Analog Circuit Outputs

[Akshatha Venkatesh]

an analog circuit's output is given to an ADC pin of the microcontroller , does the current flow to the pin too ? I have some very fundamental doubts, It would be a great help if anyone can clear it ?
Thank you.

 

[Harald Kapp]

Where there's voltage and a resistance, there's current, too: I=V/R.
An ADC's input has a finite resistance (less than infinity). Do the math.
It also has some capacitance which needs to be charged and discharged by the input voltage.
The direction of the current depends on many parameters like kind of input (single ended, differential), supply voltage (single, dual), direction of change in input voltage (to charge or discharge the input capacitance) etc.

 

[Herschel Peeler]

an analog circuit's output is given to an ADC pin of the microcontroller , does the current flow to the pin too ? I have some very fundamental doubts, It would be a great help if anyone can clear it ?
Thank you.

The input of the ADC is usually a very high resistance so very little current should flow. Mostly it just "feels the voltage".
I think Kapp is only an engineer and doesn't understand simple questions. He only sees complicated solutions. Not wrong. Just not simple.

 

[Audioguru]

Kapp simply explained the simple basics of electricity (Ohm's Law) that was not taught in India.

 

[Herschel Peeler]

Kapp simply explained the simple basics of electricity (Ohm's Law) that was not taught in India.

That he did but did it answer his question in a way he understood? Simple questions need simple answers,

 

[chopnhack]

An ADC's input has a finite resistance (less than infinity). Do the math.
It also has some capacitance which needs to be charged and discharged by the input voltage.

Kapp touched on an excellent point that most (if not all) μcu's have as a limitation of their ADC port. The input pin has a switched capacitor on it - thus if the input impedance is too high (i.e. - using high resistances to reduce current coming into the pin) there will be a delay in acquiring the voltage as the cap charges up. Too long of an RC delay and the timing maybe altered. Some pic's have a max impedance of 2.5k Ω's!!

More information is needed to discuss possible solutions. To give some food for thought, you may isolate the ADC's limited impedance with a log amp.[/QUOTE]

 

[(*steve*)]

I am not a fan of excessive formality (see Peeler, Hack, et al; posts #3-6) and I think of my fellow moderator as Harald rather than Kapp. Unless you're referring to some text he has authored?

 

[Harald Kapp]

I think Kapp is only an engineer and doesn't understand simple questions. He only sees complicated solutions.

I think I did understand the question. I could have replied simply with 'yes'.
What help would that have been to the op in clarifying his doubts?

With reference to 'only an engineer': A good engineer imho not just solves a problem - this any tinkerer can do given enough time. A good one at least tries to understand the issue at hand (he may not always succeed) to avoid the same issue in the future.

And yes, call me Harald

 

[Herschel Peeler]

Kapp touched on an excellent point that most (if not all) μcu's have as a limitation of their ADC port. The input pin has a switched capacitor on it - thus if the input impedance is too high (i.e. - using high resistances to reduce current coming into the pin) there will be a delay in acquiring the voltage as the cap charges up. Too long of an RC delay and the timing maybe altered. Some pic's have a max impedance of 2.5k Ω's!!

More information is needed to discuss possible solutions. To give some food for thought, you may isolate the ADC's limited impedance with a log amp.

[/QUOTE]

No argument at all with the correctness of his answer. I only suggest that the answers fit the level of the question.

 

[Herschel Peeler]

I think I did understand the question. I could have replied simply with 'yes'.
What help would that have been to the op in clarifying his doubts?

With reference to 'only an engineer': A good engineer imho not just solves a problem - this any tinkerer can do given enough time. A good one at least tries to understand the issue at hand (he may not always succeed) to avoid the same issue in the future.

And yes, call me Harald

Yes you were quite correct in your answer. You could have answered in French or German and been quite correct. If a tinkerer asks a question a tinkerer's answer is all that is needed.

 

[Harald Kapp]

How do you know Akshata is a tinkerer? For all we know he could as well be a student eager to learn something.

Too bad Akshata himself/herself hasn't replied.

By the way: my French isn't that good. Chnaces are an answer in French would have been not too correct

 

[davenn]

Thread closed

Yes you were quite correct in your answer. You could have answered in French or German and been quite correct.

Herschel please stop being so argumentative
And in future PLEASE refer to people by their first name