Expanded Scale Voltmeter

[Pic_User]

The Problem
There is a limit to the Voltage allowed on PIC’s A/D input pin.
Generally limited to the supply Voltage (Vdd) of the PIC.
This is fine till we come to the problem of measuring a voltage higher than the allowed maximum.

A Solution
A resistance Voltage divider reduces the higher Voltage to less than the 5V maximum allowed.
See this thread: Voltage monitor for car battery
http://www.picbasic.co.uk/forum/showthread.php?t=6809

This circuit works to protect, the A/D input pin, from over voltage but it also divides all of the Voltage by four. That is okay but with batteries we are generally most interested in some portion of voltage near the high end of the range.

To read specifically the range of voltage that is within the permissible range that will not harm the battery. Suppose we want to monitor a “12Volt” battery.
For our purposes we might need to keep the range from 10.0V to 15.0V.

If we use a “simple Voltage divider” (Figure A) this would divide by 3 to protect the PIC. So 0V to 15V would divide to 0V to 5V.

More Problems
0V to 5V with the 10 bit (1024) input would increment 0.00488V per step (resolution). However we divided the Voltage by 3 so the steps at the battery end become 0.0146V per step. This may be okay for a lot of applications. Some need better resolution.

More Problems
See this thread: Stable Adc Reading Routine
http://www.picbasic.co.uk/forum/showthread.php?p=50499

Consider that we don’t expect our battery Voltage to fall below 10V or it will be damaged chemically. Now we are not using the bottom two thirds of the remaining resolution. So we only use one third of the available resolution.

That’s right, first we divided by 3 to get rid of the “top end” then we ignored the “bottom” two thirds of what we had left.

We can make an “Expanded Scale Voltmeter”.
See this thread: ADC - 2 channels but not same voltage reference - how to?
http://www.picbasic.co.uk/forum/showthread.php?t=7578

If we could find a way to loose the 10V (Figure C) that we consider our low end reading, we would not need to “resistance Voltage divide” by three. Also we would not need to ignore the bottom two thirds of the reading. So the whole 1024 steps would be used by the Voltage of interest. This “expands” the span of Voltage expected, to fit the whole resolution.

More Solutions
Of course Figure C would be difficult to couple to a PIC because, the “low side Voltage is floating 10V above the Negative terminal of the battery. Figure D is the same circuit with the components moved to allow the PIC common (Vss) at the same potential as the Negative terminal of the measured battery.

This circuit is not ideal, battery Voltages very near and below the Zener Knee will not display correctly
The Voltage on the reference Zener will vary slightly. Zener current changes, between 10V and 15V battery Voltage, this will affect the clamping Voltage. There are other ways to generate the 10V to subtract from the larger voltage. The Zener method has worked for me and I though you guy might like to try it.
<IMG SRC="http://www.picbasic.co.uk/forum/attachment.php?attachmentid=2335&stc=1&d=120235939 5">

Any thoughts or improvements on this topic?

-Adam-