Announcement

Collapse
No announcement yet.

Power supply build quality pictorial. part 2

Collapse
This is a sticky topic.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Re: Power supply build quality pictorial. part 2

    Originally posted by PeteS in CA View Post
    That Optiplex PSU is interesting, still 5V/3.3V heavy, but with a healthy 5VSB current capability (2A).
    A lot of late Pentium 4 -era PSUs were built like this - i.e. enough Amps on the 3.3V and 5V rail to still work with older 5V-heavy systems, but also enough Amps on the 12V rail for more modern PCs (typically 14-18 Amps for that time).

    The 250 Watt HiPro "tanks" from that era are my favorite. You can load them straight to their maximum 14-16 Amp limit on the 12V rail, and they won't say a peep. Not to mention the voltage regulation remains solid. Hence, they can still be used in a modern PC today - albeit one without a high-power graphics card. And the lack of SATA power connectors can be a bit of a bummer unless you have lots of Molex->SATA power adapters around.

    Comment


      Re: Power supply build quality pictorial. part 2

      If there are some vacant holes in the PCB, you can always add one more SATA cable taken from something else

      I myself do prefer the dual-12V rail units which could also support a single graphics with 6pin power connector.
      Last edited by Behemot; 06-10-2018, 06:24 AM.
      Less jewellery, more gold into electrotech industry! Half of the computer problems is caused by bad contacts

      Exclusive caps, meters and more!
      Hardware Insights - power supply reviews and more!

      Comment


        Allied AL-D500EXP (SL-8500BTX) ā€“ blown fuse andā€¦?

        Oh Deer!
        What have we here?!

        ... an Allied AL-D500EXP "500 Watt" power supply.

        I got this from my previous job last year. Came out of a customer’s machine with, if I remember correctly, an i5 CPU and possibly a high-end Radeon graphics card (it looked chunky, so an R9 2x0 maybe?) Plugging the PSU revealed 5VSB was not present, so I had a look through the vents, expecting blown caps. Instead I immediately spotted the primary fuse was blown. Further analysis revealed the primary BJTs were also short-circuited. One of them even had a crack in its body and its Base solder joint partially melted.

        Nevertheless, I decided to take this PSU with me with intention to repair it, as it otherwise looked like a fairly decent PSU, especially for an Allied unit. Moreover, it’s a classic half-bridge Allied with a 2012 “COTY” (“chip-of-the-year” ) PWM controller, so that means it should be fairly straight-forward to fix. Unlike cheaper Deer, L&C, and Allied units, this one has good output wiring and decent selection of output connectors: 1x 24-pin ATX (of course), 1x 4-pin 12V CPU, 1x 8-pin 12V CPU, 2x 6-pin PCI-E power (but coming from the same strand of wires ), 3x Molex, 4x SATA power, and 1x floppy. All output wires are rated 18 AWG, 300V, and with a valid UL number (i.e. not some cheap no-name wires). So for all of these reasons, I thought that maybe this PSU deserved a second chance. After all, it looked like it blew because it was overloaded. While rated for 500 Watts max, we all know that Deer/Allied/L&C labels should rarely be trusted. Speaking of which, here is that:
        https://www.badcaps.net/forum/attach...1&d=1529256227
        Apparently, this PSU has two separate 12V rails, according to the label. But we shall see about that (I am very doubtful, to say the least). Either way, I personally think the system it was powering was too much for it (it would have been okay if it was a true 500 Watt PSU, that’s for sure).

        Here is a top-shot picture:
        https://www.badcaps.net/forum/attach...1&d=1529256227
        I see decent-sized heatsinks (well, maybe not for 500 Watts, but decent anyways), a fan controller board, 35 mm core main transformer, all input filtering components installed, and two Anodia 820 uF primary caps. And actually, here is a better close-up shot of those caps, as well as the rest of the primary:
        https://www.badcaps.net/forum/attach...1&d=1529256227
        We also see a GPU808 bridge rectifier, and that the 5VSB and driver transformers are of good size, having 19 mm and 16 mm cores, respectively (the real ones, not the fake short ones). You can also see on the right that the 5VSB output rectifier (TO-126 device) is mounted on the secondary heatsink as well.

        Next, a picture from the front of the PSU:
        https://www.badcaps.net/forum/attach...1&d=1529256227
        Here, you can see the faint crack near the Base leg of transistor Q4. Speaking of the primary BJTs, they are a pair of 2SC3320 in a TO-3P case. But what I like even more about this PSU is that there is no APFC circuit nonsense to deal with. Only the 2-transistor 5VSB is a bit more concerning, since it has a critical cap. That said, the 5VSB circuits in these Deer/Allied PSUs are usually pretty stable, as long as all of the associated caps with that circuit are okay.

        Moving onto the secondary side:
        https://www.badcaps.net/forum/attach...1&d=1529256227
        https://www.badcaps.net/forum/attach...1&d=1529256227
        Each major rail (3.3V, 5V, 12V, and 5VSB) has two caps and a PI coil. The output toroid inductor for the 5V/12V rails is a bit on the small side, and its wire turns aren’t as thick as I’d like to see in a “500 Watt” PSU. However, it does have a light green core, meaning it is rated for higher frequency and lower loss. Nevertheless, I think this and the 35 mm core main transformer are the major limiting factors for this PSU, in terms of power output (probably 300-350 Watts max, if we had to be honest). On the other hand, the 3.3V output toroid inductor is almost as big as the 5V/12V toroid. It too has a lower-loss higher-frequency core. PWM chip is, of course, a “2012” COTY – I mean, could we really expect anything else in a Deer/Allied PSU?

        Of course, no PSU analysis is finished without a picture of the solder side, so here comes that next:
        https://www.badcaps.net/forum/attach...1&d=1529256227
        Remember the dual 12V rail rating on the label? Yep, that’s not true. Surprised?
        What I also like about looking at the solder side is that it allows me to see if the PSU has any hot spots – at least on green and yellow PCBs (I hate working on black/dark-colored PCBs for this reason). And this has hot spots, indeed – we can see that the NTC inrush thermistor (a Sensicom 5D-15) looks like it ran hot, along with the primary-side snubber resistor, 5V/12V rail rectifiers and possibly output toroid. On that note, I don’t understand why deer has two 30 Amp rectifiers (STPS3045CT) in parallel for the 5V rail. This PSU should be meant for a modern 12V PC, so it really doesn’t need that kind of 5V capacity! Meanwhile, the 5V/12V output toroid has two thick wires for the 12V rail and only one for the 5V rail. What the #$%^, Allied?!?! At least be consistent with the design. The 12V rectifier does have a 40 Amp -rated part, though.
        Also, there are no tiny SMD components, so nothing hard to decipher should anything else be blown.

        And last but not least two pictures of the case:
        https://www.badcaps.net/forum/attach...1&d=1529256227
        https://www.badcaps.net/forum/attach...1&d=1529256227
        Unit has a 120 mm fan and decent ventilation. Speaking of the fan, it’s a Shenzhen Xin Wang Xin Electronics Co., LTD., model XWX1225M12S, rated for 0.22 Amps @ 12V. Case is also of the clam-shell style, which is what allowed me to check the primary BJTs and a few other components very quickly without taking anything else apart – neat! The case steel is a bit thinner than on quality PSU, but not terribly thin.

        Anyways, I’ll put up a repair thread on this PSU sometime (but who knows when ). Meanwhile, here is a more detailed component breakdown.

        Primary Side
        Primary RF/EMI filter: 2x 2.2 nF, 2x 3.3 nF and 1x 4.7 nF safety Y-class caps, 3x safety X-class caps (can’t read values), 2x CM chokes
        T8AL 8 Amp glass fuse (blown), 2x MOVs (1x across each primary cap)
        GBU808 bridge rectifier, Sensicom 5D-15 inrush NTC thermistor
        2x Anodia PS, 200 V, 820 uF, 23 x 46 mm (Ćø x h)
        2x 2SC3320 NPN BJTs (TO-3P) for main PS (half-bridge)
        Main BJT driver circuit caps: 2x Yang Chun LE, 50V, 2.2 uF, 5 x 11 mm
        Snubber network, main PS: RC-type with 47-Ohm 5-Watt resistor and __ nF 1 kV ceramic cap
        Main Transformer h-bridge coupling cap: 2.0 uF, 250V metal-film
        1x JCS2N65FB N-ch MOSFET (TO-220MF) for 5VSB + 2N2222A NPN BJT (TO-92) as its driver (5VSB is 2-transistor self-oscillating with feedback).
        Critical 5VSB cap: Yang Chun LE, 50V, 47 uF, 6.3 x 11 mm
        Transformers: 35 mm core (main); 19 mm core (5VSB); 16 mm core (BJT driver)

        Secondary Side
        5V Rail:
        - 2x STPS3045CWC (TO-247) rectifiers in parallel
        - 2x Yang Chun LE, 10V, 2200 uF, 10 x 20 mm  with PI coil in between
        - 27-Ohm, 3-Watt resistor for minimum load (~1 Watt of power wasted as heat!)

        12V Rail:
        - 1x SBL4060P (TO-247) rectifier
        - 2x Yang Chun LE, 16V, 1000 uF, 10 x 20 mm  with PI coil in between
        - 110?-Ohm, 3-Watt resistor for minimum load (1.3 Watts of power wasted as heat!)

        3.3V Rail:
        - 1x STPS3045CWC (TO-247) rectifier
        - 2x Yang Chun LE, 10V, 2200 uF, 10 x 20 mm  with 5-turn, 5mm core, 14 AWG, PI coil in between
        - 10-Ohm, 3-Watt resistor for minimum load (1 Watt of power wasted as heat!)

        -12V Rail:
        - 2x UF1007 diodes (as rectifiers) + SR505??
        - 1x Yang Chun LE, 16V, 470 uF, 8 x 16 mm  with PI coil before it
        - 510-Ohm, Ā¼-Watt resistor for minimum load

        5VSB Rail:
        - 1x SR-540FC (5 Amp, 40V, TO-126 case) rectifier
        - 2x YC, 10V, 1000 uF, 8 x 16 mm + 1x 8 mm free spot  with PI coil in between
        - 470-Ohm, 1/8 -Watt resistor for minimum load
        Attached Files
        Last edited by momaka; 06-17-2018, 11:43 AM.

        Comment


          Re: Power supply build quality pictorial. part 2

          Wow what a quality Deer power supply!

          Lol the amount of sarcasm is enough to fill a

          Anodia caps, they are crappy, but 12 Kuang Jin and Yang Chun are just as crap...

          I have seen the same used in the Ultra LS they are ok for testing an old system but not for prolonged use. (although LSP was a little better built)

          I feel sorry for the components that had to share with the caps
          Yes i love Nichicon MUSE Audio Capacitors...they would look awesome all over any motherboard

          Comment


            Re: Power supply build quality pictorial. part 2

            Originally posted by coreAngel View Post
            Wow what a quality Deer power supply!

            Lol the amount of sarcasm is enough to fill a
            To quote Steve Lav. "Nothing but the best here."



            Originally posted by coreAngel View Post
            Anodia caps, they are crappy, but 12 Kuang Jin and Yang Chun are just as crap...
            "Anode-I-Ain't" caps...
            The YCs may as well be on Saturn- bet they're practically open circuit.

            Originally posted by coreAngel View Post
            I have seen the same used in the Ultra LS they are ok for testing an old system but not for prolonged use. (although LSP was a little better built)

            I feel sorry for the components that had to share with the caps
            Let's not forget the cooked Super-I/O and southbridge on the motherboard(s) that thing destroyed.

            Originally posted by momaka
            AL-D500EXP
            I know, I know! That's short for "Exp-lodes!"

            Not gonna comment on the other oddities- too early for "Deer" season.

            "pokemon go... to hell!"

            EOL it...
            Originally posted by shango066
            All style and no substance.
            Originally posted by smashstuff30
            guilty,guilty,guilty,guilty!
            guilty of being cheap-made!

            Comment


              Re: Power supply build quality pictorial. part 2

              Seventeam ST-220WT-P large AT unit.

              Standard half bridge topology from 1991.

              Proper EMI input filtering.
              2x470uF Nitsuka primary caps.
              2x13009 transistors.
              IR3M02 PWM driver
              ER35 main transformer.
              Seperate transformer for current sensing
              CEC PCE-TUR caps on secondary
              LM339 comparator on the secondary.
              5V rectified by S30C45.
              12V rectifiedby C33-02C.
              -12V rectified by FR205.
              -5V by L7905C voltage regulator from -12V output.

              AWG 18 wires on all output rails.

              I really like the circuit traces on his one.
              Attached Files

              Comment


                Re: Power supply build quality pictorial. part 2

                Originally posted by pdavid View Post
                Seventeam ST-220WT-P large AT unit.

                Standard half bridge topology from 1991.

                Proper EMI input filtering.
                2x470uF Nitsuka primary caps.
                2x13009 transistors.
                IR3M02 PWM driver
                ER35 main transformer.
                Seperate transformer for current sensing
                CEC PCE-TUR caps on secondary
                LM339 comparator on the secondary.
                5V rectified by S30C45.
                12V rectifiedby C33-02C.
                -12V rectified by FR205.
                -5V by L7905C voltage regulator from -12V output.

                AWG 18 wires on all output rails.

                I really like the circuit traces on his one.
                Appears to be a pretty good power supply, looks pretty old considering it's a 5v heavy unit and the fact it does not have a 3.3v rail (that's if it is a PC PSU like i think it is). The secondary capacitors will have to be replaced because i found an old PSU with the same caps and most of them were dead, one of the secondary capacitors on your board appears to be on it's way out too. But other than that, it appears to be pretty good.
                Last edited by RukyCon; 06-20-2018, 03:21 PM.
                I'm not a expert, I'm just doing my best.

                Comment


                  Re: Power supply build quality pictorial. part 2

                  The PCB is dated 15 May 1991, he did say it was an AT power supply (386/486 era) rather than ATX.

                  Comment


                    Re: Power supply build quality pictorial. part 2

                    Yeah, because lack of +5VSB Rail and also no 3,3V either.
                    Can't be ATX...

                    Comment


                      Re: Power supply build quality pictorial. part 2

                      Well I have other AT units from Seventeam that have 5V standby and 3.3V DC-DC plus PS-on wired to to the circuit to make it work like ATX. There were plenty of similar solutions in mid 90'. Some really horrible ones too.

                      Comment


                        Re: Power supply build quality pictorial. part 2

                        Originally posted by coreAngel View Post
                        Wow what a quality Deer power supply!
                        Well, in all seriousness... compared to other Deer/Allied units, this one is relatively good.

                        Originally posted by coreAngel View Post
                        Lol the amount of sarcasm is enough to fill a
                        Oh fine! Since you guys aren't satisfied, I'll post something else later this week that you might find a bit better.

                        Originally posted by kaboom View Post
                        The YCs may as well be on Saturn- bet they're practically open circuit.
                        Not these ones.
                        Looks like the fuse, primary BJTs, driver transistors, and 12V rectifier bit the dust before the output caps did. I'm not sure who took out who first, but this Deer is pretty dead right now

                        Originally posted by kaboom View Post
                        Let's not forget the cooked Super-I/O and southbridge on the motherboard(s) that thing destroyed.
                        Only if the 5VSB goes out of control. But I find that this doesn't happen that often to Deer/Allied units, even when things look pretty crispy on the PCB. You're more likely to get that kind of damage from old CWT-built Antecs and of course, the Bestec ATX-250-12E.

                        Originally posted by kaboom View Post
                        Not gonna comment on the other oddities- too early for "Deer" season.
                        Yeah, save them for when I make a thread for it.
                        No really, I will be fixing this thing.

                        Originally posted by pdavid View Post
                        Seventeam ST-220WT-P large AT unit.

                        Standard half bridge topology from 1991.
                        ...
                        I really like the circuit traces on his one.
                        Really nice unit for its time. And even today, it's still a good AT power supply. Most "replacement" AT power supplies nowadays are usually garbage.

                        Comment


                          Re: Power supply build quality pictorial. part 2

                          Originally posted by momaka View Post
                          but this Deer is pretty dead right now
                          IOW, you mean Bambi's mama is EOL??



                          Originally posted by momaka View Post
                          Only if the 5VSB goes out of control. But I find that this doesn't happen that often to Deer/Allied units, even when things look pretty crispy on the PCB. You're more likely to get that kind of damage from old CWT-built Antecs and of course, the Bestec ATX-250-12E.
                          Not only can it run away, Bestec-style, the caps go bad resulting in extreme ripple. I've noticed a trend with Deers- +5VSB usually doesn't come directly from the rectifier on the sec side of the aux transformer. They seem to be wound for 8-12V out, which feeds a 7805- at least the older ones were...


                          Originally posted by momaka View Post
                          Yeah, save them for when I make a thread for it.
                          No really, I will be fixing this thing.
                          Looks like a great learning project. What's nice is there aren't critical parts or SMDs, so they can be rebuilt and missing components added (common-mode chokes, X/Y caps).

                          Originally posted by momaka View Post
                          Really nice unit for its time. And even today, it's still a good AT power supply. Most "replacement" AT power supplies nowadays are usually garbage.
                          Those are nearly "bangbad" or similar... as are those "replacepower" things cluttering ebay and elsewhere.

                          Originally posted by pdavid View Post
                          Well I have other AT units from Seventeam that have 5V standby and 3.3V DC-DC plus PS-on wired to to the circuit to make it work like ATX. There were plenty of similar solutions in mid 90'. Some really horrible ones too.
                          Seen some Deers labeled "Crown" (hardly!) which had an entire +5VSB/AUX/3.3 linear reg PCB mounted above the PCB of the previous year's AT supply. With board-board jumper wires. PS-ON came from this "afterthought" board (which then supplied +12 to the '494) ; as did +3.3, which had a linear reg from +5. This was in the day where low-end systems mainly had the RAM, perhaps a weak video card as the only 3.3v loads. Mainly in cheep clones during the K6-2 era.

                          They really messed something up with those. Had more loss in the snubber and min. load resistors on the hack PCB than those same resistors on the main (AT) PCB!
                          They ran warm while in standby.

                          Last edited by kaboom; 06-22-2018, 07:08 PM.
                          "pokemon go... to hell!"

                          EOL it...
                          Originally posted by shango066
                          All style and no substance.
                          Originally posted by smashstuff30
                          guilty,guilty,guilty,guilty!
                          guilty of being cheap-made!

                          Comment


                            Re: Power supply build quality pictorial. part 2

                            Originally posted by kaboom View Post
                            Not only can it run away, Bestec-style, the caps go bad resulting in extreme ripple. I've noticed a trend with Deers- +5VSB usually doesn't come directly from the rectifier on the sec side of the aux transformer. They seem to be wound for 8-12V out, which feeds a 7805- at least the older ones were...
                            Those 5VSB circuits are found only on very old Deer and L&C PSUs - late 90's to early 2000's. The giveaway is the 7805 regulator, as you noticed, and also lack of optocoupler for feedback for the 5VSB. On that note, those circuits are not even 2-transistor - they are single-transistor oscillator "wonders". And because there is no opto, the primary side is always working "at full blast", meaning that whatever power is not used up by the secondary / output gets dissipated ("eaten") by the primary-side snubber components and transistor. Hence, that's why those early Deers almost always look well-cooked in the 5VSB section. It's a really shitty and inefficient design.

                            Miraculously, however, this actually might be safer than a 2-transistor circuit, because the 7805 regulator can do quite a bit to protect the motherboard from over-voltage. Unfortunately, the PWM controller in the PSU does not get this benefit as it is fed straight from the 5VSB's secondary auxiliary winding... meaning, it gets a chance to fly to the moon *when* (not if) that 5VSB circuit overshoots.

                            Originally posted by kaboom View Post
                            Looks like a great learning project. What's nice is there aren't critical parts or SMDs, so they can be rebuilt and missing components added (common-mode chokes, X/Y caps).
                            Yup.
                            Post/thread about that coming soon. I just opened it today and desoldered the heatsinks (with all parts on them). As I mentioned in my last post above, at least this Allied unit comes in a clamshell case, so I could do all of the work without removing anything.... from the comfort of my own chair, as they say. (in this case, it really was )

                            Originally posted by kaboom View Post
                            Those are nearly "bangbad" or similar... as are those "replacepower" things cluttering ebay and elsewhere.
                            Haha, yeah the "replacepower" units are goofy indeed.

                            Next thing you know when you use one is "replacesystem", then "replacehouse"

                            Originally posted by kaboom View Post
                            Seen some Deers labeled "Crown" (hardly!) which had an entire +5VSB/AUX/3.3 linear reg PCB mounted above the PCB of the previous year's AT supply. With board-board jumper wires. PS-ON came from this "afterthought" board (which then supplied +12 to the '494) ; as did +3.3, which had a linear reg from +5. This was in the day where low-end systems mainly had the RAM, perhaps a weak video card as the only 3.3v loads. Mainly in cheep clones during the K6-2 era.

                            They really messed something up with those. Had more loss in the snubber and min. load resistors on the hack PCB than those same resistors on the main (AT) PCB!
                            They ran warm while in standby.
                            Yup, that be those single-transistor oscillator circuits I'm talking about.

                            I still have a few Deer PSUs with those - not used in a PC, of course. I'm just keeping them around for fun (as in, literally poking fun at them .) They still work, though!

                            Comment


                              Re: Power supply build quality pictorial - Jewel JS-12030-2E power adapter

                              Picked up this 12 Volt, 3 Amp power adapter at a local thrift shop for $3, because one of my portable 3.5ā€ HDD enclosures uses a slightly more weird power jack.
                              It's made by Jewel, model number JS-12030-2E. Had what seemed like a legit UL listing (E233940) too
                              https://www.badcaps.net/forum/attach...1&d=1531433621

                              So I took it home and gave it a quick test: first plugged it in with nothing connected and measured the voltage.
                              https://www.badcaps.net/forum/attach...1&d=1531433621
                              Result: 11.97V ā€“ not exactly ideal. I usually like to see slightly over 12V with these adapters, as the cable can further drop the output voltage under load. But overall, I gave it a pass.

                              However, the slightly low voltage made me a bit suspicious. Thus, I decided to test the adapter with a load. I used a 12V, 20 Watt Halogen light bulb, which comes out to a little over 1.6 Amps of load. How did the adapter do? ā€“ Not well!
                              https://www.badcaps.net/forum/attach...1&d=1531433621

                              But I noticed that immediately after connecting the load, the voltage would start to raise.
                              https://www.badcaps.net/forum/attach...1&d=1531433621
                              https://www.badcaps.net/forum/attach...1&d=1531433621

                              After about 10-15 seconds, it came back up to 11.xx Volts reading with no load. No bueno!. So, I cracked her open, as I had a hunch of what was going on.
                              https://www.badcaps.net/forum/attach...1&d=1531433621
                              Initial impressions: looks fairly packed with a large transformer and main input cap, along with adequate input RF filter and decent heatsinks running the whole side length of the adapter walls. Also has a bridge rectifier (__206 part). The primary switcher is in a TO-220 isolated case (could not read label), and the output rectifier is a Y2010PN (i.e. most likely 20 Amp, 100V) also in TO-220 case. For a 12V, 3 Amp (36 Watt) power adapter, I think that's adequate.

                              What's not so good are the cap choices. Some of you may already have noticed the ā€œKFā€ writing on the input bulk cap, suggesting this unit has CapXon caps. And some of you may also have noticed the problem too, already. No?
                              https://www.badcaps.net/forum/attach...1&d=1531433621
                              See it now? (I bet so! ) ā€“ Yup, bad capsā€¦ or in this case, just a single bad cap right after the output rectifier. The output caps and their arrangement is as follows:
                              1x CapXon LZ, 16V, 1000 uF, 10x16 mm (after output rectifier, before the PI coil)
                              1x CapXon KF, 16V, 1000 uF, 10x16 mm (after PI coil)

                              I decided to replace only the bulged CapXon LZ, as the other one tested okay on my component/transistor tester for both capacitance and ESR. The temporary recap was done with an Elite EB, 16V, 1000 uF cap that I pulled out of an LCD monitor many years ago (which had an issue not related to the caps, but I recapped it anyways).

                              That said, after replacing the bad cap, the power adapter function was restored. It now outputs 11.99V unloaded. Also, using the same 20W halogen bulb as a load, the voltage did not go out of spec. I don't remember exactly how low it dipped, but IIRC it wasn't more than 11.85V (which is not ideal but acceptable, I suppose). So the bad cap was definitely the issue here.

                              That aside, here is a shot of the bottom (solder) side of the PCB:
                              https://www.badcaps.net/forum/attach...1&d=1531433621
                              Just like the voltage regulation, the soldering isn't ideal, with some of the SMD resistors sitting a bit oddly and their joints looking blobby. But again, it's acceptable.

                              Here is another shot with the white silicone glue removed from the upper-left corner in the picture above, so that you can see the traces underneath (I had to remove it anyways in order to get to the bad cap):
                              https://www.badcaps.net/forum/attach...1&d=1531433621

                              Design-wise, it looks okay too: flyback converter with feedback, driven with a single transistor by a PWM chip (IC1, labeled ā€œAAHBWā€, which from what I searched comes out to SG6848). The ā€œstart-upā€ cap for the PWM chip appears to be that small electrolytic one on the primary next to the bulk input cap. It is a CapXon KF, 50V, 22 uF, 5x11 mm. Unfortunately, I could not read the value of the primary bulk cap, as that would have required me to remove a lot of the white goop. However, the can size appears to be 16 x 25 mm. Since the adapter is rated for 100-240V input voltage, I think we can safely assume it is either a 400V or 450V cap. Going by the KF datasheet, that means it is either 47 uF or 33 uF.

                              And that is all I got for this adapter.
                              Attached Files

                              Comment


                                Re: Power supply build quality pictorial. part 2

                                I wouldn’t be surprised if that power adapter was basically unused and if that CapXon LZ bulged on the shelf. While it could have definitely been used and the CapXon LZ could have outgassed due to electrolysis, because of high ripple current duty, since it’s on the output filter of a flyback transformer, here are several examples (not mine since I don’t keep such crap caps around for long...) of many CapXon LZ bulging and leaking despite being stored at room temperature, retaining their full length leads, without ever having bias applied or without ever having been soldered to a PCB:

                                One
                                Two
                                Three
                                Four
                                Five
                                Six
                                Seven
                                Seven from a different angle

                                Yep... that bad. Worse than KZG judging by how many 16V rated ones failed on their own. And I doubt they are counterfeits, since they are very consistent with CrapXon’s track record. That CapXon KF may test okay for now but who knows how long it’s going to last, especially since it’s 12 years old now going by the datecode. CrapXon are still replace-on-sight for me, always have been and always will be.
                                Last edited by Wester547; 07-12-2018, 07:10 PM.

                                Comment


                                  Re: Power supply build quality pictorial. part 2

                                  WOW! ^
                                  Didn't know CapXon LZ series were that bad. Look how many of them bulged without even being used (new leads) - that's insane! Not worse than just KZG... it's worse than almost anything I've seen!

                                  Funny thing is, I do have two CapXon LZ 16V, 470, 8 mm dia. caps in my spare parts bin when I'm messing with a project on a breadboard. Last I checked them (a few years ago) they were still in spec. I don't use them much and I don't trust them either. But for GP use on a breadboard project (mostly testing/playing with old audio ICs), they seem to be holding up okay... or at least they haven't spilled their guts out.

                                  Originally posted by Wester547 View Post
                                  CrapXon are still replace-on-sight for me, always have been and always will be.
                                  Same here.
                                  When I need something fixed properly, I never leave them. In the case of this power adapter, I haven't even closed up its case yet. It sees very little and occasional use right now. When I need it, though, I will be recapping it properly. I'm tempted to even change the input bulk cap, give how many CapXon primary caps I've seen failed. But it may not be worth it for this adapter.

                                  Also, it seems that only the older CapXon KM general purpose caps are actually somewhat stable - at least when used in a non-stressful applications. A few years back when I was recapping a PSU for myself and ran out of caps for the -12V and -5V rails, I used a few CapXon KM caps from another gutted PSU I had. That recapped PSU still has those CapXon KM caps and they are still okay. The PSU is actually the Inno Power 300W one by Macron I posted in this thread many years back.
                                  *EDIT*
                                  Here it is, actually:
                                  https://www.badcaps.net/forum/showpo...postcount=2230
                                  Posted in 2015, so I guess not that long ago. However, looks like I did the recap in 2010. And those CapXon KM caps came from a gutless Meico / J&C that we had in a computer way back in 1999! That means, they are close to 20 years old now!
                                  ... I guess it really may be time to replace them.
                                  Last edited by momaka; 07-12-2018, 06:49 PM.

                                  Comment


                                    Re: Power supply build quality pictorial. part 2

                                    This one surprised me. Other then crap caps, its build quality isn't bad. But it blew for seemingly no reason. This was the charger for my Yaesu FT-70D. One day it just lost output and assumed the FZK cap on the output went short, well, not the case. The output cap has an ESR of 0.31Ohm.


                                    Solder looks good...


                                    Overall design looks fairly good for how high of a switching speed it has. And yeah, i tried changing the output caps anyway.


                                    I think the switching chip went bad. All the current resistors blew open, as well as the power resistor for the driver chip.
                                    Attached Files
                                    Last edited by goontron; 07-18-2018, 11:52 PM.
                                    Things I've fixed: anything from semis to crappy Chinese $2 radios, and now an IoT Dildo....

                                    "Dude, this is Wyoming, i hopped on and sent 'er. No fucking around." -- Me

                                    Excuse me while i do something dangerous


                                    You must have a sad, sad boring life if you hate on people harmlessly enjoying life with an animal costume.

                                    Sometimes you need to break shit to fix it.... Thats why my lawnmower doesn't have a deadman switch or engine brake anymore

                                    Follow the white rabbit.

                                    Comment


                                      Re: Power supply build quality pictorial. part 2

                                      I see the label says, "Yaesu AC Adapter SAD-18B"...
                                      Maybe that should have been a hint about what to expect from it.

                                      In all seriousness, it indeed looks okay. Maybe the startup cap was on its way out and someone how that blew the controller? Also, 0.31 Ohms ESR doesn't look that good to me for the output caps. One of them looks big enough to be either 16V/470 uF or 25V/330uF. In any case, I'd expect a cap like that to have better ESR - possibly less than 0.1 Ohms.

                                      Other than that, the adapter looks capable of its rating. The soldering is okay too.
                                      Last edited by momaka; 07-19-2018, 09:50 PM.

                                      Comment


                                        Re: Power supply build quality pictorial. part 2

                                        I present this HIPRO 230W HP-A2307F3P from an IBM Thincentre.
                                        I do not have many photos, because the font does not have much space and I do not want to take it apart, but it is feeding an HP dc7900 SFF, with a core 2 quad q8300, 1HDD, ati radeon 4360 (tomorrow I replace it with an asus silent gt1030) I installed this HIPRO because it is very quiet, and of much better quality than the original Delta that it was carrying (it was damaged)
                                        I think it has an ST switch, then the capacitors have 2 rubycon of primary (I think remember), on the secondary, 1 rubycon, 2 nippon chemicon, 1 nichicon.
                                        Attached Files
                                        Last edited by kevin!; 08-07-2018, 11:39 AM.
                                        Gaming pc:
                                        nVidia RTX 3080 TI, Corsair RM750I.
                                        Workshop PC:
                                        Intel core i5 8400, Intel SSD 256GB, nvidia gt1030, asus b365-a.
                                        Server:

                                        Comment


                                          Re: Power supply build quality pictorial. part 2





                                          https://cdn.badcaps-static.com/pdfs/...bf1a346042.pdf
                                          'nen ON Semi 79M12

                                          Switcher are International Rectifier IRLR8726

                                          and for +5VSB an L5973D
                                          https://cdn.badcaps-static.com/pdfs/...2e3aac8bb9.pdf

                                          Protection is a SiTi PS222S
                                          Attached Files

                                          Comment

                                          Working...
                                          X