Also be aware that with any bipolar transistor there is a thing called second breakdown. It is entirely possible to be within the voltage and current rating of a device and be outside the second breakdown curve. It may be rated at 1 amp and 100 volts but it will not sustain one amp of current with 100 volts across it at the same time. Manufacturers provide curves that show "safe operating area". If you get outside this area catastrophic failure is immanent.
__________________
The hardest thing to overcome, is not knowing that you don't know.
Its no clear if you want to know "what can cause the failure" or "how the IC operates after it has failed"
One long term reliability problem that is usually forgotten about is "metal migration" due to excessive current. This can take place over a long period of time where the current denisity in the metal conductors (on the IC) is so great that electron velocities high enough to dislodge and move molecules in the metal conductors. Usually the conductor will go to an open circuit but it could it actually short two circuits together.
As far as symptoms of failure are concerned, there are two main failure modes:
1. Two nodes are on/connected all the time.
2. Two nodes are never connected.
I experienced a strange failure mode of a power transistor. When the transistor heated up, the collector would open circuit. Once the transistor cooled to room temperature, the collector function was normal. When investigated by the manufacturer, it was discovered that the transistor die was craked and when it was hot, the silicone expanded and the metal conductor was pulled apart. When the transistor cooled, it returned to normal dimentions and the broken metal conductor ends would remake contact.
is there any transient protection on the circuit? If the Darlington is turning on and off an inductive load you can get an inductive kick-back, which is the magnetic field collapsing to produce a high voltage. If this is not dealt with via "snubbing diodes" you can have this problem. Many Darlington arrays have them built in.
Hi friends, good do see all of your responses , here i want to throw some more light on the type of failure i'm coming across in my product:
the driver ic i'm using is SDH02 having having a current rating of 1.5 amp and voltage rating of 100 volt as one of our friends have mentioned in his reply. and the operating current of dirver ic in my application ranges from 0.65 to .85 amps, in this scenario where the operating current is well below the max spec specified in the datasheet, what could be the reason for failure???
one more thing i want to add here is.... its not that it fails immediately when i switch on the application, but it fails after 3 to 4 months of operation.
Possibly an inrush current limiter would help. transient heating from high inrush currents can cause a gradual fault by short term temperature spikes in the device which cause gradual diffusion based failure. Early indications would be(possibly) lack of saturation on the output causing partial turn on.
Cure, some inrush current limiting by resistor or inductor?
How does the failure show up?
inspection of the safe operating area chart might help. Operating outside it can cause that progressive diffusion failure by heating. If you want to operate a little outside the SOA, pay good attention to heat sinking with the proper heat sink and thermal paste
Hwell do you drive the IC? In operation what is the voltage drop across the IC. More than 1.4 = operating in linear mode = higher heat dissipation.
Excess motor current = burnt parts, thus my current limiter comment.
You may have to redesign it for a max worst case current with a current limiter in each leg. This is not as hard as it look as you can make a 3 terminal regulator do this easy enough. It all depends on the cost and extra $$ you have versus this problem. It looks like it was not properly designed in the first case. Possibly someone here can help him in this area?
You may want to consider that the margin you indicate is not adequate. If I were driving a motor I would tend to use a little more than the less than 100% margin you have indicated. As aurizon mentioned if you don't drive the driver hard enough or if it doesn't have adequate internal drive to handle periodic surge conditions you will end up in the linear region of the part and very likely in the second breakdown region. Second breakdown is certain catastrophic failure and heat sinking will not prevent it. I would suggest you go to a larger current capacity on the driver.
__________________
The hardest thing to overcome, is not knowing that you don't know.
Driver IC Failure Modes
Re: Driver IC Failure Modes
