next topic »

[Reply #15 on:]

Using a voltage divider from an unknown supply voltage to test the voltage of that voltage doesn't sound logical to me.  The reference point will change with supply voltage, proportionately, so how does that solve anything?

Use a precision voltage reference.  It's kind of like a Zener diode.  They are usually 1.225v or so, and only need some minimal current running through them to function.  Easily achieved through a current-limiting resistor from your battery.  Of course, that's only if you really need to know your battery level.  Using an LDO regulator is probably sufficient.  Dropping from 7v to 5v should be doable.  If the battery is SO low that it can't maintain a regulated 5v, you're in pretty dire straits.  Brown-out detection can be enabled to hold the micro in reset if the V+ drops below a threshold of 3 to 4v.

[Reply #16 on:]

So it will output less than 5V if the input is for example 6V?

Could I use an analog input of a microcontroller to find out if this point is reached?

Read this
http://provideyourown.com/2012/secret-arduino-voltmeter-measure-battery-voltage/

[Reply #17 on:]

Okay, thanks for your replies.

The article is very helpful.

What I want to do is use a NiMH Racing pack (7.2V), regulate it down to 5V and power a USB jack with it.
I want to use this as a cellphone charger.

Is it a good idea to put an ATtiny behind the regulator and let it check wheter the voltage dropped under 5V and then make an LED indicate the battery level (like discharged and not discharged).
A NiMH battery can hardly be damaged by exhausting it (as opposed to LiPo batteries)
I bought 7805 regulators.

(I made a thread for this project as well but I didn't get any replies.)

[Reply #18 on:]

Measuring battery voltage under charge can be misleading.. The voltage at the terminals Must be higher than the nominal battery voltage to charge it, It's called the forcing function and is used to determine the battery condition under charge.. If you know approximately the battery charge then a comparison of the voltage or current at several places during the charge cycle can flag a failing battery early and is a common technique to monitor and return those values to the controller to indicate apparent battery charge warning conditions early. Very commonly used to report battery conditions in un-monitored remote control points.

Bob

[Reply #19 on:]

actually i think u dont need a uC at all. there is simple circuitry involing transistor and zener diodes and leds to accomplish what u want. unforunately i dont have the schematic right now

[Reply #20 on:]

Do I need to check battery level at all?

Can I damage something on the USB device or battery if I just leave it connected?

[Reply #21 on:]

You can damage the NI-MH or CD packs by totally discharging them and leaving them in that state for very long. Small nickel crystals form and poke between the separator shorting the cell out. This doesn't happen except when the cell is completely discharged and is the major cause of cell failure.
Use a battery monitor of some kind and remember that it too draws current.

Bob

[Reply #22 on:]

something using this principle as example

[Reply #23 on:]

Is there a way to do this without zener diodes?

For me its totally ok to use an ATtiny because I have one here and I would have to order new stuff again and wait for it to be delivered

However, I just tested the 7805 and it works just as I expected

I connected my cellphone to it and it indicates that its charging, but it only consumes 0.11A and the battery level does not rise. Any idea whats wrong? The schematics says that it can provide 1A (thats also the limit of the cellphone wallwart).

[Reply #24 on:]

The cellphone will pull however much current it needs.  It won't pull 1A just because it can.  The charge controller will vary the charging current based on the condition of the battery and whatever algorithm it's programmed with.  I don't know if 110mA is considered typical or not.