Zener Diodes Always Fail Shorted. NOT.
I found a simple self-oscillating class D amplifier schematic and built it on a stripboard. Despite the ugly layout it works (quite well i must add), and although i was expecting higher output before distortion for the given rail voltage, it seems like this topology "runs out of modulation" earlier than a clocked design, and certainly earlier than a class AB amp, and i can always use higher voltage rails when i design the final PSU. And things are certainly going to improve with a proper PCB anyway.
I was adding a couple more bypass capacitors to the driver circuit and i shorted something, as i popped the fuse when i plugged the thing in. I replaced the fuse with a lightbulb and proceeded to troubleshoot the unit, i found the short and cleared it. However, after i did this i was getting about 2/3 the power output, and the output transistors would get hot. As they always sat at room temp before (that's the main point of class D isn't it), there was clearly something wrong.
There are two back-to-back zeners that limit the voltage swing to the mosfet gates. On the top mosfet one of them was shorted. I replaced it, but the problem persisted. Upon additional testing, the other zener was found OPEN! After the second zener has been replaced the amp works fine again. And zeners are commonly used as overvoltage protection devices since they are "guaranteed" to die shorted. Wonder what would happen when your overvoltage protector fails open...
I was adding a couple more bypass capacitors to the driver circuit and i shorted something, as i popped the fuse when i plugged the thing in. I replaced the fuse with a lightbulb and proceeded to troubleshoot the unit, i found the short and cleared it. However, after i did this i was getting about 2/3 the power output, and the output transistors would get hot. As they always sat at room temp before (that's the main point of class D isn't it), there was clearly something wrong.
There are two back-to-back zeners that limit the voltage swing to the mosfet gates. On the top mosfet one of them was shorted. I replaced it, but the problem persisted. Upon additional testing, the other zener was found OPEN! After the second zener has been replaced the amp works fine again. And zeners are commonly used as overvoltage protection devices since they are "guaranteed" to die shorted. Wonder what would happen when your overvoltage protector fails open...
I've shot the comparator now, and it happens that i don't have a spare. It'll have to wait until tomorrow. Well, at least i got to know it well enough to know what blows. 
Comment