Author Topic: Power supply effecting milliohm meter readings  (Read 1754 times)

0 Members and 2 Guests are viewing this topic.

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Power supply effecting milliohm meter readings
« on: October 28, 2019, 01:01:10 am »
Hello everyone! I have designed and built a milliohm meter which is capable of better than 10micro ohm accuracy (schematic attached)! I ordered a couple of precision resistors in order to test the stability and accuracy of the meter.


The basic principle is to generate a known current, using a 1.024V voltage reference and and shunt resistor that can vary between 1Ω to 100KΩ, then using the 4 wire measurement method, getting the resistance values. That basically means generating something between 10uA to 1A depending on the test resistor to maximize the voltage fed to the ADC.

I built a prototype and then connected it to a generic 12V switch-mode adapter and it seemed to work really well. here is the reading that I got from the 1mΩ resistor:


that's well within the tolerance of the uncalibrated unit + tolerance of the 1mΩ resistor. However I had designed the unit to work with a 9V battery, so I tried that next. Here is the reading I got from the same resistor, just using a 9V Zinc Chloride battery:


That's really odd! It's reading a much lower value which is way out of the tolerance range! so I thought maybe it's just an old battery and bought a new one! here is the result using the new Zinc Chloride battery:


not as bad as the first one, but this time it's reading higher than the actual value and it's again, outside the tolerance range. That made me think that there's just something weird going on with Zinc Chloride batteries specifically, so to prove that, this time I made a battery pack using AA Zinc cells:


nope! A well behaved, stable and within tolerance reading, so it's probably not the chemistry that's making the other 2 batteries not work, maybe it's just the voltage since the battery pack added to 12V, not 9(theoretically anything above 6V input should be fine and not affect the circuit, but that was the only thing I had to go by). I wanted to investigate further so I ordered 2 rechargable 9V batteries; One Li-ion and one NiMh. Here is the reading with the NiMh battery:


This is really odd now! The rechargeable batteries have the lowest voltage (8.4 compared to 9) yet the reading seems perfectly fine! Next was the Li-ion:


Again, another perfectly fine reading  :palm:
This means I have good readings with a 12V and very noisy switch mode power supply, a 12V battery pack made out of AA Zinc cells and any type of rechargeable 9V battery, but not with the generic zinc 9V ones.

Chemistry of the battery doesn't seem to be an issue, voltage doesn't seem to be an issue (I plan on testing that further with a lab PSU and varying the voltage from 7 to 30 which is the tolerable input voltage for the device) and it's not even consistently higher or lower with the same battery type! can anyone explains this and possible fixes for the design if I've missed something? I really can't explain how this happens with certain batteries and not the others!
« Last Edit: October 28, 2019, 01:06:50 am by OM222O »
 

Online ejeffrey

  • Super Contributor
  • ***
  • Posts: 4033
  • Country: us
Re: Power supply effecting milliohm meter readings
« Reply #1 on: October 28, 2019, 01:31:11 am »
Without looking at the schematic, my guess would be that you are overloading the batteries.  Drawing 1A from a standard 9V battery is marginal.  Check the voltage rails (battery, 5V and reference) with a dmm while measuring the 1mohm resistor.
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: Power supply effecting milliohm meter readings
« Reply #2 on: October 28, 2019, 02:18:59 am »
overloading the battery can explain the low reading of the first battery, since it's not supplying enough current, but not the high reading of the second one  :-// like is it somehow pushing more current through the resistor, despite all of the regulations? that doesn't make a lot of sense to me. I will check the rails with a scope as soon as I can.
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15797
  • Country: fr
Re: Power supply effecting milliohm meter readings
« Reply #3 on: October 28, 2019, 02:37:10 am »
Looking at all power rails will make you see what happens. Anything else will just be guesses.

But generally speaking, drawing as much as 1A from a 9V battery? I have some doubts. Maybe with rechargeable (NiMh based) ones...
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Power supply effecting milliohm meter readings
« Reply #4 on: October 28, 2019, 02:53:02 am »
When running the test, try scoping the power supply pins to see if there is a voltage drop during measurement.
 

Online ejeffrey

  • Super Contributor
  • ***
  • Posts: 4033
  • Country: us
Re: Power supply effecting milliohm meter readings
« Reply #5 on: October 28, 2019, 03:10:31 am »
overloading the battery can explain the low reading of the first battery, since it's not supplying enough current, but not the high reading of the second one  :-// like is it somehow pushing more current through the resistor, despite all of the regulations? that doesn't make a lot of sense to me. I will check the rails with a scope as soon as I can.

Circuits operating outside their design conditions can do unpredictable things. The supply might oscillate, your amplifier or ADC might be misbehaving.  It's not really worth speculating about, just measure the rails.  A scope would be best, but my guess is that a $2 DMM will be enough to show you where the problem is.
 

Offline MosherIV

  • Super Contributor
  • ***
  • Posts: 1530
  • Country: gb
Re: Power supply effecting milliohm meter readings
« Reply #6 on: October 28, 2019, 07:47:09 am »
Hi, as others have said, the 9V battery will strugle to supply 1A! I doubt it can supply 100mA for any decent length of time. The 9V batteries are really only able to supply 10 maybe 20mA max. In most dmms they typicaly supply 5-15mA, less if you want it to last a decent amount of time.
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9321
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: Power supply effecting milliohm meter readings
« Reply #7 on: October 28, 2019, 01:07:11 pm »
In my experience, milliohm meters usually use bursts of current, in order to reduce power dissipation in the unit, heating of the DUT, and allow offset compensation.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline OM222OTopic starter

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: Power supply effecting milliohm meter readings
« Reply #8 on: October 28, 2019, 05:00:27 pm »
In my experience, milliohm meters usually use bursts of current, in order to reduce power dissipation in the unit, heating of the DUT, and allow offset compensation.

I only apply the 1A for a maximum of about 0.5s before allowing significant cool down time and since the ADC is quite capable, I can compensate for the seebeck effect of the DUT, so let's not go into the design aspects of the meter for now. I'm just trying to figure out why the different power sources can affect the readings in such significant ways.

Hi, as others have said, the 9V battery will strugle to supply 1A! I doubt it can supply 100mA for any decent length of time. The 9V batteries are really only able to supply 10 maybe 20mA max. In most dmms they typicaly supply 5-15mA, less if you want it to last a decent amount of time.

due to the pulsed nature of the load, I'm not worried about the battery life at all. the power source only needs to be capable of delivering 1A peaks (which the rechargeable ones and even the AAs manage just fine). I tested the 9V Zinc batteries with a DMM and the short circuit current seems to be about 1.4A so I'm not sure what the issue is. maybe the voltage drops way too much when delivering that current. I will have access to a scope soon, so my best bet is checking the power lines like others mentioned.
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9333
  • Country: fi
Re: Power supply effecting milliohm meter readings
« Reply #9 on: October 28, 2019, 05:28:10 pm »
Testing and photographing all batteries in existence is a colossal waste of time and produces nothing else than guesswork.

The batteries won't communicate through magic. You only have one parameter to measure: the terminal voltage during operation.

Measure it.
 

Offline TimNJ

  • Super Contributor
  • ***
  • Posts: 1720
  • Country: us
Re: Power supply effecting milliohm meter readings
« Reply #10 on: October 28, 2019, 05:29:37 pm »
On average, you can expect "Heavy Duty" batteries to have 3 - 6 times higher series resistance compared to an alkaline battery of the same size.

http://www.g-holbeck.com/english/elteknik/indexri.htm
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf