Announcement

Collapse
No announcement yet.

Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

    Today I am posting about a PSU that has been on my “wanted” list for a very long time: the Aspire ATX-AS520W.

    Nope, I am not kidding here folks. I've been curious to put my hands on one of these PSU ever since I got into power supplies. In fact, even before that – back in the mid-2000's before I got into electronics – I've always wanted to have one of these Bling-Bling boxes. Why? I have no clue. They just looked interesting. I know they are not so good. There's actually plenty of examples of them with bad caps and whatnot, including on the forums here – on that note, here are a few threads for those who want to see more of this PSU, in case my wall-of-text post I am about to make is not enough:
    https://www.badcaps.net/forum/showth...ghlight=apevia
    https://www.badcaps.net/forum/showth...ghlight=apevia
    https://www.badcaps.net/forum/showth...ghlight=apevia
    https://www.badcaps.net/forum/showth...ghlight=apevia

    So, about a year ago, I finally got the chance to own one of these. I found it on eBay (as the saying always goes, lol) for $12 shipped to my door. Seller said it was working fine when last removed from his system but warned that he is selling it “as-is”. I think we can all see where this is going.

    Without further delay, let's get started here!
    As usual, first a picture of the PSU's case front and back…
    https://www.badcaps.net/forum/attach...1&d=1508117050
    https://www.badcaps.net/forum/attach...1&d=1508117050
    This PSU features an all-aluminum case, which is one of the reasons I wanted one of these (after all, the case itself was worth the money I paid for it). There are three 80 mm fans: one of the front, one on the back, and one on the top. The top fan also has blue LEDs. Otherwise all fans are re-branded by Aspire and rated for 0.25 Amps.
    Unfortunately, the Aspire PSU I got was not version with the see-through acrylic top. :\ What a bummer, right? … Well, maybe not really, unless you like all kinds of EMI and RFI noises. After all, there is a reason why SMPSs are enclosed in a metal case.

    Moving on to the label:
    https://www.badcaps.net/forum/attach...1&d=1508117050
    I love it! Just look at the table with the rails: not a single of those numbers seems to be aligned with the others. It's like they had a 5-year old kid do their table (well, maybe they really did. ). No combined rating for the “two” 12V rails either.
    The UL number, E126556, points to YoungYear, and still appears to be active for this model number:
    http://database.ul.com/cgi-bin/XYV/t...374&sequence=1
    Well, at least that's some good news – looks like the unit was UL tested and certified. Not that this means much, but better than nothing.

    Next, a shot of the cables.
    https://www.badcaps.net/forum/attach...1&d=1508117050
    As you can see, there are plenty of regular drive Molex cables – eight to be exact. This power supply also provides two floppy cables, two SATA cables, and one Auxiliary power connector (in 2005? WHY?!?!). Main ATX connector is 24-pin, and there is also a 4-pin 12V CPU connector. All output wires are 18 AWG, which is good to see. The only exception is the SATA connectors, which use 20 AWG for whatever odd reason (and of course the floppy connectors – standard 22 AWG).

    To be continued in next post. Please wait as I upload it
    Attached Files

    #2
    Apevia / Aspire ATX-AS520W power supply…. aspires to be decent [pt2]

    And now (*rolling up sleeves, spitting on hands, cracking fingers before typing on keyboard* ) for the internal pictures, starting with an overall top-view:
    https://www.badcaps.net/forum/attach...1&d=1508117050
    This PSU is based on PCB model ATX-66b REV:1.0, 2005/04/07.
    I see full input EMI/RFI filtering, decent sized primary input caps, HUGE main transformer (42 mm size core, actually), very big output toroid inductors, and… that’s where the good stops. Unfortunately, the heatsinks are a bit on the small side, especially for this design – which I can tell you right off the bat is a half-bridge design. Not that there is anything wrong with half-bridge designs – they are quite simple and reliable, actually. But they are also not very efficient. So 520 Watts will definitely not be possible out of this unit, as the PSU will probably start to dissipate more heat than those heatsinks can handle above 400 Watts power draw.
    That said, look at those dust bunnies! Seems like this thing has been in service for a while. This pretty much convinced me that I should actually try and repair this PSU - if nothing major was wrong with it, of course.
    Well, besides all the dust, what do you think?

    As I started to disassemble it, I stumbled onto something a little amusing:
    https://www.badcaps.net/forum/attach...1&d=1508117050
    Don’t see it? Okay, I’ll give you a clue: look at the screws holding the voltage selector switch and tell me what you see.
    While talking about that picture, I do want to note that the ON/OFF switch for this PSU is quite nice – it has a very defined switching action when being toggled. I’m thinking if I fix this PSU, I might actually end up installing a white or green LED in it to make it glow when the PSU is plugged in the wall. After all, this is an “all-show” PSU (don’t know about the “go” part yet).

    With the PCB removed, this is what the bottom of the case looks like:
    https://www.badcaps.net/forum/attach...1&d=1508117050
    Besides the dust (some of which is scorched/burned), I want to point out the PCB mounting system here – it is done by thin metal clips that snap onto two channels at the aluminum case’s bottom half. IMO, that’s not a very good way to safely mount the PCB.

    And here is the PCB itself after a brief de-dusting session:
    https://www.badcaps.net/forum/attach...1&d=1508117721
    Man, look at all those bloaters! I’m sure some of you already spotted that in the first overall shot I posted above, but here it is even more clear. You can also see quite a few darkened spots on the PCB. Whatever this computer was pulled out of sure was NOT in working condition. And I bet the seller knew that. But I knew very well what I was buying, so of course I didn’t hold it against him. For just $1 (shipping ate the rest), I think that’s pretty fair, actually.

    Moving onto the primary side…
    https://www.badcaps.net/forum/attach...1&d=1508117721
    We have full input filter, big primary caps (820 uF in 25x45 mm cans), a bridge rectifier, and also a big film cap for the half-bridge transformer coupling. The only thing that kind of bothers me is the primary heatsink – it could have been bigger. Mainly, YY could have extended it a few cm to the right by removing that small heatsink for the 5VSB transistor and actually mounting the 5VSB transistor to the (now extended) heatsink. Also, the wires for the selector switch are quite a bit on the thin side, rated for only 300V and 22 AWG.

    Here is another interesting shot:
    https://www.badcaps.net/forum/attach...1&d=1508117721
    Hmmm, that snubber resistor does not look very happy. Though its rubber heatshrink cover appears to have broken down along with its ceramic coating, the PCB underneath it does not indicate any signs of overheating. It showed a healthy resistance of 68 Ohms on my multimeter, out of circuit. This snubber resistor is in series with a 4.7 nF, 630V, film capacitor.

    Next, a shot of the transformer labels before moving to the secondary side:
    https://www.badcaps.net/forum/attach...1&d=1508117721
    The two coils next to the main transformer are both for the 3.3V rail’s mag-amp circuit, so it’s dual-side regulated - which is good (most PSUs usually only regulate on one side). And worth noting is that all of the rectifiers on the secondary heatsink are TO-247 and TO-3P parts.

    And a picture from the front:
    https://www.badcaps.net/forum/attach...1&d=1508117721
    Here you can actually see how big the main transformer and output coils are relative to the primary input caps. In terms of component size, YY definitely didn’t cheap out. It’s really only the bad caps and a few bad design decisions that make this PSU mediocre.

    On that note, let’s see where things go bad:
    https://www.badcaps.net/forum/attach...1&d=1508117721
    Look at that cramped secondary side – it’s a mess! I know the caps are crappy brands (Fuhjyyu, Y.C. and Su’scon) so they probably bloated on their own. But I’m sure the lack of airflow from the mess of wires helped them die faster. To add insult to the injury, YoungYear added a 10-Ohm load resistor on the 5V rail (which translates to a whopping 2.5 Watts of heat there).

    But it gets better… as in more bad.
    https://www.badcaps.net/forum/attach...1&d=1508117721
    https://www.badcaps.net/forum/attach...1&d=1508117721
    https://www.badcaps.net/forum/attach...1&d=1508117721
    See how that main output toroid inductor is connected with the wires going everywhere and plugging into various spots on the PCB? It’s like the whole thing was a complete after-thought , as if they forgot to include an output inductor in the final design and did it literally seconds before it got sent to the production floor. If I can imagine a sketch for that, it would probably go like this: YoungYeah started the design with the input EMI/RFI filter and dedicated enough space to it. Then they placed the primary side components, again allocating enough space for them. Then the transformers and secondary heatsink… and by the time they got to the output inductors and caps, they realized, oh shit, we totally forgot the PSU needs those. Let’s just shove them here under the output wires. Should be fine!

    At the same time, this PSU has not one, not two, but THREE fans – which are totally unnecessary and take up a lot of valuable space inside the case. What YoungYear really should have done is get rid of all fans (minus the exhaust) and made the PCB bigger, placing the components with more space between them for better airflow. Not to mention this would have probably saved them time and money, as it would be easier for the workers to install the components. You can bet it took a lot of time for someone to fiddle with the output inductor’s wires to get them in the right holes. Not to mention this makes the production even more likely to suffer due to the possibility of the inductor’s wires shorting to something, like they almost did on my unit above. And to top it all off, the 5V and 12V rails have spots on the PCB for only one cap each. However, YY installed two additional small caps on both rails (470 uF each) by having the caps connect with extra-long leads above the other caps.

    Ugh, I need to take a break from typing here. The more I think about all this “design” nonsense, the more my head hurts. I might need a “special” soldering therapy after posting this. (More like snort an entire can of flux to set my head straight. )

    But before I do that, let me just finish with a few more images. First, a solder-side shot:
    https://www.badcaps.net/forum/attach...1&d=1508117721
    At least the soldering is not too terrible. Might need to fix a few solder blobs (more like solder lakes - look at that trace for one of the primary caps!) and re-align some of the SMD resistors. Some of them barely made it into their targeted spots.
    You can also see there appears to be something that has burned. Let’s investigate…
    https://www.badcaps.net/forum/attach...1&d=1508117846
    Looks like an SMD resistor: CR211.
    Of course, all of that damage was probably caused by these bad boys:
    https://www.badcaps.net/forum/attach...1&d=1508117050
    Despite being different brands, all of these caps have yellow sleeves. I suspect they were all custom-ordered by YY for these PSU – particularly the ones with the clear top case. Well, there is only one cap that is brown – but that’s only because that cap got burned to a crisp. Don’t believe me? Here’s some interesting proof:
    https://www.badcaps.net/forum/attach...1&d=1508117050
    See that tiny yellow spot on the right side of its sleeve? That’s where the cap was touching the (much cooler) heatsink, so the sleeve didn’t burn up. Amazing!
    This cap was on the 5VSB output filter, by the way. So if anyone is still wondering why the PCB has so many burned spots, that’s exactly why – 2-transistor 5VSB circuits do NOT like bad output caps.

    Holy crap, 10k char. Limit AGAIN! To be continued… part 2 lol
    Attached Files
    Last edited by momaka; 10-15-2017, 07:46 PM.

    Comment


      #3
      Apevia / Aspire ATX-AS520W power supply…. aspires to be decent [pt3 lol]

      Welp, that does it for my talk. Here’s a summary of all the major components….

      Primary Side Summary:

      - three 4.7 nF Y2-class certified caps
      - two MCY/HT X2-class certified caps (0.22 uF at receptacle and 0.47uF uF on PCB)
      - two common-mode chokes + one single-mode choke, two MOVs across primary caps.
      - PBU805 bridge rectifier (8 Amps, 600 V DC)
      - 250V, 8A glass fuse in fuse holder
      - two Koshin KL5 bulk caps (in series): 200 V, 820 uF, 25 x 45 mm (dia. x h)
      - two 2SC5058 NPN BJTs (TO-3P case) for main PS (H-bridge design)
      - main PS snubber network: 68-Ohm, 5W resistor in parallel with 4.7nF, 630V film cap
      - two Fuhjyyu TN, 50V, 10uF, 5x11 mm caps for BJT drive circuit
      - one 2SC5353 NPN BJT (TO-220FP case) for 5VSB (2-transistor design)
      - transformer sizes: 42 mm (main), 19 mm (BJT driver), 19 mm (5VSB)
      - 18 AWG, 600V input wiring (except selector switch, which has 22 AWG, 300V)

      ICs
      - KA7500c (PWM), AAT1014 (output supervisor), “817” optocoupler (5VSB)

      Secondary Side Summary:
      3.3 V rail:
      - mag-amp regulated (both sides)
      - PEC SB3040PT (30 Amp, 40V, TO-247) schottky recrifier
      - 2x Fuhjyyu TNR, 10 V, 3300 uF, 10x25 mm
      - PI coil between caps: 8 mm core, 5 turns, 14 AWG wire

      5 V rail:
      - PEC SB3040PT (30 Amp, 40V, TO-247) schottky recrifier
      - 1x Fuhjyyu TNR, 16 V, 3300 uF, 12.5x30 mm before PI coil
      - 1x Su’scon SK, 16 V, 470 uF, 8x15 mm after PI coil
      - PI coil: 8 mm core, 5 turns, 14 AWG wire
      - 10-Ohm, 3 Watt load resistor

      12 V rail:
      - SIRECT MBR30100PT (30A, 100V, TO-3P) schottky recrifier
      - 1x Su’scon SK, 16 V, 470 uF, 8x15 mm before PI coil
      - 1x Fuhjyyu TNR, 16 V, 3300 uF, 12.5x30 mm after PI coil
      - PI coil: 6 mm core, 4 turns
      - 150-Ohm, 3 Watt load resistor

      5 VSB rail:
      - SB340 diode (3 Amp, 40V, DO-201) schottky type
      - 1x Y.C. 10 V, 2200 uF, 10x20 mm before PI coil
      - 1x Y.C. 10 V, 1000 uF, 8x15 mm after PI coil
      - PI coil: 4-5 mm core, 10 turns
      - 47-Ohm, 1 Watt load resistor

      -12 V rail:
      - two 1 or 1.5 Amp diodes
      - 1x Su’scon SK. 16 V, 470 uF, 8x15 mm after PI coil
      - 1-kOhm, ½ Watt load resistor

      -5 V rail:
      - two 1 or 1.5 Amp diodes
      - 1x Su’scon SK. 10 V, 470 uF, 8x12 mm after PI coil
      - 300-Ohm, ½ Watt load resistor

      OMG! You are still reading. Why?
      OK, this is the end... for today.

      Comment


        #4
        Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

        For some reason I'd actually see the design work a bit better in a 120mm casing rather than 80mm. Something about the heatsinks tell me a 120mm fan suits better.

        Other than that, they packed it up pretty good! If it wasn't for the Fuhjyyu and Su'scons and the load resistors on the secondary side, this would have been a pretty good PSU.

        On a side note, does the design have the silkscreen for another 12v in parallel? Maybe unrelated but I just found out that one of my rebuilt Delux PSUs actually has place for another 12v rectifier in parallel. Might upgrade the heatsinks and add it

        Wonder how much it does in spec if recapped?
        Main rig:
        Gigabyte B75M-D3H
        Core i5-3470 3.60GHz
        Gigabyte Geforce GTX650 1GB GDDR5
        16GB DDR3-1600
        Samsung SH-224AB DVD-RW
        FSP Bluestorm II 500W (recapped)
        120GB ADATA + 2x Seagate Barracuda ES.2 ST31000340NS 1TB
        Delux MG760 case

        Comment


          #5
          Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

          That's new
          yellow fuhjyyus
          is it true
          that they're still poo?

          Comment


            #6
            Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

            i want the mains inlet!!
            is it uv reactive??

            Comment


              #7
              Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

              fuhjyyu!
              Attached Files

              Comment


                #8
                Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                Originally posted by eccerr0r View Post
                That's new
                yellow fuhjyyus
                is it true
                that they're still poo?

                That's good poetry, man!

                Yup, Apevia/YoungYear custom-ordered their caps from Fuhjyyu, Su'scon, and Y.C. in that yellow color. I've also seen yellow CapXon and Koshin (like my primary caps) in these Apevia PSUs.
                ... and yes, they were still poo.

                Originally posted by eccerr0r View Post
                *Yawn*
                You still have a long way to go, my rookie friend. Only when you obtain a pure rare sample of the 16V, 1650 uF Fuhjyyu cap will I consider you a pro.

                Originally posted by stj View Post
                i want the mains inlet!!
                is it uv reactive??
                I believe so. In fact, I think most of the connectors are, along with the cable sleeves too. Or at least I remember some reviews way back in the day mentioning that it was.
                As crappy as this PSU is/was, it still did make for a cool-looking PC.

                Originally posted by Dan81 View Post
                For some reason I'd actually see the design work a bit better in a 120mm casing rather than 80mm. Something about the heatsinks tell me a 120mm fan suits better.
                Indeed both heatsinks are low enough that a 120 mm fan could fit over them. After all, the top 80 mm fan is 25 mm thick.

                That said, it's still very disapointing how small the heatsinks are. I actually had forgotten how short the secondary heatsink was. As I opened the unit today to check something, I noticed that the secondary heatsink is almost 1 cm shorter in height than the primary. And I really don't understand WHY!

                Originally posted by Dan81 View Post
                Other than that, they packed it up pretty good! If it wasn't for the Fuhjyyu and Su'scons and the load resistors on the secondary side, this would have been a pretty good PSU.
                Fair point. Though I should mention that NONE of the Su'scon output caps were bad, despite being the smallest, and one of the a little overheated from the fan controller series resistor, IIRC. Meanwhile, the Fuhjyyu were bad, as expected. But that browned Y.C. cap I showed above - that didn't even register on my meter, IIRC. Just open. I forgot what the other one measured, but it wasn't a whole lot better. How the 5VSB still functioned is a mystery. Oh wait, did I say function? I'll type up a post tomorrow of what happened when I tested the PSU as is and also some improvements that I made already.

                Originally posted by Dan81 View Post
                On a side note, does the design have the silkscreen for another 12v in parallel?
                I think there is only one for the 5V rail, looking at the solder-side pictures. But I'll double-check, as I don't recall there being any. For the 12V and 3.3V rails, definitely no.

                Originally posted by Dan81 View Post
                Maybe unrelated but I just found out that one of my rebuilt Delux PSUs actually has place for another 12v rectifier in parallel.
                If it's a Deer/Allied, I wouldn't be surprised. They often have spots for more (and bigger) rectifiers, but only really the "high-end" models *might* have them.

                Originally posted by Dan81 View Post
                Wonder how much it does in spec if recapped?
                Bah. Probably still no more than 400 Watts safely. Maybe up to 450W if you are ballsy, but I wouldn't risk it. Sure everything is nice and big and oversized, but remember that this PSU has a KA7500 IC, so it probably runs at a relatively low switching frequency, which means the transformer and primary BJTs need to be big anyways. On the plus side, a simple designs like this cares more about the size of the output capacitors than their ESR. So I can use even general purpose caps and still get good results.
                Last edited by momaka; 10-17-2017, 12:36 AM. Reason: mistakes and typos galore

                Comment


                  #9
                  Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                  Strange Fuhjyyu had special runs of different colors though.
                  The ones I have are all TM(R) and TN(R) - supposedly TM is their blue low ESR, TN are general purpose black 105C and I would guess that R is ROHS...
                  I guess any company can be bribed?

                  Comment


                    #10
                    Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                    Originally posted by eccerr0r View Post
                    Strange Fuhjyyu had special runs of different colors though.
                    The ones I have are all TM(R) and TN(R) - supposedly TM is their blue low ESR, TN are general purpose black 105C and I would guess that R is ROHS...
                    I guess any company can be bribed?
                    Wave enough cash around in the right places and they'll find a way to have chartreuse sleeves with magenta polka dots.
                    PeteS in CA

                    Power Supplies should be boring: No loud noises, no bright flashes, and no bad smells.
                    ****************************
                    To kill personal responsibility, initiative or success, punish it by taxing it. To encourage irresponsibility, improvidence, dependence and failure, reward it by subsidizing it.
                    ****************************

                    Comment


                      #11
                      Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                      Originally posted by PeteS in CA View Post
                      Wave enough cash around in the right places and they'll find a way to have chartreuse sleeves with magenta polka dots.
                      Yeah, absolutely.
                      Look at the CapXon or Teapo inside Super Flower units.

                      If you order enough and wave around enough money, they do everything YOU want.
                      They even print YOUR name onto the capacitor...

                      Comment


                        #12
                        Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                        True, some probably would even print someone ELSE'S name... and there we get the phony capacitors...

                        Comment


                          #13
                          Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                          actually, there are company's that just sell rolls of heatshrink with whatever you want printed on it.
                          i have seen reels on sale with xxxFIRE printed on it - anybody want to re-sleeved any lithiums?!!

                          Comment


                            #14
                            Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                            It seems like a lot of PSU manufactures like to use custom colored capacitors.
                            The only custom yellow capacitors i've seen are GL and Canicon, IIRC these capacitors were from a Ultra PSU that i was recapping. (I think i made a post of it somewhere on the site)
                            I'm not a expert, I'm just doing my best.

                            Comment


                              #15
                              Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                              i have seen the yellow fuck-u caps before, in atx psu's with acrylic lids.

                              Comment


                                #16
                                Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                                Originally posted by stj View Post
                                i have seen the yellow fuck-u caps before, in atx psu's with acrylic lids.
                                And all of those were Aspire/Apevia (YoungYear) PSUs, just like this one. Same/similar layout and all (though some had red PCB and green-anodized heatsinks).

                                Originally posted by ruky con View Post
                                It seems like a lot of PSU manufactures like to use custom colored capacitors.
                                The only custom yellow capacitors i've seen are GL and Canicon, IIRC these capacitors were from a Ultra PSU that i was recapping. (I think i made a post of it somewhere on the site)
                                AFAIK, those Ultra PSUs were also YoungYear and similar to the unit I posted here.

                                Originally posted by stj View Post
                                actually, there are company's that just sell rolls of heatshrink with whatever you want printed on it.
                                i have seen reels on sale with xxxFIRE printed on it - anybody want to re-sleeved any lithiums?!!
                                Ha! Someone should order that with the name "HouseFIRE", then see how many noobs on eBay and Ali Express will actually pickup the joke. If someone buys them, then they are stupid.
                                Last edited by momaka; 10-18-2017, 09:59 AM.

                                Comment


                                  #17
                                  Apevia / Aspire ATX-AS520W power supply - repairs v1

                                  I forgot to mention that before I did anything to the PSU (other than opening the top upon arrival ), I tried plugging the power supply to the wall (though an incandescent bulb, of course) to see if the 5VSB came up. Guess what? It did – and that was with those horribly blown Y.C. caps . BUT! The PSU was not a happy camper for sure – the 5VSB squealed for a few moments before settling into “regulation” at around 4V with *no load*, meanwhile the PSU was drawing a whooping 8 Watts from the wall. At the same time, the secondary auxiliary rail was sitting at 26.5 Volts (which is technically still well within the KA7500’s Vcc range, but a bit high for unloaded 5VSB from what I am use to seeing.)

                                  So, the first thing I did: recap the 5VSB to see if I could get it to output a healthy 5V. Out went the Y.C. caps and in went… United Chemicon KZJ (16V, 1500 uF) and Teapo SC (6.3V, 1000 uF).
                                  https://www.badcaps.net/forum/attach...1&d=1508342847
                                  Okay, I know those aren’t exactly ideal choices, and both were board pulls (with the Teapo SC coming from HP DC5000 PCs that had many of those blown – though the one I used measured “reasonably” good ESR ). After the recap, I hooked the PSU again to my incandescent bulb setup and plugged it in. Result: proper 5V on the 5VSB and no squealing. Also as I expected, the secondary auxiliary rail dropped a bit – to 23.5V. First I tried putting a 0.3A load on the 5VSB, and all went well. The sec. aux. rose only to about 24V, which is good. Then I tried a 1A load. Unfortunately this time, I got only 3.75V out of the 5VSB, and it was squealing like mad at me (but regulation was stable). Very strange, I thought. Then it crossed my mind that the incandescent bulb I had in series with the live line was only rated for 40W. Having no mercy, I swapped it for a 100W bulb and tried it again, expecting something to cook or smoke. But nope – 5VSB came back up to 5V again, no fuss. And aux. sec. came out at 26V with this load, which is alright again. For repeatability, I swapped the bulb one more time back with the 40W. As expected, the 5VSB was back to 3.75V. So conclusion: it seems the 5VSB of this PSU does *not* like low line voltages.

                                  Next item on the list: the snubber resistor on the primary – the broken heat shrink was such an eyesore to look at!
                                  https://www.badcaps.net/forum/attach...1&d=1508342847
                                  I cleaned it well with IPA, wrapped it with a few layers of electrical tape, and added some heatshrink tubing only at the edges where it is closest to the primary heatsink. Then I soldered it back in – it barely fit. The result:
                                  https://www.badcaps.net/forum/attach...1&d=1508342847

                                  Speaking of burned resistors, remember this one?
                                  https://www.badcaps.net/forum/attach...2&d=1508117846
                                  I found a picture of a similar Aspire PSU (not exactly the same model/layout), and R211 was 1.5 kOhms, if I remember correctly. While I do have 1.5 KOhm SMD resistors of that size from some broken CRT TV boards, I decided to do something different (as usual). Hint: it is a bit
                                  https://www.badcaps.net/forum/attach...1&d=1508342847
                                  Yup, those are two 1-kOhm SMD resistors in series. I figured since the original resistor burned, higher resistance with two series resistors should avoid that, because they can dissipate more heat. Of course, this comes with the risk of making some circuit not work. But from what I remember tracing, R211 provides current to the BJT drive circuit on the secondary side. Thus, if the resistance was too high, the main power supply would just not work when requested to turn ON.

                                  The last step was to install some new caps and test the PSU. But before doing that, I thought now would be a good time to fix this:
                                  https://www.badcaps.net/forum/attach...1&d=1508342847
                                  There! No more bare wire running between random components and nearly touching their leads. It was a pain too, as I did this without removing the whole output inductor (used tweezers and pliers to slip that insulation on the wire).

                                  And finally, the caps:
                                  https://www.badcaps.net/forum/attach...1&d=1508342847
                                  Out went the yellow caps and in went the dark blue caps. Oh wait, did I say dark blue caps? What replacement caps in the world did I just use? Rubycon MBZ?
                                  –Nah, those don’t have “X”-type vents. So what then, OST RLX?
                                  –Nope.
                                  Before I tell you, take a deep breath.
                                  I *re*-used ….. wait for it………. GSC caps. Yes folks, I’ll say that again – I re-used GSC caps. They are all ME series that came from my Jetway N2PAP-LITEmotherboard. To save you all the reading in that thread – yes, many of these caps were starting to go (electrically) leaky and registering higher than normal capacitance. But they passed my “spark” test – i.e. charging them up to maximum voltage and discharging on a metal object, observing for sparks. This method works on larger caps and will reveal if a bad leaky cap is fooling an ESR meter to show good ESR and capacitance when the cap is dead. Thus, for a temporary recap, I figured these would do. After all, I had no idea if the PSU was going to nuke itself, so I figured it wasn’t worth using brand new parts. The caps I used were:
                                  2x GSC ME, 16 V, 2200 uF, 12.5 mm dia. (one for the 12V rail and one for the 5V rail)
                                  3x GSC ME, 6.3 V, 1500 uF, 10 mm dia (two for the 3.3V rail and one for the 5V rail)
                                  The -12V and -5V rails were left alone with their Su’scon caps back on duty, as they registered good ESR and capacitance *and* passed my spark test.

                                  After all this mad business, I connected the power supply to the wall with the 100W bulb in series with the line again and plugged it in. Bulb barely flashed, then pretty much went dark. The 5VSB came up and nothing smoked. My new replacement resistors in place of R211 also remained cool. So then came the moment of truth: with a 12V 20W Halogen bulb on the 5V rail (1A of current draw) and a 12V 10W car bulb on the 12V rail (about 0.8A of current draw) I jumpered the PS_ON signal to ground, getting ready for something to smoke. The two load bulbs lit up as the 5V and 12V rail voltages came up for about a second. Then the 100W series bulb became dim and OFF went the PSU. I tried it a few more times, and same thing.
                                  Then it occurred to me that I’ve seen this before with a series light bulb – it usually means the voltage input to the PSU is getting too low and so the PSU cuts out. I measured it, and sure enough, the input voltage was dropping below 90V. Thus, I swapped the 100W incandescent bulb for a 450 Watt heating element. This time, WE HAD SUCCESS! All voltages came up normal and stayed that way. I tested the PSU for about 5 minutes and then unplugged it. The PSU remained working for almost an entire second with this load before shutting down, so the hold-up time looks okay. I then started checking for hot parts. Needless to say, I didn’t find any. The “hottest” part perhaps was that snubber resistor on the primary side, and by hottest, I mean just a little warm. Not burning or anything like that.

                                  So folks, I think this PSU is going to be saved after all. It may not be the best, but I am going to put more work into fixing up some of its problems – that means giving it better caps too (well, I don’t know if it’s worth new Japanese caps, but it definitely deserves better than GSC). Until that happens, this is all the updates I have.
                                  Attached Files
                                  Last edited by momaka; 10-18-2017, 10:17 AM.

                                  Comment


                                    #18
                                    Re: Apevia / Aspire ATX-AS520W power supply - repairs v1

                                    Originally posted by momaka View Post
                                    So folks, I think this PSU is going to be saved after all. It may not be the best, but I am going to put more work into fixing up some of its problems – that means giving it better caps too (well, I don't know if it's worth new Japanese caps, but it definitely deserves better than GSC). Until that happens, this is all the updates I have.
                                    This has been a really fun read. It obviously takes a long time for you to write out your details.

                                    I will say that I was rather surprised that some of the changes were a bit - to use your own term:

                                    I figured you were going to take poor PSU and turn it into the Cadillac of PSUs. Haha. Of course you also mentioned you didn't think it was worth new Japanese caps - so I imagine "fun experiment" would best describe your endeavor?

                                    Comment


                                      #19
                                      Re: Apevia / Aspire ATX-AS520W power supply - repairs v1

                                      Originally posted by ckrtech View Post
                                      so I imagine "fun experiment" would best describe your endeavor?
                                      Yeah, more or less. A lot of the repairs I do for myself are like that. Eventually if the device does "prove" itself as reliable and trustworthy, I do invest more time and better parts into it.

                                      So for this PSU, I pretty much just did the "minimum" number of repairs that I thought were important to get it going.

                                      Originally posted by ckrtech View Post
                                      This has been a really fun read.
                                      Thanks, glad toy enjoyed.

                                      Originally posted by ckrtech View Post
                                      It obviously takes a long time for you to write out your details.
                                      Not that long. I'm used to writing long posts by now. Usually two to three 30-minute sessions in MS Word is all I need for one of the long posts like above. And yes, I do like putting all the details I can think of - helps me later if/when I get something similar to repair again. Plus, I am also enjoying myself with these posts, as one might tell from all the smilies.

                                      Comment


                                        #20
                                        Re: Apevia / Aspire ATX-AS520W power supply…. aspires to be decent

                                        i have tested the fuckyou 1650 pulled from an unused psu.
                                        it just barely made 1000uf and .21 esr.
                                        garbage from the start.
                                        Originally posted by momaka View Post


                                        *Yawn*
                                        You still have a long way to go, my rookie friend. Only when you obtain a pure rare sample of the 16V, 1650 uF Fuhjyyu cap will I consider you a pro.

                                        Comment

                                        Working...
                                        X