Equipment for measuring low currents

[LarsT]

I'm looking for "cheap" equipment for measuring low currents (100nA - uA) and I've checked now and again for the last two years for the uCurrent meter, but it has always been out of stock. Either it isn't made any more, which makes me think there are comparable alternatives out there, or it's ripped away shortly after put for sale, which makes me think there's a huge market for this and more alternatives out there. Yet I can't seem to find any user friendy and reasonable cheap equpment (for hobbyists) for this purpose.

So what should I get, where should I get it and should I just give up the dream of ever owning a uCurrent meter?

[mimmus78]

There is keithley 485 in this range (and if you don't want more than 2mA). It has also an analogue output so you can contact it to a scope. It can be found on eBay usually less than 200 USD.

Inviato dal mio Nexus 6P utilizzando Tapatalk

[The Soulman]

What accuracy do you need and what meter do you have?

[Gyro]

You should try making your own. If you select a few appropriate current sense resistors then you can use a low offset x10 gain stage (using something like an OP07 opamp [Edit: or something more modern] ) to get your sense voltage within the 200mV range of an ordinary DMM.

With more current sense resistors than the uCurrent uses (use 1 resistor per decade) you can keep the gain down without suffering from significant voltage drop across the current sense resistor.  X10 gain would give you 20mV maximum drop. That should be insignificant in practice.

[colorado.rob]

There are a lot of meters that can do 10nA precision, including this one on the front page right now for under $20: https://www.eevblog.com/forum/testgear/an8008-us-$19-10000count-1uv-0-01ua-0-01ohm-resolution-meter

[MadTux]

Any decent benchDVM/DMM can measure currents in the 100nA range, even something old like HP 3455A/3456A without current measurement can do it. Just clamp something like a simple 10M resistor between the terminals and you get 10mV/nA. 3456A has fancy math functions, so you get the reading directly and don't have to do the math yourself.

Keithley stuff is usually much better than that, usually goes down into the pA and fA ranges. They also use other/better principle (balanced opamp) for measuring low currents (no more shunts, no more voltage drop over shunt). Basically the instrument acts as a current sink and uses an opamp with an feedback resistor to keep the input at 0V. More current into the instrument, opamp applies more (negative) voltage to feedback resistor to keep the input at 0V. Want a lower range, simply use a larger feedback resistor, the input resistance of the opamp is more or less the only limit. (read an old Keithley manual, if interested in this measuring principle)

 Oldish Keithley 602/610C are sometimes quite affordable (but only has analog display with something like 1% accuracy). Keithley 616/617/619 usually cost more, 6517 is even more expensive.

[joeqsmith]

Cheap powered shunt.

[joeqsmith]

The pre-production 121GW compared with my highly modified UT61E reading 100nA.  This is the only range that my UT761E has lower burden than the 121GW.   

Video showing my 61E mod to the low current input.

[Shock]

I own a Keithley 480 Picoammeter which I recently repaired, they are about the cheapest dedicated instruments around if you are willing to go the secondhand route. I see two just sold in the last 2 weeks on ebay for $26 and $10, a pretty sweet bargain I think. Shown in this photo it's the 3rd from the top (Kelthley 179, 179, 480 and an IET 510a), with 1 pico drift on the micro range.

[LarsT]

What accuracy do you need and what meter do you have?

I have a cheap multimeter purchased at a cheap hardware store a long time ago. It's highly inaccurate.

I'm not familiar at all with the price ranges for different accuracy levels, but I would like something to at least be able to measure uA up to at least 100mA, preferably with <1uA accuracy.

[LarsT]

There are a lot of meters that can do 10nA precision, including this one on the front page right now for under $20: https://www.eevblog.com/forum/testgear/an8008-us-$19-10000count-1uv-0-01ua-0-01ohm-resolution-meter/

This one seems interesting (=cheap). Any idea what quality and accuracy can be expected?

[colorado.rob]

There are a lot of meters that can do 10nA precision, including this one on the front page right now for under $20: https://www.eevblog.com/forum/testgear/an8008-us-$19-10000count-1uv-0-01ua-0-01ohm-resolution-meter/

This one seems interesting (=cheap). Any idea what quality and accuracy can be expected?

They state the accuracy in the advert, but until someone does a stringent test, it's impossible to know.  It's also not clear how much it drifts over time, how much it  is affected by temperature, or if it is possible to recalibrate it.

[Shock]

This one seems interesting (=cheap). Any idea what quality and accuracy can be expected?

Specs can be found here: http://szzotek.com/en/col.jsp?id=144
Quality seems ok for a $20 multimeter delivered to the door, it is not safe for using on highly energized circuits so take the ratings with a grain of salt. For low voltage AC/DC (e.g. under 30V) and low current work it should be fine.

[MrW0lf]

Somewhat unexpected device to measure low currents:
https://www.picotech.com/data-logger/tc-08/usb-tc-08-specifications
It's thermocouple datalogger with 8 channels and optional accessory:
https://www.picotech.com/accessories/sensors/tc-08-terminal-board
for  "±5 V, ±500 mV and ±50 mV, or 4–20 mA"
Noise free resolution claimed 16.25bits (20bit DAC).
I'm using this datalogger mostly for tuning my smart home project heating stuff, but out of curiosity tried adapter also - seems to work ok and waiting for some further use. Obvious advantage is fact that it's datalogger, not just single number on DMM etc. Not strictly budget solution but maybe idea to try if there is already some high-bit thermocouple temperature logger lying in drawer unused etc.

[Fungus]

They state the accuracy in the advert, but until someone does a stringent test, it's impossible to know.  It's also not clear how much it drifts over time, how much it  is affected by temperature, or if it is possible to recalibrate it.

I really doubt it's going to be accurate/stable down in the nanoamp range.

[tronde]

Maybe some inspiration from this thread?

https://www.eevblog.com/forum/projects/picoammeter-design/msg790045/#msg790045

[David Hess]

A good multimeter with its built in 10 megohm input resistance can be directly used as a low current meter in DC volts mode.  The 10 megohm input resistance yields 10mV/nA so 1mV/100pA and 0.1mV/10pA.  Using a multimeter this way has the nice property of including good input protection.

Cheap multimeters often have a 1, 9, or 11 megohm input resistance and their input resistance may change on different ranges.  Laboratory multimeters often have effectively infinite resistance on their low voltage ranges requiring an external shunt to be used.

I have a 3.5 digit DM502 which I use for this.  On its low voltage DC ranges, it can be configured to have an infinite input resistance so a simple external 1 megohm shunt converts the 200.0mV range to a 200nA range with 100pA resolution.  The DM502 was produced before excessive safety requirements so it uses standard banana binding posts making it easy to mount an external shunt resistor but the same thing can be done with a more modern meter by using an old dual banana plug and a set of non-safety leads.  If I use my 4.5 digit DM501 the same way but with an external 1.11 megohm shunt, then I get the same 100pA resolution but over a range of +/-2uA instead of +/-200nA.

If you want a zero burden voltage low current meter, then build a simple floating transimpedance amplifier using an LMC6081 and 9 volt battery to connect to your favorite 0 to 2 volt DC voltmeter.  Use a metal box for low noise and include a series resistance between the inverting input and summing point to protect the operational amplifier's inverting input.  With a 1 megohm feedback resistor and a 4.5 digit meter, that will allow measurements to +/-2uA with 100pA resolution and with a 1 gigohm feedback resistor, +/-2nA with 100fA resolution.  The burden voltage will be dominated by the input offset voltage of a couple hundred microvolts but this could be trimmed out if necessary.

In principle an LMC6001 or graded LMC6081 can work down to about 2fA (about 6,000 electrons per second!) and I have done this however it is probably better to make a dedicated tester if you want to measure currents that low.

[alm]

Cheap multimeters often have a 1, 9, or 11 megohm input resistance and their input resistance may change on different ranges.  Laboratory multimeters often have effectively infinite resistance on their low voltage ranges requiring an external shunt to be used.

Any idea what the input bias current is like on cheap meters? Lab meters might have something like 100 pA max, so if the cheap meters are substantially worse, this might limit their performance.

[David Hess]

Cheap multimeters often have a 1, 9, or 11 megohm input resistance and their input resistance may change on different ranges.  Laboratory multimeters often have effectively infinite resistance on their low voltage ranges requiring an external shunt to be used.

Any idea what the input bias current is like on cheap meters? Lab meters might have something like 100 pA max, so if the cheap meters are substantially worse, this might limit their performance.

For mine which have input bias current specifications, about 50pA max but in practice they are 10pA or better and modern CMOS input ones should be that good or better.  For a CMOS design, the leakage should all be from the input protection circuits.

If you know the input resistance, then you can calculate the input bias current based on the voltage reading with nothing attached but usually it will be zero volts so not very informative.

[Neomys Sapiens]

Apart from the specialized instruments, the Unigor 6e has a 1µA FSD range with a nice, big scala.