I have a quick question,

If I want to measure the voltage at an input of a PV panel (24 - 48V, 220W), and give that to a microcontroller for ADC purpose (0 - 3.3V ADC range), will it be enough to just use a voltage divider (by a factor of 16 = 48/3)? Is this safe and/or efficient?


Or do I need to use some fancy circuitry using Op amps as an ADC interface (for better resolution)?

 

48Vac → 15K+1K resistor divider (3V) → Op Amp Voltage follower(3V) → 3.3V clamp → ADC

Op Amp Voltage follower(3V) - for isolation.

You also should see this.
http://forum.allaboutcircuits.com/showthread.php?t=80082

 

It is "best/safest" to do as Scott suggested above as you get some isolation/high impedance.

Here is a neat site with some "simple" uC interface circuits
http://cq.cx/interface.pl#5

 

Assuming your micro had a common and very good ground to the panel a simple resistive divider is fine, just match it's equivalent resistance so it's not greater then the input R of your A2D. The equivalent resistance is just the resistors in parallel.

The 1K/15K ratio should be fine if the panel max voltage is 52.8 volts (full sun no load in winter I believe gives the max voltage). Even if it goes above this the ESD protection diode should clamp the voltage so as to not blow the A2D input, though the reading will of course be incorrect.

 

The output from the PV panel will be DC, not AC. So do I need the clamping circuit? I do not understand the purpose of the clamping circuit here.

 

The clamp is to ensure that if the voltage exceeds your 48V, the input to the micro (or output from the opamp or voltage divider) could be over 3.3V which can damage the micro input/adc.

 

The output from the PV panel will be DC, not AC. So do I need the clamping circuit? I do not understand the purpose of the clamping circuit here.

The clamping circuit is for the AC input signal, moving the negtive power to Zero, if the input is DC, then you can add a 3.3V zener, that is the protection for the ADC.

 

Ok that clears things up a bit.

Also, the "isolation" from the Op Amp voltage buffer doesn't deal with separating grounds, right? In order to implement it, I wouldn't have to separate the analog ground with the digital ground? Because I want to keep it simple, I'm going to short the analog ground with the digital ground. Having two grounds will make the layout more tedious.

 

Ok that clears things up a bit.

Also, the "isolation" from the Op Amp voltage buffer doesn't deal with separating grounds, right? In order to implement it, I wouldn't have to separate the analog ground with the digital ground? Because I want to keep it simple, I'm going to short the analog ground with the digital ground. Having two grounds will make the layout more tedious.

Unless you have a special purpose, otherwise you just need a common Ground.
If you no need to care the interference, then using a common Ground is ok.

 

But wait, if I supply the op amp with 3v power supply, then the maximum output from the op amp will be 3v. So adding a Zener diode will be redundant, right?

 

But wait, if I supply the op amp with 3v power supply, then the maximum output from the op amp will be 3v. So adding a Zener diode will be redundant, right?

Adding 3.3V Zener is used to protect the ADC, because the ADC is more expensive then Op Amp, if the ADC is included in the uC, then the uC could be damaged, if you think that you can afford it then you dont' need the 3.3V Zener.

48Vdc → 15K+1K resistor divider (3V) → Op Amp Voltage follower(3V) → 3.3V Zener → ADC