Power supply build quality pictorial. part 2

[Behemot]

Re: Power supply build quality pictorial. part 2

Oh yea, didn't even look at the upper part before. The platform is very similar to Xhy though, wonder if they licenced it or just copied it.

Can see that now, HEC ATX500TD-M, is da a new platform? Haven't seen it around.

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

I like the Rosewill psu, definitely worth recapping! And keep the Teapos to recap a cheaper psu.

[pdavid]

Delta DPS-200EB-1H

Heres an old AT unit from 1990 made by Delta.( Yup, another vintage psu... recently I dug out a lot of these)

This unit has great EMI input filtering. Lacks transient supression parts but look at all the filtering coils and capacitors plus EMI socket. The primary bulk caps are 1000uf each! 8 amp bridge diode...

The topology is half bridge. Two 2SC3164 trasistors controlled by a TL494 thats on a seperate board.

The power supply is rated 220W, but it may be able to deliver mode. 5V rail is rectified by two S30SC4M diodes. The 12V rail is taken care of by a single 12KC20 diode. All output diodes have ferrite beads on their leads.
The soldering is clean, near perfect. Primary and secondary caps are by Marcon.
The unit works, but the Delta fan makes quite a noise.

[Dan81]

Re: Delta DPS-200EB-1H

Heres an old AT unit from 1990 made by Delta.( Yup, another vintage psu... recently I dug out a lot of these)

This unit has great EMI input filtering. Lacks transient supression parts but look at all the filtering coils and capacitors plus EMI socket. The primary bulk caps are 1000uf each! 8 amp bridge diode...

The topology is half bridge. Two 2SC3164 trasistors controlled by a TL494 thats on a seperate board.

The power supply is rated 220W, but it may be able to deliver mode. 5V rail is rectified by two S30SC4M diodes. The 12V rail is taken care of by a single 12KC20 diode. All output diodes have ferrite beads on their leads.
The soldering is clean, near perfect. Primary and secondary caps are by Marcon.
The unit works, but the Delta fan makes quite a noise.

Funny thing,that grille in the first picture reminds me of Okia PSUs.

Speaking of which,aren't some of them based on the CWT ISO platform? Why did they not pay a bit more to have a decent line based on CWT ISO? I have a G-Alantic PSU that looks as generic as possible outside and it actually is decent inside. And it uses the CWT ISO platform. (with a ERL-35 transformer)

[pdavid]

Re: Power supply build quality pictorial. part 2

I'v only seen some guttles low wattage Okia units with 8cm fan. This dense small diameter grille style is something I've seen on some random server psu and older Chieftec units too, very random.

[momaka]

Re: Power supply build quality pictorial. part 2: Mitac SPU-75

I guess while on the topic of old AT power supplies...

Here is a Mitac SPU-75 AT PSU. It came out of a Compaq PC. Not sure what model. Possibly Presario 4000 and 5000 series based on my research online (old socket 7 stuff).

Anyways see attached pictures below. The PSU uses flyback topology (i.e. what is often used in your typical 17"-21" LCD monitor). Hence, I think the 75 Watt rating is pretty honest. Overall, the PSU looks well-made with very clean soldering. I've never heard of Mitac before (no UL number on the label either), but seems they are (or were?) a legit company based on research. Perhaps a Delta subdivision? (Based on the clean work.)

Detailed Specs:

PRIMARY SIDE ====

* 3.15 A ceramic fuse in holder + NTC thermistor
* six 4.7 nF Y2 safety-rated caps
* one 0.47 uF X2 safety-rated cap before common-mode choke
* one common-mode choke, EE core, size 19 (?)
* one 0.1 uF X2 safety-rated cap after CM choke
* KBP206 bridge rectifier
* primary filter cap: one Hitachi HP3 400V, 330 uF, 25 x 50 mm (dia. x height), 85C
* one FS10KM-12 MOSFET (N-ch, 10 Amps continuous, 40 Watt Pd, TO-220FN case) for main PS
* flyback topology
* ST UC3842 PWM controller + BT169d SCR + 4N35 optocoupler for feedback
* size 33 (?) EE core main transformer


SECONDARY SIDE ====

voltage regulation: 431 shunt + BT169d SCR
power good generator: M51957b
NO PROTECTIONS OF ANY KIND (except under-voltage, which = loss of PG signal)

5V rail:
* two Rubycon YXB, 10 V, 3300 uF, 12.5 x 25 mm caps before PI coil
* one Lelon (can't read specs), 12.5 x 20 mm 105C cap after PI coil
* 18-turn, 5 mm core PI coil
* one PBYR2040CT (20 A, 40V) schottky rectifier, TO-220 case
* minimum load resistor: two 47-Ohm, 3 Watt resistors in parallel

12V rail:
* one Lelon CE, 16 V, 2200 uF, 12.5 x 25 mm, 105C cap before PI coil
* one Lelon CE, 16 V, 1000 uF, 10 x 20 mm, 105C cap after PI coil
* 18-turn, 5 mm core PI coil
* one rectifier (can't read part number) in TO-220 case
* minimum load resistor: one 150-Ohm, 3 Watt resistor

-12V rail:
* rectifier: HER103 diode (raw -12V rail output)
* one Lelon CE, 35 V, 220 uF, 10 x 13 mm, 85C cap filtering the raw output
* one ST L7912CV linear regulator (TO-220 case, no heatsink)
* one Lelon CE, 16 V, 100 uF, 6.3 x 11 mm, 85C cap filtering 7912 regulator
* minimum load resitor: one 5.1-KOhm, 1/8 Watt resistor

-5V rail:
* "raw" input for -5V rail is -12V rail output
* one ST L7905CV linear regulator (TO-220 case, no heatsink)
* one Lelon CE, 16 V, 100 uF, 6.3 x 11 mm, 85C cap filtering 7905 regulator
* minimum load resitor: one 1-KOhm, 1/8 Watt resistor


OTHER: ====

* Input wires: 18 AWG, 300 V
* Output wires: 18 AWG, 300 V
* Connectors: AT pair (for motherboard), 3x 4-pin standard drive connectors, 2x floppy
* Fan: Delta Electronics AFB0612L, 12 V, 0.09 Amp, 60 mm, used as exhaust and attached to a speed controller
* Ratings: 5 V @ 11.5 Amps, 12 V @ 1.5 Amps, -5 V @ 0.1 Amps, -12 V @ 0.12 Amps

[Wester547]

Re: Power supply build quality pictorial. part 2

^ Nice power supply momaka. Mitac is the company who produced the "Bluford" model motherboards for the Dell Dimension 2300/2350/2400 series PCs. The ones that would often have bad Hermei 470uF 6.3V caps or 2200uF 6.3V Nichicon HMs with 2002 datecodes in the VRM output, and occassionally bad 1500uF 16V KZGs in the VRM input. Never seen a Lelon with a honeycomb vent until now either, although it seems to be a popular vent among many a brand (Hermei, Choyo, Jamicon, Fuhyin, even Sacon, etc...). I also spot some Taicons among the Marcon capacitors in pdavid's old Delta.

...

And now another retro PSU... a Win-Tact Electronics Corp. WP603A12 300W PSU, made in Taiwan R.O.C., manufactured in the 13th week of 2000 (late March or early April 2000). Totally comprimised of 18 gauge connectors, this +5V heavy PSU lacks an ATX +12V connector and sports two 20-pin ATX connectors. I acquired this PSU a few months ago and it came with a rack that was never used before - its zip ties were still intact (before the PSU was removed from the rack) which suggests the PSU was never used prior to acquiring it. The primary capacitors are 22x50 820uF 200V Teapo LXKs (datecode "9J09" printed atop the plastic caps - week 9 1999 perhaps? Or September 9th 1999?). The primary switchers are 2SC3320s (yes it's half-bridge topology) and the standby switcher appears to be 2SK320 but the text on the plastic package was rather faded so it was hard to tell. The bridge rectifier is a Taiwan Semiconductor KBU805 bridge without a heatsink. For input filtering it has two toroidal coils, three Y capacitors, two Teapo (-40C to +85C rated) X capacitors, two MOVs, etc. The primary switchers are biased by two Sanyo OS-CON SC 2.2uF 50V solid capacitors (which use an organic semiconductor material known as TCNQ complex salt, a type of charge transfer complex, rather than conductive functional polymer). A coupling film capacitor to also keep the BJTs balanced is present on the primary side as to be expected. The main transformer appears to be size 35 (made week 5 2000), and the feedback and flyback transformers much smaller. The single optocoupler present on the primary side is made by SHARP.

Moving on to the secondary side, we have an intermingling of Teapo SX, Teapo SEK, Su'scon SD, and Su'scon SK capacitors (all have 1999-2000 datecodes). The +5VSB rail is rectified by a single freestanding FR302 and filtered by a 10x20 Teapo SX 470uF 25V (with a honeycomb style vent, batch "A8" rather than "A3") before the ferrite filtering coil with a 470uF 16V 10x16 Su'scon SD after (also seems to have a 9k ohms minimum load resistor with a 5% tolerance). The secondary supervisor chips are an Enhance LS-168A and Texas Instruments TL494CN. The main toroid uses -52 material and the +3.3V toroid -26 material. Also present is the extra toroid used to derive the +3.3V transformer pins from the +5V pins. The +3.3V rail is rectified by a single MBR3060PT by Lite-on semiconductor (datecode week 49 of 1998), and the +5V rail MBR6040PT by Lite-on semicondutor (datecode week 7 of 2000). The +12V rail has a single Philips BYV32E-200 (manufactured week 48 1998, or late November 1998). +3.3V is filtered by a single 2200uF 6.3V 10x30 Teapo SX before the filtering coil with a 3300uF 10V 12.5x25 Su'scon SD after, and also bears a 90 ohms (5% tolerance) minimum load resistor and an extra sense wire. +5V is filtered by the same arrangement, and +12V has a sole 12.5x30 2200uF 16V Teapo SC and a filtering coil (whatever good that will do with only one capacitor...). -12V is rectified by an axial FR104 and filtered by another 470uF 25V 10x20 Teapo SX (this time with the regular Teapo vent stamp, "A3") along with a filtering coil, with a Motorola MC7905CT on an extra heatsink linear regulating -5V from -12V, and a 22uF 16V Teapo SEK after the regulator.

Really overbuilt considering the ratings on the label: +5V rated for 30A, +3.3V for 22A, +12V for 10A, -12V for 0.8A and -5V for 0.5A. No +5V/+3.3V combined rating exists on the label. This PSU has ample ventilation on the back and a very nice wire grille on the front, with an ADDA AD0812HX-A70GL fan (made in March of 2000, has 24 gauge wires), whose shaft is still very wet after all these years, although it has never been used before. The PSU has a pretty aggressive controller and the fan spins rather fast (around 2000RPM) by default. It's not too loud for the speed at which it spins. Of a lesser pleasant note is the conductive snot about the supply (which hasn't blackened yet)... but needless to say the unit works fine with its voltages in spec. Haven't tested it under heavy load as yet. The heatsinks aren't the greatest but are hardly bad given the primary and secondary silicon.

[Compgeke]

Re: Power supply build quality pictorial. part 2

Mitac was actually pretty big for OEM cases in the 90s. Almost all the 90s Compaqs in the round fashion (like the 5200s), the case used for the Dell Dimension XPS and early Micron Milienia towers, the later round Micron towers, etc. All Mitac cases.

[Th3_uN1Qu3]

Re: Power supply build quality pictorial. part 2

Mitac also made some laptops in the P3-Pentium M-P4-Core Duo-C2D era, don't know if they still make them. They generally OEM'd for some of the lesser known brands.

[Wester547]

Re: Power supply build quality pictorial. part 2

As an addendum and a correction, I meant Teapo SX for the +12V rail and not SC.

[Wester547]

Re: Power supply build quality pictorial. part 2

Sorry for the double post, but also forgot to mention that +5VSB is rated for 1A in the Win-tact. Also had the +3.3V and main toroid's material mixed up (+3.3V uses -52 material and the main toroid -26 material, which is somewhat worse).

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

What's that stuff mounted to the fan?
What's on there??

And do you have pics of that thingy?

[Wester547]

Re: Power supply build quality pictorial. part 2

Are you referencing the attached image?

That's an extra sister PCB. It's mounted to the fan with what appears to be metal grommets? of sorts. On it rests some red dipped film coupling capacitors (and what appears to be one small film capacitor), an STMicro triac of sorts (BTA16600B, a "glass passivated PNPN device"). There's also an LM393N comparator by OnSemi, some axial resistors (and one big resistor), some small electrolytics by Teapo and Su'scon (SEK and SK series), some axial diodes, two potentiometers, and some TL431 shunt regulators. It's connected to the primary and secondary side through those two connectors. And there are actually five Y capacitors on the primary. The PSU also has a two transitor +5VSB circuit, but no critical cap fortunately, as well as six molex connectors and two berg (floppy) connectors.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

And where is the fan connected to?

Seems like it has something to do with +5VSB regulation and some additional protection...

[Wester547]

Re: Power supply build quality pictorial. part 2

The fan is connected to a header on the secondary side (not directly connected to +12V). There's a thermistor on the secondary heatsink to regulate the fan speed (although it seems to be a bit aggressive by default).

[PeteS in CA]

Re: Delta DPS-200EB-1H

Heres an old AT unit from 1990 made by Delta.( Yup, another vintage psu... recently I dug out a lot of these)

This unit has great EMI input filtering. Lacks transient supression parts but look at all the filtering coils and capacitors plus EMI socket. The primary bulk caps are 1000uf each! 8 amp bridge diode...

The topology is half bridge. Two 2SC3164 trasistors controlled by a TL494 thats on a seperate board.

The power supply is rated 220W, but it may be able to deliver mode. 5V rail is rectified by two S30SC4M diodes. The 12V rail is taken care of by a single 12KC20 diode. All output diodes have ferrite beads on their leads.
The soldering is clean, near perfect. Primary and secondary caps are by Marcon.
The unit works, but the Delta fan makes quite a noise.

Interesting. Marcon was good stuff, bought by Nippon Chemicon. A fan that old has "earned" the "right" to be noisy. Looks like the PCB has positions for surge protection components. Interesting derating: +12V rated for 8A, but the rectifier for 12A; +5V rated for 23A, but two 30A rectifiers paralleled. Potential power O/P above the rating is probably dependent on cooling and the rating of the switch transistors (max. current, beta at operating current, switching speed vs. switch frequency).

[PeteS in CA]

Re: Power supply build quality pictorial. part 2: Mitac SPU-75

I guess while on the topic of old AT power supplies...

Here is a Mitac SPU-75 AT PSU. It came out of a Compaq PC. Not sure what model. Possibly Presario 4000 and 5000 series based on my research online (old socket 7 stuff).

Anyways see attached pictures below. The PSU uses flyback topology (i.e. what is often used in your typical 17"-21" LCD monitor). Hence, I think the 75 Watt rating is pretty honest. Overall, the PSU looks well-made with very clean soldering. I've never heard of Mitac before (no UL number on the label either), but seems they are (or were?) a legit company based on research. Perhaps a Delta subdivision? (Based on the clean work.)

Detailed Specs:

PRIMARY SIDE ====

* 3.15 A ceramic fuse in holder + NTC thermistor
* six 4.7 nF Y2 safety-rated caps
* one 0.47 uF X2 safety-rated cap before common-mode choke
* one common-mode choke, EE core, size 19 (?)
* one 0.1 uF X2 safety-rated cap after CM choke
* KBP206 bridge rectifier
* primary filter cap: one Hitachi HP3 400V, 330 uF, 25 x 50 mm (dia. x height), 85C
* one FS10KM-12 MOSFET (N-ch, 10 Amps continuous, 40 Watt Pd, TO-220FN case) for main PS
* flyback topology
* ST UC3842 PWM controller + BT169d SCR + 4N35 optocoupler for feedback
* size 33 (?) EE core main transformer


SECONDARY SIDE ====

voltage regulation: 431 shunt + BT169d SCR
power good generator: M51957b
NO PROTECTIONS OF ANY KIND (except under-voltage, which = loss of PG signal)

5V rail:
* two Rubycon YXB, 10 V, 3300 uF, 12.5 x 25 mm caps before PI coil
* one Lelon (can't read specs), 12.5 x 20 mm 105C cap after PI coil
* 18-turn, 5 mm core PI coil
* one PBYR2040CT (20 A, 40V) schottky rectifier, TO-220 case
* minimum load resistor: two 47-Ohm, 3 Watt resistors in parallel

12V rail:
* one Lelon CE, 16 V, 2200 uF, 12.5 x 25 mm, 105C cap before PI coil
* one Lelon CE, 16 V, 1000 uF, 10 x 20 mm, 105C cap after PI coil
* 18-turn, 5 mm core PI coil
* one rectifier (can't read part number) in TO-220 case
* minimum load resistor: one 150-Ohm, 3 Watt resistor

-12V rail:
* rectifier: HER103 diode (raw -12V rail output)
* one Lelon CE, 35 V, 220 uF, 10 x 13 mm, 85C cap filtering the raw output
* one ST L7912CV linear regulator (TO-220 case, no heatsink)
* one Lelon CE, 16 V, 100 uF, 6.3 x 11 mm, 85C cap filtering 7912 regulator
* minimum load resitor: one 5.1-KOhm, 1/8 Watt resistor

-5V rail:
* "raw" input for -5V rail is -12V rail output
* one ST L7905CV linear regulator (TO-220 case, no heatsink)
* one Lelon CE, 16 V, 100 uF, 6.3 x 11 mm, 85C cap filtering 7905 regulator
* minimum load resitor: one 1-KOhm, 1/8 Watt resistor


OTHER: ====

* Input wires: 18 AWG, 300 V
* Output wires: 18 AWG, 300 V
* Connectors: AT pair (for motherboard), 3x 4-pin standard drive connectors, 2x floppy
* Fan: Delta Electronics AFB0612L, 12 V, 0.09 Amp, 60 mm, used as exhaust and attached to a speed controller
* Ratings: 5 V @ 11.5 Amps, 12 V @ 1.5 Amps, -5 V @ 0.1 Amps, -12 V @ 0.12 Amps

Mitac was/is a monitors manufacturer and contract manufacturer. Decent stuff.

Concur that this is a flyback design - no O/P inductor (the main transformer primary is a coupled inductor).

The 10A rating of the switch MOSFET is the result of picking a part with low R(D-S On).

I'm surprised they used Lelon - for whom I have low esteem - for the +12V O/P caps. The flyback topology needs really good O/P lytics (e.g. the Rubycons on the +5V O/P).

Nice conservative derating for the O/P rectifiers. The rectifier heatsink looks like heavy metal goodness.

[PeteS in CA]

Re: Power supply build quality pictorial. part 2

^ Nice power supply momaka. Mitac is the company who produced the "Bluford" model motherboards for the Dell Dimension 2300/2350/2400 series PCs. The ones that would often have bad Hermei 470uF 6.3V caps or 2200uF 6.3V Nichicon HMs with 2002 datecodes in the VRM output, and occassionally bad 1500uF 16V KZGs in the VRM input. Never seen a Lelon with a honeycomb vent until now either, although it seems to be a popular vent among many a brand (Hermei, Choyo, Jamicon, Fuhyin, even Sacon, etc...). I also spot some Taicons among the Marcon capacitors in pdavid's old Delta.

...

And now another retro PSU... a Win-Tact Electronics Corp. WP603A12 300W PSU, made in Taiwan R.O.C., manufactured in the 13th week of 2000 (late March or early April 2000). Totally comprimised of 18 gauge connectors, this +5V heavy PSU lacks an ATX +12V connector and sports two 20-pin ATX connectors. I acquired this PSU a few months ago and it came with a rack that was never used before - its zip ties were still intact (before the PSU was removed from the rack) which suggests the PSU was never used prior to acquiring it. The primary capacitors are 22x50 820uF 200V Teapo LXKs (datecode "9J09" printed atop the plastic caps - week 9 1999 perhaps? Or September 9th 1999?). The primary switchers are 2SC3320s (yes it's half-bridge topology) and the standby switcher appears to be 2SK320 but the text on the plastic package was rather faded so it was hard to tell. The bridge rectifier is a Taiwan Semiconductor KBU805 bridge without a heatsink. For input filtering it has two toroidal coils, three Y capacitors, two Teapo (-40C to +85C rated) X capacitors, two MOVs, etc. The primary switchers are biased by two Sanyo OS-CON SC 2.2uF 50V solid capacitors (which use an organic semiconductor material known as TCNQ complex salt, a type of charge transfer complex, rather than conductive functional polymer). A coupling film capacitor to also keep the BJTs balanced is present on the primary side as to be expected. The main transformer appears to be size 35 (made week 5 2000), and the feedback and flyback transformers much smaller. The single optocoupler present on the primary side is made by SHARP.

Moving on to the secondary side, we have an intermingling of Teapo SX, Teapo SEK, Su'scon SD, and Su'scon SK capacitors (all have 1999-2000 datecodes). The +5VSB rail is rectified by a single freestanding FR302 and filtered by a 10x20 Teapo SX 470uF 25V (with a honeycomb style vent, batch "A8" rather than "A3") before the ferrite filtering coil with a 470uF 16V 10x16 Su'scon SD after (also seems to have a 9k ohms minimum load resistor with a 5% tolerance). The secondary supervisor chips are an Enhance LS-168A and Texas Instruments TL494CN. The main toroid uses -52 material and the +3.3V toroid -26 material. Also present is the extra toroid used to derive the +3.3V transformer pins from the +5V pins. The +3.3V rail is rectified by a single MBR3060PT by Lite-on semiconductor (datecode week 49 of 1998), and the +5V rail MBR6040PT by Lite-on semicondutor (datecode week 7 of 2000). The +12V rail has a single Philips BYV32E-200 (manufactured week 48 1998, or late November 1998). +3.3V is filtered by a single 2200uF 6.3V 10x30 Teapo SX before the filtering coil with a 3300uF 10V 12.5x25 Su'scon SD after, and also bears a 90 ohms (5% tolerance) minimum load resistor and an extra sense wire. +5V is filtered by the same arrangement, and +12V has a sole 12.5x30 2200uF 16V Teapo SC and a filtering coil (whatever good that will do with only one capacitor...). -12V is rectified by an axial FR104 and filtered by another 470uF 25V 10x20 Teapo SX (this time with the regular Teapo vent stamp, "A3") along with a filtering coil, with a Motorola MC7905CT on an extra heatsink linear regulating -5V from -12V, and a 22uF 16V Teapo SEK after the regulator.

Really overbuilt considering the ratings on the label: +5V rated for 30A, +3.3V for 22A, +12V for 10A, -12V for 0.8A and -5V for 0.5A. No +5V/+3.3V combined rating exists on the label. This PSU has ample ventilation on the back and a very nice wire grille on the front, with an ADDA AD0812HX-A70GL fan (made in March of 2000, has 24 gauge wires), whose shaft is still very wet after all these years, although it has never been used before. The PSU has a pretty aggressive controller and the fan spins rather fast (around 2000RPM) by default. It's not too loud for the speed at which it spins. Of a lesser pleasant note is the conductive snot about the supply (which hasn't blackened yet)... but needless to say the unit works fine with its voltages in spec. Haven't tested it under heavy load as yet. The heatsinks aren't the greatest but are hardly bad given the primary and secondary silicon.

Interesting. Ordinarily I'd say the heatsinks were marginal for the 300W power rating, but between the conservative ratings of the rectifiers - over-rated means lower forward voltage drop and lower dissipated power - and the faster fan speed, it could be OK.

Re the mix of core materials on the O/P inductors, -52 material is not necessary for use with the switch frequencies common with TL494-type PWMs (usually under 50KHz), but would be lower loss. I'm guessing -52 material was used to improve efficiency and/or to reduce heat in that area of the P/S.

Around 2000 was when power demands were exceeding the capability of the standard 20-pin ATX connector arrangement, and a newer, more capable, arrangement was not yet agreed on. So some idiosyncratic models were produced around that time frame.

[Wester547]

Re: Power supply build quality pictorial. part 2

Interesting. Ordinarily I'd say the heatsinks were marginal for the 300W power rating, but between the conservative ratings of the rectifiers - over-rated means lower forward voltage drop and lower dissipated power - and the faster fan speed, it could be OK.

Well, it has 820uF primaries and 2SC3320 BJT switchers as well - I think it can do 300W, especially if the fan is spinning at 3000RPM (full speed).

Re the mix of core materials on the O/P inductors, -52 material is not necessary for use with the switch frequencies common with TL494-type PWMs (usually under 50KHz), but would be lower loss. I'm guessing -52 material was used to improve efficiency and/or to reduce heat in that area of the P/S.

Well, what I did also notice is that the main toroid is quite a bit smaller than the +3.3V toroid, so it makes sense. The lower "ripple" (switching) frequency on the output could also be a reason for the somewhat elevated choices of capacitance on the output of the +3.3V and +5V rail (5500uF per rail).

Around 2000 was when power demands were exceeding the capability of the standard 20-pin ATX connector arrangement, and a newer, more capable, arrangement was not yet agreed on. So some idiosyncratic models were produced around that time frame.

True, I suppose this was also a reason for auxiliary power connectors.

I forgot to mention that the PSU does not have a voltage selector switch, so I'd guess that it uses not only wide range flyback topology but wide range half bridge topology as well (or something like that?). I also incorrectly identified the voltage rating on the Sanyo OS-CON SC BJT biasing capacitors - it's 25V and not 50V. Furthermore, the actual transformer datecodes marked into the back are "9951" (for the main transformer), "0009" (for the feedback transformer) and "0011" (for the flyback transformer). There are also no "A3" marked Teapos in this PSU despite using the associated vent stamp, although all the Teapo SEKs are from the "A2" batch.

[PeteS in CA]

Re: Power supply build quality pictorial. part 2

Well, it has 820uF primaries and 2SC3320 BJT switchers as well - I think it can do 300W, especially if the fan is spinning at 3000RPM (full speed).
...
I forgot to mention that the PSU does not have a voltage selector switch, so I'd guess that it uses not only wide range flyback topology but wide range half bridge topology as well (or something like that?). ...

I'm not familiar with the 2SC3320, but in Googling it I see its rated for 15A, pretty impressive compared to the 12A MJE13009 which could be adequate.

I wonder if that model was hard-wired for one voltage range or the other. Does the label say it's wide-range? I know STMicro produced an IC that sensed the I/P voltage and did the selection automatically, using a triac. That's another possibility.