12V battery indicator

[George L.]

Hello,

I attached a simple circuit I created that acts as a 12V battery indicator. It tells you how much power is left in the battery by lighting up 4 LEDs. Would the circuit work the way I think it will?

Also, is the diode right off positive necessary, I just want to protect against connecting it up the wrong way, or will the LEDs do that job for me?

Lastly, what do you think of the Zener values?

Would really appreciate some help.

thanks,

George L.

 

[Klaus]

hi George,
As a battery "power" indicator your circuit has much too coarse steps. You also did not say which type of 12V battery, whether its for a flooded, AGM, gel or for a different battery technology altogether, all would have a different discharge curve.

Ideally, to be useful, your circuit should have LED's indicating in 0.2V steps from 11.6 to 12.6V for a standard lead acid 12V battery discharge voltage. If you want to include the charging voltage as well then have LED's up to 14V. This is easier done with a LM3914 IC. There is heaps of info on the net about these chips. Also about battery terminal voltages as an indicator of charge remaining and about how far down these batteries should be discharged before their life span is greatly reduced. I suggest you do a bit of googling on this subject.

And as a final tip, LED's do also consume power so, if the battery is of the smaller size the indicator should have a push button turn on feature to light the LED's only while one wants to know the battery state.
Klaus

 

[Russlk]

The diode in series is a good idea because the reverse breakdown of the LED is about 5 volts. A problem with the design is that forward voltage of the LED is not taken into account. A red LED drops about 2 volts, so only the two left most LEDs will light on a full charge. further, the tolerance of the zeners and the LEDs has not been taken into account. A 5% 10 volt zener will vary plus or minus 1/2 volt. That is too much for a measuring instrument.

 

[George L.]

Hello Guys,

thanks for the replies. The battery I am talking about is a 12V 18AH Deep Cycle Exide Battery. I am trying to create this circuit for a FIRST robotics competition. The batteries are usually used up very quickly. When being used by the robots, they only last minutes. This circuit is supposed to be a "give away," something my team can give to other teams as a gift. Therefore I can't make it to complicated or expensive.

My circuit is very coarse, but it is just a rough indicator of which batteies are used, and which are already charged... basically.

Klaus, I understand your concern that it will be somewhat unaccurate and REALLY COARSE.

Also, the circuit is not permanetly connect to the battery. It is just hooked up for a few seconds to determine the charge.

Russlk brough up a major flaw...

A red LED drops about 2 volts, so only the two left most LEDs will light on a full charge.

Major problem, If I use for example a 8, 9, 10 and 11V zeners, I won't have this problem, right?

Would really appreciate your help.

thanks a lot,

George L.

 

[philba]

zeners are pretty inaccurate.

how about a quad comparator - a voltage divider for each comparator selects the 4 voltage levels. Power the circuit with a 7805 and make the reference 2.5V. You could use a real reference chip for better accuracy or just another divider. Each comparator's output drives an led. The TLC3704 can drive up to 20 mA per comparator so you could get by without led drivers.

Or you could just use an 8pin PIC... :shock:

 

[chemelec]

This circuit would be suitable.

**broken link removed**

 

[Klaus]

This circuit would be suitable.

**broken link removed**

Chemelec, you do a neat PC board if I may say so.

However, rather than going through all that exercise the OP should see if kits for this are available. I have had one for 15 years in my 4WD, its only a 10LED / single LM3914 version but it was quite cheap here in Australia. These kits are still available AFAIK.
Klaus

 

[Oznog]

zeners are pretty inaccurate.

The threshold voltage of the LEDs is even more inaccurate.
The intensity ends up pretty variable. For example, a red LED designed with a 9v threshold would be 8x brighter at 14v than it would at 9.5v. So it can't be made very bright at lower voltages because it'll blow the LED at higher voltages.