DMM for measuring 1 > ohm at a resonable cost

[lk]

Greetings,

I'm on the lookout for a DMM that with a good decent accuracy can measure resistance at less than 1 ohm (down to 0.01 ohm maybe). I have been looking at a Vichy VC8145 bench DMM, but i am somewhat suspicious about the claims that it can measure in these ranges. link to the meter on eBay
http://www.ebay.com/itm/DMM-VICHY-VC8145-4-7-8-Digital-Bench-Top-Multimeter-/300459570747?pt=Photo_Frames_Display&hash=item45f4c9363b

I suspect that i wont get better readings with the Vichy compared to my Exetech 330.

If i had 600$ i would buy a used Agilent 34401A bench meter in a heartbeat, but my funds are limited.

For the time being i just measure voltage drop over the wire/resistor and with a known current then calculate the, but i would like to to the measurements directly.

Thinking a little about it, maybe i will just be more happy if i save up the money and get a proper instrument

-lk

[eevblogfan]

hey

question : what are you planning to meassure ? 

if it's fuses etc , build yourself some 1 A CC , and just meassure the voltage on the mV range , really , each 1mA will be 0.01 mohm , really doesn't need to spend that much  ,

with you're exetech 330 . you'll be able to see as low as 0.1 mohm . so do the math

as for the CC , open a theard (or  search for one ) about building 1A constant current ,

good luck  

[lk]

What am i going to measure, right this instance its a shunt resistor for a CC, but tomorrow, im sure its something totally different, maybe  pieces of wire. 

-lk

[EEVblog]

A good LCR meter will measure down to 1mohm resolution, with just two terminals.
e..g the Extech 380193
http://extech.com/instruments/resources/datasheets/380193.pdf

Just over $200:
Extech 380193 Passive Component LCR Meter

Plus you get valuable LC capability too of course.

Dave.

[robrenz]

Accuracies of most meters below 1 ohm are not very good.  You most certainly need to be doing 4 wire kelvin type measurement at this level.  Most 6.5 digit bench meters are only good to 0.3% at 1 ohm. The reason is that these measurements tend to be at the low end of the range and the =/-% of range part of the error starts to become very large.
Example:
Fluke 8846A 6.5 digit meter Calculated actual error values on 10 ohm range
10 Ohm  = 0.04%
1.0 Ohm  = 0.31%
0.1 Ohm   = 3.01%
0.01 Ohm  = 30.0%
0.001 Ohm = 300.0%

Your answer to inexpensive low ohms is a decent LCR meter.  A decent one will be 4 wire measurement and AC excitation so it eliminates thermal emf errors.

The IET DE-5000  will give you +/- 1% +3 counts on the 20 Ohm range at 1kHz  Calculated actual error values would  be:
10 Ohm  = 1.%
1.0 Ohm  = 1.3%
0.1 Ohm   = 4.0%
0.01 Ohm  = 31.0%
0.001 Ohm = 300.0%

The Agilent U1733C will give you +/- 0.7% +50 counts on the 2 Ohm range at 1kHz  Calculated actual error values would  be:
10 Ohm  = 0.75%
1.0 Ohm  = 1.2%
0.1 Ohm   = 5.7%
0.01 Ohm  = 50.7%
0.001 Ohm = 500.0%

Notice the higher resolution .0001 Ohm on the Agilent does not translate to higher accuracy at the low end.

[robrenz]

Sorry Dave, I was posting when you posted an ignored the warning.

[EEVblog]

Yes, the accuracy isn't too crash hot at these levels without proper 4 wire measurement, but it's usually a lot better than the spec limits would have you believe. If you are careful to zero out and do multiple measurements to make sure it's reasonably repeatable, then you can be reasonably confident in the absolute results at the low end of the range.
The 10mOhms minimum you mention should be doable on a decent LCR meter, but as rebrenz said, 4 wire is the right way to do it at these sorts of levels.
If you don't already have an LCR meter, then it would likely be a lot more useful than a bench DMM.

Dave.

[robrenz]

Same  examples on the Extech's

The Extech 380193 will give you +/- 1.2% +8 counts on the 20 Ohm range at 1 kHz but only measures to .020 Ohm.  Calculated actual error values would  be:
10 Ohm  = 1.2%
1.0 Ohm  = 2.0%
0.1 Ohm   = 9.2%
0.01 Ohm  = out of range but would be 81.2%
0.001 Ohm = out of range but would be 801.2%

The Extech LCR200 will give you +/- 1% +5 counts on the 20 Ohm range at 1 kHz Calculated actual error values would  be:
10 Ohm  = 1.0%
1.0 Ohm  = 1.5%
0.1 Ohm   = 6.0%
0.01 Ohm  = 51%
0.001 Ohm = 500%

[eevblogfan]

hey

Dave . from curiosity , why shouldn't he build 1A constant current and measure the mV drop on the resistor ?

really , I can't understand why should he spend 200 $ (LCR is good , but buy it special for resistance measurement  ? )

BTW, I'll mabye open a thearad on 0-1A 55V electronic load , I think 20W will do

have anice day !

[EEVblog]

The other thing I forgot to mention is that you can also use your existing current/voltage measurement technique to check and build up confidence in your LCR or other meters at reading those low values. And in most practical cases you'll find it's a fair bit better than the theoretical worse case absolute specs robrenz has mentioned.

Dave.

[robrenz]

Yes, the accuracy isn't too crash hot at these levels without proper 4 wire measurement, but it's usually a lot better than the spec limits would have you believe. If you are careful to zero out and do multiple measurements to make sure it's reasonably repeatable, then you can be reasonably confident in the absolute results at the low end of the range.
The 10mOhms minimum you mention should be doable on a decent LCR meter, but as rebrenz said, 4 wire is the right way to do it at these sorts of levels.
If you don't already have an LCR meter, then it would likely be a lot more useful than a bench DMM.

Dave.

I dont know about the Agilent or the Extechs but the DE-5000 is true 4 wire at the built in test clips and on the aligator clips and tweezers.  Also using the AC resistance will be more accurate at these levels because it cancels any thermal emf issues. But if your resistance is significantly inductive you will need to stick to DC resistance measurement.

[EEVblog]

Dave . from curiosity , why shouldn't he build 1A constant current and measure the mV drop on the resistor ?
really , I can't understand why should he spend 200 $ (LCR is good , but buy it special for resistance measurement  ? )

It's just a nicer and less messy way to do it, and gains you a valuable bit of kit.
he's asking for alternative to doing it manually like that.

Dave.

[EEVblog]

http://www.ebay.com/itm/DMM-VICHY-VC8145-4-7-8-Digital-Bench-Top-Multimeter-/300459570747?pt=Photo_Frames_Display&hash=item45f4c9363b

I think a $200 LCR meter (that has AC measurement capability) is a much better value for money and more useful at not much more cost.
Actually, this one is cheaper:
http://www.ebay.com.au/itm/Handheld-bridge-LCR-Meter-0-3-accuracy-0-1-resolu-/220356080266?pt=BI_Electrical_Equipment_Tools&hash=item334e3ef28a
and there is a thread on the forum about this meter.
https://www.eevblog.com/forum/product-reviews-photos-and-discussion/tonghui-th2822-lcr-meter-review-teardown-and-possible-hacking/msg82662/#msg82662

Dave.

[robrenz]

The other thing I forgot to mention is that you can also use your existing current/voltage measurement technique to check and build up confidence in your LCR or other meters at reading those low values. And in most practical cases you'll find it's a fair bit better than the theoretical worse case absolute specs robrenz has mentioned.

Dave.

I agree completely,  Your worst case error if you used a 87V to measure your current and voltage for the 1A through the .01 Ohm would be 1.51%   Also agree most quality meters will be much better than the worst case absolute specs. My main point is showing how important it is to actualy do the math and compare +/-% of value specs over the full range to accurately compare instrument specs.

 

[free_electron]

The proper way to do this is by using a real microohm meter...
How do they work ?
First of all : kelvin sensing. 2 wires deliver a reference current, 2 other wires sense the voltage. Simple enough principle.
The problem is that, with such small resistances, the current needs to be large to get a good reading ( if you don't want to sink waya in noise ..)
Large current means that you would heat up the device under test. This will falsify you reading... Most parts, especially wirewound resistors, drift largely under temperature.
So, how to solve this ?
The answer is a pulsed current source. They apply this test current for a few microseconds, sample the voltage and turn the current off. The conversion is done afterwards.
True microohmmeters use sense currents of up to half an ampere. In the advanced machines you can select the test current as well as the polarity of the test current and the sampling rate.

[lk]

Greetings Again,

I turns out i have a Cheap LCR meter, i forgot about it DM4070, it measures a 1% 0.22 resistor at 0.33 ohms, i guess this is why i forgot about it. If i feed 10 mA through the resistor i get ~22.3 mV drop across it, which if my math is correct is  0.223 Ohm.

What can i learn from this? i should properly not get cheap ass LCR meters

I just watched your blog on the extech LCR meter with the silly product number, but I'm still unsure if i should get that, or save up for a good used 4-wire bench multimeter.

-lk

[robrenz]

A good MicroOhm meter will also do offset ohms along with the pulsed measurement to eliminate thermal emf errors. It will take voltage measurement with no current to get the thermal emf voltage and then another voltage measurement during the current pulse. the meter then does the math to cancel the thermal emf error.  Anything highly inductive like a transformer will need straight DC offset ohms measurement to be accurate.  The IET LOM-510A is probably the most bang for the buck.

I did a thread about it here https://www.eevblog.com/forum/product-reviews-photos-and-discussion/ie/msg96311/#msg96311

[T4P]

Umm ... The Mastech MS5308 from precisegauge.
150$ and pretty darn accurate

[lk]

Dave S. I'm really impressed with the capabilities of the ms5308, i could see myself buying that, i still need to do some more, reading on it. The post on the the board from aurora, does make it look promising.

The 4 wire measurement capability makes it a very strong candidate, and the price seems hard to beat as well

-lk

[alm]

When measuring power components like shunts or wires, you should put in some extra thought into what heating means to your circuit. What's the point of measuring the cold resistance of a shunt that you will use to measure 10 A? Wouldn't it make more sense to measure the resistance it reaches when heated by those 10 A, i.e. put 10 A through it and measure the voltage drop? Current shunts are usually calibrated at several current levels for exactly this reason.

This may not fit with the textbook model of an ideal resistor, but will actually give you the information you need: how much voltage does my shunt drop at X amount of current?

[free_electron]

especially since the resistance will change. every metal has a tempco...

[lk]

Hi Alm and Mr. Electron,
I realize that i need to take heating into account, i will mainly be working with low currents, only exception, is Daves variable CC source thingie, where i have added a fan to cool the heat sink.
But yes i need to keep an eye on component temperature, but i still need a good way to measure those components with the low values, the mastek ms5308 seem like a good choice there.

-LK

[qno]

I had a milli ohm meter once that used a 1500 Hz sin. wave in a 4 wire setup and an AC milivolt meter.
It send  the 1500 Hz true the dut at low amplitude so it will not blow out the component.
The trick is that the 1500 AC signals are much easier to amplify without offset and filter to have little influence from 50/60 Hz.