the thing is of you couple all of the inductor, then your transient response is worse.?
-because in a coupled inductor, its inductance as seen from one of the outputs depends on the current being flowed through it by the other output?
-it must do , because current flowing through it will put flux in the common core
-because in a coupled inductor, its inductance as seen from one of the outputs depends on the current being flowed through it by the other output?
-it must do , because current flowing through it will put flux in the common core
Well that's why you couple them, so they track each other. With both positive and negative consequences. But remember that if you use ferrite, it has a hard saturation curve, and unlike powdered iron materials which have soft saturation, the ferrite cored inductor would be designed so that it never saturates. Then, its inductance does not change.
Sharing the same transformer and fully coupling the output inductors are the only way of proper cross regulation if the supply is regulated, and also allows sensing of the positive rail only.
Thans Th3 Un1qu3
.......Though surely you agree that the coupled inductor gives a vchanging inductance even if there is no saturation.
Suppose one half is on load and the other is not............(ie i refer to the halfs of the split power supply output)
...then the one that is on load flows current through the coupled inductor........that will put a flux in the core that is in the direction such as to make the coupled inductor look much higher inductance to the one thats on light load.
so i say to people, that the coupled inductor, as used at the output of split output, is not a constant inductance, but varies with the loading of each half of the split output.
-this is not a good thing...since we would ideally prefer to have the output inductors the same value all the time.....ie a low value which would give fast transient response.
.......Though surely you agree that the coupled inductor gives a vchanging inductance even if there is no saturation.
Suppose one half is on load and the other is not............(ie i refer to the halfs of the split power supply output)
...then the one that is on load flows current through the coupled inductor........that will put a flux in the core that is in the direction such as to make the coupled inductor look much higher inductance to the one thats on light load.
so i say to people, that the coupled inductor, as used at the output of split output, is not a constant inductance, but varies with the loading of each half of the split output.
-this is not a good thing...since we would ideally prefer to have the output inductors the same value all the time.....ie a low value which would give fast transient response.
Flux is irrelevant. Only flux change per current change accounts.
Effective inductance varies between 2 values: L, when only one side is loaded, and 2*L, when both rails are loaded. Designing controller for L...2*L can't be a problem.
Do you prefer a completely unstable SMPS?
Or do you prefer to build two independent SMPSs for the two rails?
Ideally we needed two independent SMPSs, with synchro rectification (against supply pumping). Do you want ideal solution, or optimal?
Coupled output inductors have proven to have a lot of advantages.
We have found that the transient response of out PSU is still excellent after this change and, anyway, in practical audio use you won't have a lot of load difference from one rail to the other (averaged over time), only small differences in the tens of mA range.
We have found that the transient response of out PSU is still excellent after this change and, anyway, in practical audio use you won't have a lot of load difference from one rail to the other (averaged over time), only small differences in the tens of mA range.
- Status
- Not open for further replies.