[cal40john]

Originally Posted by Rick

John,
First I apologise for not paying sufficient attention to the details, like failing to hit "shift" for a "+" and getting a "=" as you found, kinda like not checking my work.

One source for the formula giving voltage versus SG comes from an old IBE (a large manufacturer of batteries in this country) book and I've seen it quoted elsewhere as well. IN addition, I've verified the general accuracy.

I note a definite departure in the voltages versus SG for the first chart provided in your link in that it shows 12.478 Volts for a 1.271 SG @ 65 deg F when, for absolutely sure a 1.26 SG at 65 deg will show a much higher voltage (12.66 V). Of this I am sure. I question, therefore, all those voltages quoted.

The next chart down (in green and red) show a voltage for a 0% charged battery as being 10.5V. 10.5V is the end of test voltage for a 20 hour current discharge rate for making capacity tests. Obviously the voltage will rise after removal of the test current to a higher voltage (11.76 Volts) so this table is also suspect as not meeting standards.

According to Storage Batteries, by Vinal, aside from voltage variations caused by temperature of the electrolyte there are slight variations depending upon the grid makeup such as pure lead (like AGM and Gell-cell batteries and those grids of lead-antimony. There are also differences between the results made by various well-known electrochemists yet these variations occupy the second place past the decimal, i.e.; 2.1x on a volts-per-cell reading. Therefore, the third place past the decimal for nominally 12V batteries is of a highly dubious accuracy.

A comment was made about voltage during discharge rising after time. This is due to the discharge current being sufficiently high to cause electrolyte heating and plate heating which yields an increase in voltage. The electrolyte temperature increase also can result in a lower apparent cell resistance.

Looks like the table on the Trojan site uses the equation.

Trojan Battery Company - Testing

The differences from the table to the equation were a constant. Shifting the 0.85 constant to 0.845 made the equation match the Trojan table which is at 80 degrees.

John