• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

opamp - integrating circuit

Thread starter #1
I am currently writing my extended essay in Physics (high school), on a comparison between analogue and digital integration.
I have a question concerning the integration circuits, using opamps. I found that it only works, using ceramic capacitors (having a LM741 opamp and a 470nF capacitor and 100kOhm resistor). My problem is now, why is that????
I would really appreciate, if there is anyone, who could help me out on this. I mean it integrates very nicely, but i have no clue why it has to be a ceramic capacitor and not an electrolytic one (since I have to give areason in the essay, and I'm also curious myself)???

Thank u very much!

Perhaps it's because ceramics have a lower leakage current than electrolytics. BTW, are you using standard electrolytics, or bipolar ones?


Active Member
No its not true that Op-Amp integrator works only with ceramic capacitors. But the true fact is electrolytic capacitors are not avaliable in very small values (i.e. less than 0.1uF). Whereas ceramic ones are available in very small values (1pF or so). So using large value electrolytic capacitors will increase the time constant of your integrator and will behave differently than a integrator with a smaller time constant.
To see integrator with electrolytic capacitor functioning, try changing the input signal frequency and you will see the difference.

Regarding to pebe post "WHAT IS DIGITAL INTEGRATION"
Digital integration is done using digital processing chips like DSP or you can also do it using microcontrollers or microprocessors. The final output is same but the method of calculation is different. The input signal has to be digitized before integrating it digitally. Here in this case accuracy of output is very good. Also leakage problem of capacitors etc. problems are also not present.[/b]


New Member
Digital V Analogue

I don't think I agree that the two types are the same - they suffer different problems and have different strengths -

Analogue is closest to true integration, its main problem is with loss of charge from the capacitor (electrolytics are poor in this respect and will cause errors for that reason). If the integrator ramps too far it will hit a power rail and either suffer an abrupt stop or start to "wind up" with the capacitor still charging but the amp output unable to follow it (depends on design).

Digital will (as Kinjal says) use an ADC - this will have an error caused by quantinisation (rounding to next digital value up or down according to type), this value will be effectively added to some running total. If the ADC output is zero it will have no effect on this total, if negative the total will decrease. This total is (with some scaling) the integrator output.
If the total tries to exeed what the system can cope with there will be an error - the system must stop the integration or there will be an overflow with a large signed positive value suddenly appearing as a large signed negative (two's complement overflow).
The advantage here is lack of drift for a zero input (the capacitor problem). Problems include its time sampling nature, it will sample at a set frequency and can therefore give very wrong answers as it cannot see sharp spikes or differentiate between a smooth ramp and a staircase if its frequency approaches that of the integrator sample time.
Digital remains an approximation of the real world.
Analogue computers are still used for some jobs, partly for this reason !!!

As for the electrolytic, Phasor makes a good point, if the circuit tries to apply a negative charge to the cap it will really cause errors in addition to the poor leakage properties of these poxy devices (good electolytics are available at a cost but they are rare).
Thread starter #6
Thank u guys a lot, I think i got some idea now. My experiment is coming along well btw. Have some nice curves logged, the only job left now is the evaluation of my results :? .
Noise is gonna be a major part of it, the curves look like someone stepped on 'em! :lol:

Concerning the digital integration, - it is basically I using a computer, and a trapezium approximation, which will give some results, hopefully.


Latest threads

EE World Online Articles