Re: Power supply build quality pictorial. part 2
Geez! Forgot one even larger difference. MJE13009-clones (originally it was a Motorola part; not sure if On Semiconductor, who bought Mot's discretes, still makes the part) are bipolar transistors. TO-247/TO-3P switch devices in power supplies are likely to be MOSFETs.
Much, maybe most, of the power dissipation in a switch device - transistor or MOSFET - is during the switch time. During switching the device is conducting significant current at the same time there is significant voltage across the device. Power = (Voltage x Current). The slower the switch time, the more power dissipated during each transition; the greater the % switch time is of the entire cycle time, the higher the over all dissipation of the device.
A bipolar transistor switches far slower than does a MOSFET. Consequently, a transistor dissipates more power during a transition than does a MOSFET, assuming both are of similar current rating and properly driven. At the same switch frequency, a bipolar transistor will dissipate more power than will a MOSFET.
Looked at another way, if you know the maximum power power you can allow your switch device to dissipate and assuming the same output power, the maximum switch frequency at which a bipolar transistor can be operated is lower than the maximum frequency for a MOSFET. Why would one want a higher switch frequency? The magnetics for a given power level at 50KHz are smaller than for 200KHz, The P/S could be smaller, the heatsinks could be larger, or the resistance to airflow could be less (which allows considerable flexibility to the system designer). Also, filtering conducted noise at 200KHz tends to be less difficult than at 25KHz or 50KHz (BTDT, got a drawer full of T-shirts!)
Geez! Forgot one even larger difference. MJE13009-clones (originally it was a Motorola part; not sure if On Semiconductor, who bought Mot's discretes, still makes the part) are bipolar transistors. TO-247/TO-3P switch devices in power supplies are likely to be MOSFETs.
Much, maybe most, of the power dissipation in a switch device - transistor or MOSFET - is during the switch time. During switching the device is conducting significant current at the same time there is significant voltage across the device. Power = (Voltage x Current). The slower the switch time, the more power dissipated during each transition; the greater the % switch time is of the entire cycle time, the higher the over all dissipation of the device.
A bipolar transistor switches far slower than does a MOSFET. Consequently, a transistor dissipates more power during a transition than does a MOSFET, assuming both are of similar current rating and properly driven. At the same switch frequency, a bipolar transistor will dissipate more power than will a MOSFET.
Looked at another way, if you know the maximum power power you can allow your switch device to dissipate and assuming the same output power, the maximum switch frequency at which a bipolar transistor can be operated is lower than the maximum frequency for a MOSFET. Why would one want a higher switch frequency? The magnetics for a given power level at 50KHz are smaller than for 200KHz, The P/S could be smaller, the heatsinks could be larger, or the resistance to airflow could be less (which allows considerable flexibility to the system designer). Also, filtering conducted noise at 200KHz tends to be less difficult than at 25KHz or 50KHz (BTDT, got a drawer full of T-shirts!)
Half of the computer problems is caused by bad contacts 
which killed it fast I'm sure cause the efficiency has to be terrible.
Comment