I picked up a 9v 2200ma battery to replace the piles of standard 9vs I've been going through. When I plugged it in, my arduino started to smoke, but I was able to unplug it in time to prevent permanent damage. Voltage testing revealed that the battery was actually outputting ~10.5v (probably because its brand new), so I made a voltage divider to reduce it to 9v, but the arduino still smoked.
I would still like to use this battery, but I am afraid to plug it in. Could it be the current that is the problem? If so, can I build a current divider that would work?
I'm no expert on this stuff (read as: totally thick) but here are my thoughts:
According to the Arduino website you can run an Arduino (this page is for the Uno) up to 20v anyhow
The current available from a power source is not necessarily all drawn by a device, which only draws what it needs. A 9v battery the size of a house and which could provide a zillion amps, will still only deliver what the device draws.
So I'm thinking there's something else the matter..
Wow... if you mean the 2 mins literally, and assuming you're only driving the board not some motors and stuff, then I'd say the problem's not the batteries but the board's drawing too huge a current and something's wrong with it in the first place.
Those 9v batteries (I'm assuming the small matchbox sized pp3) are not very long-lasting, sure, but should drive an Arduino for more than a couple of minutes.
Edit.... it would be interesting to hook up an ammeter in series and see what the current draw is.
I'm running the arduino, a servo controller logic board (but not the servos), PS2 wireless controller receiver and a couple of LEDs from it. It goes quick, I'm lucky if I get 5 mins, usually it's less.
Sounds to me like your project is just drawing too much current for the on-board voltage regulator to support. The reason it doesn't 'smoke' when using the small 9vdc battery is that they have pretty high internal resistance and that is limiting the maximum current flow to the board, where you new 9vdc battery has much higher current capacity and is providing much more current to the project.
Maybe just power the arduino board with nothing else wired to it with the new board and load the blink sketch to see if it's running ok. If nothing smokes start adding your other components one at a time and then power up and see if the blink sketch still is running.
I anything wired to the Vin pin when you plug your battery into the external DC connector?
sandworm:
I picked up a 9v 2200ma battery to replace the piles of standard 9vs I've been going through. When I plugged it in, my arduino started to smoke, but I was able to unplug it in time to prevent permanent damage.
Ummm, if smoke came out then there's permanent damage...
The typical solution is to use a CCS probe and measure the voltage over the probe - or use an led.
In situations like this, you should be careful when batteries drain too fast, or supply voltage drops when a load is connected. Putting a stronger batter is a sure way to disaster.
When you power up, it is safer to power up with a power resistor, like 100ohm or so and monitor the voltage drop over that power resistor, to get a sense of current draw.
OMG, I hate to admit my shame, but here it is:
I solved it. Turns out since I had two 9v connectors hooked together, the polarity was getting reversed.
Danger! Danger!
Works perfectly now though, whew.
Thanks everyone for your help! Much appreciated.
In this situation, the battery's poles need to be hooked up opposite to the wire colors, or else you're sending opposite polarity into the arduino. The 9V connectors are a bit confusing that way. I'm going to switch to another method, but at the time this connector was already hooked up so it was the path of least resistance. Too bad it caused me so much grief.
Smoke or overheating [usually of the v.reg] implies too much load or battery connected
in reverse polarity - I would have guessed at #2 immediately. Arduino boards usually
have reverse-protection diodes in the power jack lines, but these may be bypassed
when connecting to the Vin header pin.
Even now that it's working, I would test the v.regs with my finger for too hot operation.
Those tiny surface-mount devices are generally deficient at handling loads of more than
a couple of hundred mA, with Vin = 9V or more.
