# SLA Battery warning circuit

I'm looking for a simple circuit to light an LED when a SLA's voltage gets below ~11.5 or some preset value. I found this one online but am having trouble understanding it. Nor does it function as it should in a circuit simulator.

I'm trying to work it out mathematically and it shows (and simulator shows) the LED will never properly turn off. At 12.5V lets say, the zener will have ~10V across it, meaning the transistor base must see 2.5V. The BC557's max Vec is .65V, the diode would drop say ~3V at best, probably less, meaning Ve must be 3.65V. Even at 13.5V, Vb < Ve, how is the LED supposed to extinguish?

The description says the LED should illuminate below 11.6V only. Will this circuit work? And if so what am I missing?

Edit: This circuit below seems to make more sense, simulator shows LED lighting at 11.8V and extinguishing at 11.9V. The only problem is it "wastes" the LEDs current to ground while it is off. Can anybody suggest an improvement on this?

• You need more precision than a Zener can provide. You will want a voltage reference of some sort (maybe a TL431). And a comparator to turn on the LED. I think the circuit you posted is designed to keep the LED on until the voltage drops to 10 or a little above, and then turn it off. Commented Jul 31, 2016 at 7:07
• The page quotes "This circuit will light an LED when the battery voltage decreases to 11.6V. When the LED turns on, you’ know it is time to charge..." And "The BC557 transistors’ base is biased by the 10V zener diode. As long as the battery voltage stays above 11.6V, the zener keeps the base of the BC557 transistor high. When the battery becomes discharged, the zener stops conducting and the base bias goes low, and BC557 begins conducting, and the LED alerts you to the low battery voltage." Precision aside, it doesn't appear to work at all? Commented Jul 31, 2016 at 7:26
• Can't believe everything you read on the internet. Maybe if you put in an NPN instead of a PNP. Commented Jul 31, 2016 at 7:30
• @JoshNZ You've analyzed the circuit and correctly identified that it does not work. You have also simulated it, giving some credibility to your calculations. Time to write that website off as a joke and move on.
– pipe
Commented Jul 31, 2016 at 8:36
• @JoshNZ You have now changed your question, you have two circuits. Remember that this is a Question & Answer site, not a forum.
– pipe
Commented Jul 31, 2016 at 9:31