vlotage detect circuit

[guest]

Hello,
I need to determine when an alternator is running on a tractor. The power system is similar to a car. Vbat=12Vdc or so when the engine is off and 14.2 or so when engine is on. My initial thought was to use a resistive divider to bring the two levels into range of 0-3.3vdc (my supply voltage to a PIC18) and use the A/D to determine what the voltage level is. Looks good on paper. Any body have any experience with this sort of thing?

Thanks in advance
Eaton

[treitmey]

The old standby is a analog current meter inline with the alt. output.
Are you just trying to do that digitally?

You talked about voltage,... but mostly you want to know the current that is feed back to the battery. Also if the tractor is old, it is probably a DC generator, and not an alternator.

Can you elaborate more,.. What are you trying to do? Is this just for fun?
or is this going to production?

[Guest]

I'm not concerned with the current. I just want to know that the alternator is spinning and the voltage regulator is putting out voltage. a ckt is already in place for other functions and this is an add-on that the customer has requested and i need to do it with the least amount of haedware possible.

[treitmey]

I think that is right.
BUT
+12 to Gnd with a simple voltage divider
+12--1k--1k--Gnd
This will half the 12 to 6V between the resisters. Too much!

DO 1/4

+12---1k---1k---1k---1k---Gnd
or
+12--3K---1K--Gnd
That way between 3k and 1k to Gnd should read +3
SCALE BY 1/4
12 volts == 3
14Volts == 3.5V when read by A to D
And you have your window.
ps 3mA 36mW


Woops.. Thats still to high for your 3.3V PIC,... but you get the idea

[treitmey]

Anyone know what would happen if you try to read 3.5V on the A to D
with a 3.3V PIC


Smoke? ? ?

[PCM programmer]

There are tons of voltage divider calculators on the net.
That's the easy way. Put in your max output voltage of 3.3v
and put in your max input voltage of (let's say) 15.0v.
Then put in some reasonable value for your R1 series resistor,
which might be (let's say) 1.0K. Then click Compute.
http://www.raltron.com/cust/tools/voltage_divider.asp

[treitmey]

Right you are!
But still. What would happen. This isn't within the spec, right??
I mean, if you try to read A to D thats above the rail.

Does the chip blow.

[PCM programmer]

He didn't say what PIC he's using, other than "18F", so I'll assume
it's an 18LF452.

In Table 22-21 (A/D Converter Characteristics) of the data sheet, it says
the max analog input voltage (Vain) is AVdd + 0.3v. Since AVdd is 3.3v
in his case, the max analog input voltage is 3.6v.

There are also input protection diodes to ground and Vdd, which will limit
the input voltage to the PIC's A/D circuit to Vdd + 0.6v (which is 3.9v in
his case). But Microchip must have a reason for spec'ing the input
voltage to AVdd + 0.3v (3.6v), so I wouldn't exceed it.

[DragonPIC]

My spec for the 16F876A says absolute max is VREF + 0.3V so you should be safe with VREF 3.3V and VAIN 3.5V. Not sure of the accuracy after VAIN>3.3V.
_________________
-Matt

[valemike]

Like PCM said, it is clamped to be one diode drop above Vdd and one diode drop below Vss.

I've programmed an application which simply had a 100Kohm resistor in series to the output of a 12V transformer, going the RB0 interrupt. And if you scoped it out, you'd get a pseudo-square wave, whose period is 16ms (60Hz).

I figured this was well above the allowed voltage on any pin, but was told that it was okay because the PIC clamps the voltage, and that we were not exceeding current since the 100K resistor was there.

[JohnKennedy]

Why use a PIC at all to do this a simple comparater circuit would suffice and be alot cheaper, or even the TL431.

JFK

[DragonPIC]