Measuring uAmps (without a good meter)

Go To Last Post
23 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have a micro which I want to measure the 'deep sleep' current, but my meters are not up to the job. I have a Fluke 115 (I think that's the number), but its not a good meter for this. I also have some cheap meters that have a 4ma range, but do not seem to be very accurate.

So I take a 100K resistor, put it in-line with the battery (lion- ~4.1v when charged), and also put a jumper across the resistor. I power up the micro, put it in deep sleep, remove the jumper, and measure the voltage across the resistor. I show 0.84v volts across the resistor. I then put the jumper back on, wake the micro to prove it was actually in deep sleep, and not powered down. (Power goes to a mcp1700 3.3 regulator, so I'm measuring that, too).

If I calculate correctly, an 0.84v drop across a 100k resistor means there is 8.4uA of current flowing through it. Is this a good way measure low currents when the meter can't handle it? (the cheap meter was showing ~200uA)

I don't have to run years from the battery, so would be thrilled if this amount of current is close to the actual value. I can also do some tweaking if this method is half-way reliable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Have you considered using an opamp as a "virtual ground" current to voltage converter.

Wikipedia Current-to-voltage converter

Stan

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you can pick up an old volt-ohmmeter, 50 microamp movements were common (20K ohms per volt) with typical <500 nanoamp resolution using a mirrored scale. A normally-closed pushbutton in parallel can be used for the sleep current measurement, as you have indicated.

I routinely use 10 ohm resistors in series with Vcc to measure running current, dont know about 100K, that seems high.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have used the same technique at times in a pinch.
You do not provide specifics about the avr or operating voltage, but your calculation is well 'within the ballpark' for the typical avr.

OHMS LAW : I = E/R
R = 100,000
E = 0.84
Therefore: 0.0000084 amperes (8.4 uA)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'll have to look over that opamp stuff in more detail. I eventually would like to build a little uAmp box (with display).

The micro is not an avr, but is supposed to be able to 'deep sleep' from 500na to 20na depending on a few things. I have a mcp1700 3.3v regulator and a battery charger ic also in the circuit, so 8uA seems like I'm in the right ballpark. There are a few things I can do to decrease it a little, but if I'm in the single digit uA, I'm happy.

If I run the circuit as explained originally, but instead use a variable power supply, and set it so that the normal battery voltage is seen at the circuit side of the 100k resistor, would that be more accurate? (would that be equivalent to the I-to-V opamp linked above?)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes but if you need the kind of accuracy for a publication get a used picoammeter - they are not all that expensive. At a minimum you need a guard voltage to minimize leakage.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You can measure very low currents with a voltmeter connected in series.As the sleep current might be around 500pA and the impedance of the instrument is usual 10Mohm then I=U/R.
But huge currents compared to the avr sleep currents are flowing through electrolytic capacitors in the power supply.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

No publication, for now just want to know which ballpark I'm in.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

µcurrent meter as shown on www.eevblog.com , is that an option ? For me it is !

A GIF is worth a thousend words   They are called Rosa, Sylvia and Tessa, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

a Unique circuit in EDN Design Ideas Measure small currents without adding resistive insertion loss. Includes a lively discussion leading to this 32 year old patent 4242741 - Floating shunt seismic amplifier.

Try running a google image search on current monitor and current to voltage converter --- I find viewing schematics directly much more helpful than the conventional text search.

Stan

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Plons wrote:
µcurrent meter as shown on www.eevblog.com , is that an option ? For me it is !
Yep. I own one also.

Cheers,

Ross

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Your Fluke 115 has 2% + 3 digits accuracy with DCV??

Get a <4$ DMM with a <0,5%+2 digits DCV accuracy. Those have a 100nA resolution at 200uA range so I think it would suit your needs perfectly. The one I own is called "DT-830".

It has 4-slope ADC so it self-cancels offset. You can calibrate gain by yourself with some cheap 0,1% voltage reference.

No RSTDISBL, no fun!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
If I run the circuit as explained originally, but instead use a variable power supply, and set it so that the normal battery voltage is seen at the circuit side of the 100k resistor, would that be more accurate? (would that be equivalent to the I-to-V opamp linked above?)

Yes that would be a tiny bit more accurate, but given your looking for a 'ballpark' figure it doesn't matter much.

Your nominal battery voltage is 4.1 volts when charged, so you would set your bench supply to obtain 4.94 volts at the regulator input.

A LiPo cell is generally considered depleted at 3.6 volts, so you would dial in 4.44 volts (again at the regulator input) to obtain the nominal 'terminal current.

I will wager that the difference between the 'terminal' and 'full charge' current is on the order of 100 to 150 nano-amperes... equivalent to a change in voltage across your resistor of .01 to 0.15 volts.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Look up the term "burden voltage," which is the voltage drop across a shunt resistance, and how can it be avoided (compensated, actually) by the use of transimpedance amplifiers (current to voltage converters).

This is the famous µCurrent that was mentioned above, with full schematics and documentation.

-George

I hope for nothing; I fear nothing; I am free. (Nikos Kazantzakis)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I will probably continue to use the simple method for now, as the measurement is made while the circuit is in a static low power state, and where I can can manually compensate for any burden voltage when needed. I think this will get me into the range I need, and I should be able to see what effect the code changes make (turning off brownout, wdt, pin states, etc).

Thanks for the info and links.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Afaik you can get it (uCurrent) in Lady ADA's shop in US

/Bingo

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Seems like OP does not have the requirement of a low voltage drop.
The problem is that OP owns an inadequate DMM with a 1mA DC resolution.

Quote:
I will probably continue to use the simple method for now
But you need to know DCV input impedance, anyway. If you do not know it - you will get erroneous results.

For example DT-830 has a 1M DCV impedance (all ranges) so XmV gives XnA current. If you connect this 1M in parallel with 100k, then you will get..
I know, you know, Ohm's law.

What is the input impedance of your (lets call it) multimeter on a mV range? Or, can you accept 20% error results?

No RSTDISBL, no fun!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The service manual shows >10Mohm for dc volts. I have not measured/tested it, though.

I'm not by my equipment now, but I suspect 10k will be a better value to use for my measurements. My meter has a 600mv range, so if I keep the voltage drop below 0.6v (measure <60uA), I don't think I will lose any resolution on the meter. The meter input impedance will then also be less of a problem, (the introduced error should be below what I can measure, and below the values I'm interested in).

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If it is 10M then with 10k in parallel you can neglect that current - the "multimeter" is over 2% inaccurate by itself anyway.

No RSTDISBL, no fun!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Not sure where the 2% comes from, its actually listed as +- 0.5% +2count.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Not sure where the 2% comes from

http://www.farnell.com/datasheet...
Quote:
its actually listed as +- 0.5% +2count.

Any reference?

No RSTDISBL, no fun!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

2% + 3 is revision A
0,5% + 2 is revision B

I hope this revision applies to a document not to a multimeter..

No RSTDISBL, no fun!