Looking for good and fast uA meter

[stan33]

Hi,
I am developing devices that have to run several years on alkaline batteries. The average power consumption of these devices is about 30uA, but in fact the idle current is few uA, with a 1ms burst of 30mA every second (and even some bursts at 400mA sometimes).
I am having troubles to measure the average current for such devices.
I tried using Agilent 34450A : not fast enough (600Hz sampling) and low accuracy (resolution of 10uA in fast mode). I also tried with Tektronix TX3 and get better value, but still to slow (1kHz), and voltage drop too high during current burst causes my device to reboot.
I am wondering if it possible to find a lab power supply with fast and high accuracy uA meter and averaging ? I guess it should be easier for a power supply to measure current on various range (100mA-10uA), and to switch quickly from one range to another one to avoid voltage drop on the load.
If not, could someone help me to suggest Amper meter (handheld or table), with fast sampling for good averaging (> 10kHz), fast range switch to prevent voltage drop on the load, and high accuracy ( ideally 100nA resolution) ? I saw Keithley DMM7510, but is quite expensive, also Siglent SDM3065X but  I am not sure how it will deal with fast current burst average.
Thanks for your help.
 

[voltsandjolts]

https://www.joulescope.com/

[_Wim_]

https://www.qoitech.com/otii/

Edit: I see this has insufficient bandwidth for your needs

[_Wim_]

Another alternative could be to put a large capacitor and resistor between the battery and the DUT. This will spread & lower the current peaks over a longer time , maybe allowing you to use your 34450A. Its will be less accurate than the solutions above, but is maybe good enough and a whole lot cheaper...

[Calvin]

Hi,

maybe You could build something like a leakage test circuit.
At the time I´m evaluating a very simple circuit originally by Jim Williams that was designed for ranging from 100pA to 100µA, selectable over 7 decades.
It basically consists of a low input bias current OPAmp and two (matched) transistors.
It puts one of the transistors in the feedback loop of a inverting OPAmp topology and translates the input current into a 100µA FS output signal that can be read from a 100µA analog meter.
As the DUT is working into a virtual ground it circumvents problems of the typical extremely high-impedance inputs associated with suitable OPAmps.
Even with a simple AD711 and a pair of roughly matched BC327 one can already measure down to 100nA with high precision.
With more suitable OPAmps like LTC6240(HV) or ADA4530-1 one should be able to reach down to 10pA with good accuracy.
I didn´t test the amplitude response yet, but I assume it should be fast enough for Your application.
If you wish I can probabely send You the link (it was in one of the early ANs of LinearTech) and the circuit schem late tomorrow.

regards
Calvin

[alm]

If you are open to buying used equipment, then a battery simulator like Keithley 2302/2306 or HP/Agilent/Keysight 66309/66311/66319/66321/66332 might do what you want. They are designed for exactly this measurement. They do the high dynamic range current measurement, but resolution/accuracy might be borderline. They can also do fastish digitization of the current, or integrating (averaging) the current over a period. They are sometimes available on eBay for $100-$200.

[Kleinstein]

The idea with a large filter cap to smooth the current does not have to be low accuracy. The alkaline battery is also a source with no so low impedance and absolute constant voltage. With such extreme variations in the current it may very well be the best bet. The method may profit from a voltmeter that has good performance in a continuous sampling mode. The 34450 has a suitable SD ADC, but not sure about input amplifier auto zero. With the filter cap and maybe an additional filter, the meter does no longer have to be very fast. It just should avoid longer breaks in the measurement, from a relatively slow auto zero mode or a classical dual slope ADC.