OK, here we go.
This is using a Rallye temp. gauge.
Connected to a spare 12v battery, in order to see any movement on the gauge needed a 150 ohm resistor in series (R = 150), and at that I could just barely see the gauge move. There was about 1v at the gauge. That means the gauge resistance at the bottom of the scale is about 13.6 ohms.
Swapping for R = 100 the gauge moved just a little way onto the scale and showed about 1.5v across the gauge posts (which is consistent with ohms law if gauge resistance was 14.2 ohms. So it seems it is true that the gauge resistance increases with more voltage (presumably due to the fact it works by heating itself up internally).
The increasing resistance is helpful as it makes the gauge more sensitive higher up, which is good for use as a charging voltage indicator because there will be more deflection at higher voltages where we want to see some movement more easily.
To get the gauge to half way I had to use R = 65. that measured at about 2.2 volts on the gauge which means it's resistance had gone up to about 14.6 ohms.
Working back on the basis that we want the gauge to show 12 volts dead center, so R=65 is about right, that means the gauge would not start to move until total voltage was about 5.8 volts.
That is also good as it means (even if it was linear) we ought to see the gauge move about a fifth to a sixth further when V=14 instead of V=12, ie. when the system is charging.
However, one bad thing is the gauge took a little over one minute to reach half scale from rest - !!
Next I calculated that to imitate a 14.5 volt supply using my 12 volt battery I needed to use a value of about R= 50. That produced a surprisingly big swing towards the top end of the scale and nearly 3 volts on the gauge. That means the gauge resistance had jumped to about 16.5 ohms
It seems the value I was assuming in my earlier post for the gauge full scale and temp. sender resistances was incorrect, but not too far out.
What is evident is the gauge responds quite a lot more in the top half of the scale which is due to the ratio of the gauge resisistance to R going up more quickly - probably then I had actually simulated more like about 15 volts if R = 65 (as I had assumed a lower value for gauge resistance).
So, on my gauage at least, and using a resistor in series of about 65 ohms, it is possible to make it work as a volt meter which begins to move at about 5 volts, reaches half scale at about 12 volts and which would move noticeably into the second half of the scale at 14.5 volts. At 18 volts it would be right off the top of the scale.
The downside seems to be its somewhat lazy response time, but that would always be the problem using one of these gauges because of the way they work. I can't see how any other system would make it move more quickly, although a system based on a pulsed 12v might move it a bit quicker maybe?
