Re: Power supply build quality pictorial. part 2
DPS-400MB-1A Rev: 02M 400W IBM server power supply
The unit has four +12V rails. Manufactured in September of 2007 according to the datecode on the PCB (0738).
Fan is AFB0812VH, appears to have been manufactured during the 2nd week of September, 2007, by way of what I can garner from the datecode on the label. Doesn't spin fast at all during low loads but hey, it's a Delta, so that low speed makes for a quiet fan at least (doesn't move too much air at such a speed, though). Capacitors are all Japanese. 330uF 450V Nichicon GU (30x50) on the input, Nichicon PW, Rubycon YXG, and Chemi-con KMG on the output filtering. As you can well see, the underside is very clean, but not without the SMD components Delta likes to use there. Bridge rectifier is D15XB60 on a finless heatsink. Double forward topology with APFC. Main switchers are 2SK2915 and APFC transistors are FQPF3N50C. ST UC3843B and Texas Instruments LM393P are also present on the primary side for the PWM controllers, and LM393A is present on the underside. Main transformer is size 39. Input filtering... at least four Y capacitors, at least two X capacitors, one NTC thermistor, a few coils (on the sister board on the primary), but can't seem to find any MOVs in the unit. Maybe that's a good thing, though. There's another tiny transformer on the primary side... perhaps for driving the MOSFET on the high side? This unit has as BIG APFC coil and four Vishay optocouplers. The startup capacitors appear to be 100uF 50V Rubycon YXG (8x11.5) and 47uF 50V YXG (6.3x11).
+5VSB appears to be rectified by a freestanding 5A diode. 12.5x31.5 3300uF 10V Nichicon PW before the ferrite coil with 8x20 1000uF 10V Rubycon YXG after for filtering. +5VSB PWM chip appears to be a proprietary Infineon part with a VERY long part number that I can find no datasheet or info on, and there appears to be another similar chip on the primary side too. +3.3V is rectified by a single STPS3045CT and has a 2200uF 6.3V 10x23 Rubycon YXG before the ferrite coil with a 2200uF 10V 10x28 Rubycon YXG after. -12V appears to use a single heatsinkless diode for rectification and has a 220uF 50V YXG (10x16) before the -12V linear regulator (L7912CV on the secondary heatsink) with a 100uF 50V (8x11.5) YXG after. +5V is rectified by STPS3045CT on the high side and STPS3045CW on the low side and has a 2200uF 10V 10x31.5 Nichicon PW before and after the ferrite coil for filtering (the Nichicons in this unit were produced in Ohno, Japan, and the Rubycons in Singapore and Japan, all with 2007 datecodes... the Chemi-cons in the unit have 2007 datecodes as well.. and they all have polyester sleeves).
+12V uses STPS20H100CT on the high side and STPS30H100CW on the low side for rectification with a 12.5x35.5 3300uF 16V Nichicon PW in parallel with a 10x20 1000uF 16V Rubycon YXG for filtering (no ferrite coil). There are some tiny 100uF 25V Nichicon PWs scattered on the secondary daughterboard and some very small Chemi-con KMGs for decoupling. +12V rail has two shunts. There appears to be two ST1002DS controllers on the secondary daughterboard. The heatsinks have a very thick base and are very nice. +3.3V toroid uses -26 material, the massive +5V toroid uses -52 material. The unit has one 24-pin connector, one 4-pin ATX12V connector, five molex connectors, four SATA connectors, one floppy connector, and one 5-pin 12V connector. Also has a nice honeycomb stamp for ventilation and rubber gromets holding the fan in. It has a fan header as well on the daughter board on the secondary, to which the fan's connector is glued. The PSU also has a +3.3V sense wire. The wires are all 18 gauge, excepting the -12V, +3.3V sense, floppy, PS-ON, and PS-GOOD wires, which are all 22 gauge. No rocker switch on the PSU. PS-ON appears to be managed through a 431 shunt on the secondary. There are lights on the unit to indicate operation. I think that's about it.
DPS-400MB-1A Rev: 02M 400W IBM server power supply
The unit has four +12V rails. Manufactured in September of 2007 according to the datecode on the PCB (0738).
Fan is AFB0812VH, appears to have been manufactured during the 2nd week of September, 2007, by way of what I can garner from the datecode on the label. Doesn't spin fast at all during low loads but hey, it's a Delta, so that low speed makes for a quiet fan at least (doesn't move too much air at such a speed, though). Capacitors are all Japanese. 330uF 450V Nichicon GU (30x50) on the input, Nichicon PW, Rubycon YXG, and Chemi-con KMG on the output filtering. As you can well see, the underside is very clean, but not without the SMD components Delta likes to use there. Bridge rectifier is D15XB60 on a finless heatsink. Double forward topology with APFC. Main switchers are 2SK2915 and APFC transistors are FQPF3N50C. ST UC3843B and Texas Instruments LM393P are also present on the primary side for the PWM controllers, and LM393A is present on the underside. Main transformer is size 39. Input filtering... at least four Y capacitors, at least two X capacitors, one NTC thermistor, a few coils (on the sister board on the primary), but can't seem to find any MOVs in the unit. Maybe that's a good thing, though. There's another tiny transformer on the primary side... perhaps for driving the MOSFET on the high side? This unit has as BIG APFC coil and four Vishay optocouplers. The startup capacitors appear to be 100uF 50V Rubycon YXG (8x11.5) and 47uF 50V YXG (6.3x11). +5VSB appears to be rectified by a freestanding 5A diode. 12.5x31.5 3300uF 10V Nichicon PW before the ferrite coil with 8x20 1000uF 10V Rubycon YXG after for filtering. +5VSB PWM chip appears to be a proprietary Infineon part with a VERY long part number that I can find no datasheet or info on, and there appears to be another similar chip on the primary side too. +3.3V is rectified by a single STPS3045CT and has a 2200uF 6.3V 10x23 Rubycon YXG before the ferrite coil with a 2200uF 10V 10x28 Rubycon YXG after. -12V appears to use a single heatsinkless diode for rectification and has a 220uF 50V YXG (10x16) before the -12V linear regulator (L7912CV on the secondary heatsink) with a 100uF 50V (8x11.5) YXG after. +5V is rectified by STPS3045CT on the high side and STPS3045CW on the low side and has a 2200uF 10V 10x31.5 Nichicon PW before and after the ferrite coil for filtering (the Nichicons in this unit were produced in Ohno, Japan, and the Rubycons in Singapore and Japan, all with 2007 datecodes... the Chemi-cons in the unit have 2007 datecodes as well.. and they all have polyester sleeves).
+12V uses STPS20H100CT on the high side and STPS30H100CW on the low side for rectification with a 12.5x35.5 3300uF 16V Nichicon PW in parallel with a 10x20 1000uF 16V Rubycon YXG for filtering (no ferrite coil). There are some tiny 100uF 25V Nichicon PWs scattered on the secondary daughterboard and some very small Chemi-con KMGs for decoupling. +12V rail has two shunts. There appears to be two ST1002DS controllers on the secondary daughterboard. The heatsinks have a very thick base and are very nice. +3.3V toroid uses -26 material, the massive +5V toroid uses -52 material. The unit has one 24-pin connector, one 4-pin ATX12V connector, five molex connectors, four SATA connectors, one floppy connector, and one 5-pin 12V connector. Also has a nice honeycomb stamp for ventilation and rubber gromets holding the fan in. It has a fan header as well on the daughter board on the secondary, to which the fan's connector is glued. The PSU also has a +3.3V sense wire. The wires are all 18 gauge, excepting the -12V, +3.3V sense, floppy, PS-ON, and PS-GOOD wires, which are all 22 gauge. No rocker switch on the PSU. PS-ON appears to be managed through a 431 shunt on the secondary. There are lights on the unit to indicate operation. I think that's about it.
Attached Files
Half of the computer problems is caused by bad contacts 
)
.
Comment