Measure battery condition?

[boriz]

You connect a capacitor from ground, to AXE pin and resistor from same pin to battery +V. You set the pin LOW, pause for a couple of mS to make sure the cap is fully discharged, then do a PULSIN on the same pin. The lower your battery +V, the lower the charging current, and the longer it takes for the cap to reach the logic threshold and read as a logic high. So you should be able to tell the battery voltage/condition by the duration of the PULSIN.

Would this work?

 

[Jeremy Leach]

Good idea, but don't think it would work Boriz, because Pulsin needs to see a transition to start timing the pulse.

I'm assuming you are talking about a pin that can be input and output.

Also, another point is that I think you'd need a resistor from the cap/resistor juntion to the pin, otherwise in theory you could have too high current through the pin when you discharge.

This is quite apart from what battery health really means ... and this has been discussed before (Beaniebots usually has some good tips about this).

 

[BeanieBots]

As Jeremy states, it won't work because pulsin requires a transition. However, your theory is sound in as much that measuring the time taken for a capacitor to reach a certain voltage fed via a resistor will give an indication of the source voltage. In fact, I believe that is the method which is used by the 08 PICAXE to do the low res ADC measurement.

Unfortunately, terminal voltage tells you very little about the CONDITION of a battery. A low voltage is certainly bad but that might just be because it heavily loaded. A voltage associated with full charge (or new if not a rechargeable type) does NOT indicate that a battery is in prime condition. As a quick example, a fully charged car battery will indicate about 13.6v on a DVM. Now stick a 1K resistor in series between your DVM and the battery. It will still read 13.6v but will drop by 1V for every mA you try to pull.

 

[Michael 2727]

Many security alarm systems (monitored ones mainly-ring out capable) do a battery
self test every month or so. This tests the 7Ah SLA backup battery.
They simply engage a small incandescent lamp for a fixed time, then measure the voltage.
When the battery gets a bit long in the tooth (4 to 5 years usually) the voltage at the
end of the test cycle starts to go lower, until a pre-determined "battery fail" voltage is
reached then a warning is sent to the control room - Send a Tech and replace the battery.

The same lamp is also an integral part of the SLA charging circuit, acting as a current
limiting device. You can see the pulsed charging circuit in action as the lamp flickers.

A capacitor across a (low) battery can sometimes make the battery appear to have more
juice than actually has, as they slowly steal charge (near full volts) and store it up.
A large capacitor can deliver a good thump of power temoprarily but if the same load
was placed on the battery by itself the voltage would plummet a lot faster.
This technique was commonly used in battery transistor radios to squeeze the
most out of the battery life.