I'm considering using an AD8397 to build a headphone amp because of its high current. However there seems to be words that AD8397 does not have built-in current-limiting and will fry itself when short-circuited.
I would really want to use/abuse this op-amp, and in AMB mini3's design (link) a 6R2 resistor is used in series in the output and in the feedback loop. But I have doubts with this method, and these are my concerns:
1) If the output is shorted to ground, then the circuit can be effectively redrawn as (op-amp's output pin) -> 6R2 resistor -> ground, and there is no negative feedback, so the output at the op-amp hits the supply rails.
2) Since output hits supply rails, then, say, 15V, exists across the 6R2 resistor. If I want to limit the current to 300mA, the resistor needs to be 50R, 4.5W. Will the 6R2 resistor allow for short-circuit of infinite duration?
3) Assuming feedback resistors are very large, then voltage at output = R_load / (R_load + 6R2) multiplied by voltage at op-amp's output pin. So the op-amp has to swing extra voltage. And if 50R is used instead of 6R2, and load is 32R, the op-amp has to swing a lot more voltage.
Are my concerns real or just thinking too much?
I would really want to use/abuse this op-amp, and in AMB mini3's design (link) a 6R2 resistor is used in series in the output and in the feedback loop. But I have doubts with this method, and these are my concerns:
1) If the output is shorted to ground, then the circuit can be effectively redrawn as (op-amp's output pin) -> 6R2 resistor -> ground, and there is no negative feedback, so the output at the op-amp hits the supply rails.
2) Since output hits supply rails, then, say, 15V, exists across the 6R2 resistor. If I want to limit the current to 300mA, the resistor needs to be 50R, 4.5W. Will the 6R2 resistor allow for short-circuit of infinite duration?
3) Assuming feedback resistors are very large, then voltage at output = R_load / (R_load + 6R2) multiplied by voltage at op-amp's output pin. So the op-amp has to swing extra voltage. And if 50R is used instead of 6R2, and load is 32R, the op-amp has to swing a lot more voltage.
Are my concerns real or just thinking too much?
The obvious solution is to NOT short the output to ground! Why would you ever want to be able to do that, anyway?
Also, just so people might have understood better, you probably should have also said that there is a 1.5k feedback resistor to the inverting input and 330 Ohms from inverting input to ground. The 6.2 Ohms is only from the output to the feedback pickoff point, i.e. in series with the output but inside the feedback loop. Also, there is a ferrite bead in series with the output, outside the feedback loop.
That is all quite typical. The only atypical thing that you have mentioned is shorting the output to ground. That is not a good situation, with most types of amplifiers.
How do you imagine such a short of output to ground could occur, when you will have headphones as the load? If there is a way, then you should first try to make THAT less likely. The amplifier topology is not necessarily the place to start protecting against that.
If you are going to worry about shorting the output to ground, then what about shorting the V+ to the V-? What if the whole unit falls into a fire? Or falls off of a tall building? Or falls into water while operating? Or gets run over by a truck? Or shot out of a cannon? Are you also going to worry about all of the other things that could occur? The amplifier topology won't protect against many of those, either.
Also, just so people might have understood better, you probably should have also said that there is a 1.5k feedback resistor to the inverting input and 330 Ohms from inverting input to ground. The 6.2 Ohms is only from the output to the feedback pickoff point, i.e. in series with the output but inside the feedback loop. Also, there is a ferrite bead in series with the output, outside the feedback loop.
That is all quite typical. The only atypical thing that you have mentioned is shorting the output to ground. That is not a good situation, with most types of amplifiers.
How do you imagine such a short of output to ground could occur, when you will have headphones as the load? If there is a way, then you should first try to make THAT less likely. The amplifier topology is not necessarily the place to start protecting against that.
If you are going to worry about shorting the output to ground, then what about shorting the V+ to the V-? What if the whole unit falls into a fire? Or falls off of a tall building? Or falls into water while operating? Or gets run over by a truck? Or shot out of a cannon? Are you also going to worry about all of the other things that could occur? The amplifier topology won't protect against many of those, either.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.