# average value of a non uniform wave/signal

Discussion in 'General Electronics Chat' started by dkthechamp, Mar 9, 2010.

1. ### dkthechamp Thread Starter Member

Mar 3, 2010
10
0
Can anybody suggest a simple low power circuit/method to calculate the average value of a continous signal/wave in real-time?

2. ### t_n_k AAC Fanatic!

Mar 6, 2009
5,448
790
A low pass filter comes to mind. But what sort of waveforms or signals are you thinking of? Your question is perhaps too general.

3. ### dkthechamp Thread Starter Member

Mar 3, 2010
10
0
The signal is the attached waveform and the frequency is 1 Hz. The objective is to find the average value of the signal amplitude after each 1 seconds.

I'm interested in integrated solution without any passive component. One apparent solution is op-amp based integrator but in that case i get a continous output !

All suggestions/remarks will be highly appreciated.

File size:
203.3 KB
Views:
28
4. ### Papabravo Expert

Feb 24, 2006
11,788
2,497
The only thing that comes to mind besides a resettable op-amp integrator is a DSP solution where you have an anti-alias filter followed by an A/D converter and a DSP to compute the average over some time interval.

Last edited: Mar 9, 2010
5. ### dkthechamp Thread Starter Member

Mar 3, 2010
10
0
Digital solution is already there but i'm interested in analog circuit consuming less power, to achieve the same objective.

Any doc/reference expalining resettable op-amp integrator solution for average calculation.

6. ### Papabravo Expert

Feb 24, 2006
11,788
2,497
Not that I am aware of. The basic idea is to setup an integrator, and place an analog switch in parallel with the integrating capacitor. To reset the integrator you discharge the cap by closing the analog switch and shorting the output to the input.

Mind you I've never actually done this so I have no idea what problems you may run into.

7. ### russ_hensel Distinguished Member

Jan 11, 2009
825
57
For a PIC 10 ma is more than enough, where does the low power requirement come from, and what sort of value are you looking for? ( be curious to know why you want to do this, can you tell? ) There are some problems with all the analog solutions.

8. ### dkthechamp Thread Starter Member

Mar 3, 2010
10
0
When i say low power, i meant circuit working in weak inversion (sub threshold) region and the consumption being into few nA.

9. ### russ_hensel Distinguished Member

Jan 11, 2009
825
57
I this what you mean: Subthreshold conduction or subthreshold leakage or subthreshold drain current is the current that flows between the source and drain of a MOSFET when the transistor is in the subthreshold region, or weak-inversion region, that is, for gate-to-source voltages below the threshold voltage. The subthreshold region is often referred to as the weak inversion region. The terminology for various degrees of inversion is described in Tsividis.[1]

And low power with no passive components, so no wire, no resistors, no caps? I would say no can do.

10. ### dkthechamp Thread Starter Member

Mar 3, 2010
10
0
Your are right though bulk driven OTA architecture seems to be a potential option ieeexplore.ieee.org/iel5/4/21268/00987079.pdf?arnumber=987079
with only one internal capacitance. I understand that a capacitive element would be required for charging the output current.

Now the issue is, since input signal is very low frequency 1-Hz, a huge capacitor order of few mF could be required. As i'm looking for a highly integrated SoC i'm interested in few p/f Farads of value !!!

Any thoughts on this ?

Thanks,