I've made a circuit to drive 7 CCFL fluorescent lights through a 12v inverter with a ULN2803A eight-count darlington array. I drive the pins on the darlington with the PWM outputs of an Arduino Mega. A photo of the circuit is below. The power is driven by a 7.5AH 12V Lead-Acid battery. The problem I am having is that the transistor IC heats up to deadly levels whenever I run more than two CCFLs at a time. It seems like the chip is rated for current like this from its Data Sheet but it fries every time. The same effect happens using this design in a breadboard, so I know it's not a wiring issue. I tried putting a 1K resistor between the darlington and the CCFL inverter but the lights didn't turn on at all...
The ULN arrays are rated for 500mA pr. pin but the package has a total heat dissipation capacity that is smaller than the sum of doing that on all the pins at the samet time.
You probably need a dedicated transistor pr. display, that can handle the current. Find the current needed for each diaplay, and find a suitable transistor.
The package is a kind of a no-information white box that says "CCFL Inverter DC12V input type" with no other information. How do I find out the current draw of this light? A similar product online lists these specs:
Input voltage of inverter: 12v
Output voltage of inverter: 680v
Current draw: 5.0mAv
which is confusing.
what sort of transistors or darlington drivers can anyone recommend?
is a darlington driver just another transistor that is driven from the IC darlington array? is the solution to use 7 of something like this http://datasheet.digchip.com/000/000-1-D45H11.pdf instead of the chip or drive transistors from the chip? drive those transistors from the chip? should i use a MOSFET? so many possibilities...
Just try it with half the number of loads and see if that helps. If so you need two chips.
Otherwise transistor or FET will do, just make sure it is a logic level FET.
If you don't know the current of the load then try measuring it with a meter. If you haven't got one then get one.
thanks, splitting the load between two of the darlington arrays seems to have done the trick. each chip still gets quite hot, but not hot enough i guess to pass the 85c mark of chip destruction. the tutorial is good reading, my troubleshooting arsenal is now expanded and my fluorescents are on.
;D