# Forum: Analog Circuits to average a current over a given time analog

Rate this post
 0 ▲ useful ▼ not useful
Hello,

I am looking for a circuit that forms the average of a current over a
given time (like as long as someone presses a push-button).  Does anyone
too.

The reason why I am asking is because I have to develop a device that
can measure a very low current and therefore it needs a very small
resolution (0,001uA) but also has got a wide measuring range (10uA) -
actually it should be able to measure up to 100mA but I divided the
range in smaller pieces.

Usually I could choose a clock-frequency for the uC independently from
the resolution of the ADC but building a average joins these two
separate aspects together. So, I am now thinking about using a ADC with
a very high resolution, which means that I can do all the calculating
within the uC, or building the average analog and using a ADC with a
ordinary resolution of (10 or 12bit).

Of course I have considered the solution to integrate first and then
diving within a uC-program. But as I have to develop this device for my
thesis I want to know something about every of the three solutions. And
thats why I am especially asking now for the complete analog solution.

Gabriel

Rate this post
 0 ▲ useful ▼ not useful
Hi Gabriel,

I think your're looking for what we in German call "Meßgleichrichter".
It consists of two OpAmps and two diodes - purely analog ;-)
One of the OpAmps compensates the forward voltage drop of the diodes so
that you can theoretically measure voltages in the scale of a few
hundret of milivolts. The second OpAmp amplifies the result with an
I think, you could search the internet using key words like "measuring
rectifier opamp" or something like that, as I don't know the exact
translation of "Meßgleichrichter".

Cheers,
Jochen

Rate this post
 0 ▲ useful ▼ not useful
The analog solution would indeed be an integration.  The advantage
over the digital solution is that you are actually averaging the
signal continuously, where the digital solution samples it at
certain points of time -- any change between two sample points
will be missed (like, a spike).

The disadvantage of the analog solution is that any offset error
in the integrator will eventually sum up as well.

Rate this post
 0 ▲ useful ▼ not useful
Hello Jörg,

what about using a low pass?

And how to stop measuring then? Whould you prefer to use a potentiometer

Thanks for your reply!

Rate this post
 0 ▲ useful ▼ not useful
And isn't a low pass the less complicated system regarding to noise and
temperature changes instead of using a op-amp for (math.) integration?

Thanks again in advance!

Rate this post
 0 ▲ useful ▼ not useful
Gabriel wrote:

> what about using a low pass?

Essentially, an integrator is nothing different than a low-pass
filter.

Rate this post
 0 ▲ useful ▼ not useful
------------------------------------------------------------------------
--------

Gabriel schrieb:
> I am looking for a circuit that forms the average of a current over a
>
> given time (like as long as someone presses a push-button).  Does anyone
>
> knows such a circuit?

Do a search for: ramp generator.
That's what does the non-digital solution.

>
> too.

It does work like a charm, but it seems to be old-fashioned.

• $formula (LaTeX syntax)$