If you want zero current into the meter below 5.4 V input here's how I think it can be done.

For some reason I can't upload the schematic so I will describe it to you. Photobucket used to work for me and now it won't upload either.

The first section switches on 35 µA to the meter for Vin > 5.4 V and the second section adds 15 µA per volt of Vin for 5.4 V < Vin < 6.4 V.
The - power supply terminal of the IC mentioned below is at ground and Vcc > 7 V.

Vin goes into the + input of a comparator (really an opamp running open loop) and the - comp input is connected to Vz, a 5.4 V zener powered through a resistor from Vcc.

The comp output connects to the meter through a resistor of such value that when the comp output goes to Vcc the meter is fed by 35 µA.
For Vin < 5.4 V the comp output is 0 V and the meter sees 0 µA.



The second section is an opamp set to, let's say, +3 gain, with the gain setting resistor that normally goes to ground going to Vz.

Vin goes into the + input, and the opamp output goes through a resistor and then to the meter so that the meter sums the current from both the 35 µA resistor and this one.

This resistor is of such value that with Vin = 6.4 V the opamp output voltage is at +3 V and another 15 µA is fed to the meter.
With Vin = 5.9 V the meter sees 35 + 7.5 µA.

Of course, I have to believe that with Vin below 5.4 V the output of this section will be at 0 V.
With dual supplies and a Vin of 4.4 V the opamp output would be at -3 V, but the output in this case cannot go below 0 V.

Hope it works!

 

I thought I'd report back before closing this thread. My circuit works perfectly. I finally breadboarded it and it works as expected. It did require a few tweaks, however, including bumping up the first stage gain a bit, Av=3.9 instead of 3, dropping the second stage gain a little, from 2 to 1.85, and to crank both pots just a bit from their calculated values.

I tested the first stage alone, setting Vbat to 5.40V, and trimmed the offset pot (1.90V instead of 2.0V) until I got the desired output. All subsequent Vbat values followed perfectly once the gain was corrected (empirically derived). The second stage then tested fine once I dropped the gain and tweaked its offset pot (1.64V instead of 1.4V).

The drawing here is the final version with the gain changes. Also, the pots in the previous versions have been replaced by fixed resistors here, with their actual measured values.

Again, thanks for your participation! This was a good exercise for me.

Attached Images

vbat4.PNG (28.1 KB, 7 views)

__________________
Corey

Rapp's Law of Inanimate Reproduction:
"If you take something apart and put it back together enough times, you'll eventually have two of them."