Belial88:
hmm let me understand this. so… I would wire the analog/ground as i’ve been doing for voltage reading. But I’d install a pot on the line… so middle pin to arduino, and one end would be to… not 5v but aref? I’d really be using a rheostat right (variable resistor…?)So i use this pot lined voltage reader… on the battery. I’m confused, why dont I just use 1.12 as my divider (2.8/2.5, or I want it to read up to 2.8 from 2.5v reference, or make the max 2.8 instead of 2.5) and I’m good to go? So the point of the pot is for calibration, but why do I need to do that?
But yea, I’m more open to using different voltages on each, because CPU VRIN really doesn’t need to be as accurate as VCORE. So 2.56v for vcore read, and then for vrin, i mean i could still probably use 2.56v but i’m still confused and interested in this idea. I’m trying to google this stuff but not having much success.
myself:
So a way to keep the better resolution is to add a voltage divider between the voltage to be measured and the ADC input. A pot is just an easy to adjust voltage divider. So with 2.8v across the pot, you could adjust the wiper of the pot so only 2.5v goes into the ADC. Of course any mobo voltage less than 2.8v is less than 2.5v, exactly 2.5/2.8 = 0.89285714285714285714285714285714 less. You then set the DIV_1 (and _2) numbers to be 2.8/2.5 = 1.12. That way when the ADC reads 2.5v, you know the mobo voltage is really 2.5v x 1.12 = 2.8v.
So the “top” of the divider goes to the voltage to be divided, in this case the mobo voltage CPU VRIN, the one voltage that might ever be > than the ADC ref voltage, the one that might be 2.8v. That way if CPU VRIN is ever > 2.5v, the voltage at the wiper (input to the ADC) will be < 2.5v, therefore less than the ADC ref voltage and thus won’t be pegged at 1023 counts. Because any voltage input to the ADC that is > the ADC ref voltage will read 1023.
1/1.12 = 0.89285714285714285714285714285714