's accurate to 0.01mv.

 

[Miles]

http://www.bellnw.com/products/0570/

 

[rkosiorek]

KEWL TOY for $325 plus shipping. i'm trying to do it for a lot less and use the meter i have.

i have a few meters and even my most expensive FLUKE ($120 or so) only goes down to 0.1ohm. most meters do not read below this because most service work would not require it. most multimeters only have what the manufacturer thinks the customers will use.

I'm not 100% sure but i think that the old Radio Shack 44 range FET-TVM analog meter only went down to a 10 ohm range but with the logarithmic scale it would be rough to read anything below 0.1ohm. that was a really nice meter. i wish i still had it. sensitive sucker.

RS for all of its detractors and cheap crap that the more commonly sold also had some real gems hidden in there.i worked at one of those stores part time while in high school (1972 thru 1975) In those days the majority of the store was parts and tubes and stuff. i bought that 44 range meter for myself while i was there.

rick

 

[Miles]

How about this beauty?: http://www.valuetronics.com/vt/assets/pdfs/HP_4328A.PDF (takes a while to load..)

You can find them cheaply secondhand - minus the probes.............

 

[rkosiorek]

thanks Miles but that HP meter will only measure shunts but not motors. it works on a different principal. it uses a 1kHz sqare wave as the constant current source and since our motor makes a lovely big inductor it will totally block the signal and act like a very large resistor.

those HP meters are great for checking switch and socket contact resistances though. but i guess so far i'm still stuck with rolling my own.

rick

 

[Miles]

Thanks for the explanation, rick.