eem2am
Banned
forward converter as led drive
Hello,
I am designing an offline one-transistor-forward-converter for 10W LED drive.
I obviously need to have a demagnetisation winding in my transformer.
However, this winding, of course, does not discharge the energy in the transformer’s leakage inductance.
In order to discharge the energy in the transformer’s leakage inductance, I must use an RC snubber.
-I must put this snubber either across the transformer primary, or across the FET.
However, when I use such a snubber, the snubber ends up doing the job of demagnetising the transformer, AND discharging the transformer’s leakage inductance.
-So the problem is that the snubber “burns” up the energy of the demagnetisation current instead of returning it to the input capacitors.
This makes the converter extremely inefficient.
Do readers know how I can get around this problem.?
I appreciate that a two-transistor-forward-converter would mitigate this problem, though the component count becomes extreme with the high side pulse transformer gate drive.
I know that bootstrap high side drivers exist, such as IR2301 –however, these cannot work with two-transistor-forward-converters, since the high side FET’s source connection is not grounded by the lower FET during the high side FET’s off time.
-this in turn means that the high side driver’s bootstrap capacitor simply cannot re-charge and cannot therefore drive the high-side FET.
“Bootstrap” high-side drivers only work for half or full bridge set-ups. ?
Hello,
I am designing an offline one-transistor-forward-converter for 10W LED drive.
I obviously need to have a demagnetisation winding in my transformer.
However, this winding, of course, does not discharge the energy in the transformer’s leakage inductance.
In order to discharge the energy in the transformer’s leakage inductance, I must use an RC snubber.
-I must put this snubber either across the transformer primary, or across the FET.
However, when I use such a snubber, the snubber ends up doing the job of demagnetising the transformer, AND discharging the transformer’s leakage inductance.
-So the problem is that the snubber “burns” up the energy of the demagnetisation current instead of returning it to the input capacitors.
This makes the converter extremely inefficient.
Do readers know how I can get around this problem.?
I appreciate that a two-transistor-forward-converter would mitigate this problem, though the component count becomes extreme with the high side pulse transformer gate drive.
I know that bootstrap high side drivers exist, such as IR2301 –however, these cannot work with two-transistor-forward-converters, since the high side FET’s source connection is not grounded by the lower FET during the high side FET’s off time.
-this in turn means that the high side driver’s bootstrap capacitor simply cannot re-charge and cannot therefore drive the high-side FET.
“Bootstrap” high-side drivers only work for half or full bridge set-ups. ?