Announcement

Collapse
No announcement yet.

Some PlayStation 3 fan repairs, just for fun

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

    Some PlayStation 3 fan repairs, just for fun

    Prepare yourself for another long thread of mine detailing a nearly worthless repair.

    I will start with an NBM-MAT 140 mm squirrel cage fan, model BG1402-B045-P00.

    The story:
    Back when I worked in the console repair shop in 2012, there was a technician there that didn't know much about electronics at the time. One day, he thought that he could plug a 12V PlayStation 3 fan into the wall to test it and it would work. He thought it would just spin faster. His logic: spin 10 times faster, because the wall provides 10x more voltage (120V AC here in the US.)

    Needless to say, when he attempted to run the fan like that, there must have been a small 4th of July fireworks show. I wasn't there when it happened, but I saw the fan in the trash and asked what was wrong with it and why it was thrown away. He said it was burned. He also gave me another one that he burned in a similar way (more on that in a bit). When he explained what he did, I was shocked and amused at the same time – the guy was dead serious. It was hard to keep a straight face, but nevertheless I explained to him why his idea/experiment failed. Hopefully he remembered. Glad I wasn't a person to give him my computers for repair. But after all, everyone has to start somewhere, right.

    Anyways, I took both bad fans with me, even though I wasn't sure what I was going to do with them at the time. They sat in storage for maybe 2-3 years until I had to make an order from Digikey for some parts and decided why not pitch in a few more dollars to see if I could get these fans going.

    Granted, one can buy replacement PS3 fans on eBay for around $10, the repair wasn't worth it from an economical stand point. But I decided to do it anyways, mainly to see if those fans were revivable. And I also wanted to have a working PS3 fan, just in case I ever think of a project that might need it.

    With all that said, here are some pictures of the fan and the aftermath of being plugged in to 120V AC.

    The fan shot:


    Label:

    Look at that current rating: 2.9 Amps at 12V! That is almost 35 Watts of wind power!!! I've ran one at full speed out of its cage before, and they are scary.

    But now for the even more interesting part: the carnage.

    Oh yes! Look at that brown and black sooth. Something really must have released a ton of magic smoke.

    Before I move to the next shot, I want to take a moment here to mention how hard it is to take apart one of these fans. If you look at the cage hub and also the stator that goes on top of it – that took a lot of force and creativity to separate them. As you can see, both are made of metal and what makes it worse is that they were heat-pressed together at the factory. To add insult to the injury, there was also a bit of adhesive between them. But why separate them? –Because the fan's PCB is attached to the stator assembly, so there would have been no way to repair the fan if I didn't take it apart.

    So how did I take it apart? Answer: heat, ice, and lots of sweat. To elaborate, basically I heated the entire fan assembly with the heat gun to a temperature as high as I could hold with a glove and not get burned. Then, because the stator assembly and cage hub were heat-pressed together, the hub has to be cooled (to shrink it) while the stator assembly is hot (expanded). After heating everything, I fed ice through the hole in the hub with one hand while turning and pulling away with the other hand.

    About 10-15 minutes of heating, cooling, and turning, I got it apart. Finally, I was able to see the PCB and its components… or whatever was left of them

    Hmmm... I think I have found the source of the sooth – look at that MOSFET!

    Say what now? Did I hear anyone say it doesn't look that bad? Well here is a close-up for you then:
    https://www.badcaps.net/forum/attach...1&d=1500759436

    And even the controller (a Sanyo LB11867FV single-phase variable speed controller) took a hit:
    https://www.badcaps.net/forum/attach...1&d=1500759436
    Mmmm. Well done!

    That said, it wasn't just one FET and the controller that were bad – ALL of the MOSFETs on the board were shorted on ALL pins. I suppose that is one good reason NOT to plug fans directly in the wall.

    Of course, before I could order any parts, I had to find what else was bad. So I started removing components. As soon as I tried to remove the incinerated FET, it just fell apart.
    https://www.badcaps.net/forum/attach...1&d=1500759436

    But eventually, I got everything that was bad removed.
    https://www.badcaps.net/forum/attach...1&d=1500759436

    Worth noting here is that SMD resistor R15 at the top was also open. It is in series with a bunch of ceramic caps, which filter power to the FETs and reduce EMI from going back to whatever the fan is powered from. I'm not sure why NMB put the resistor in series with those caps, as that essentially increases their ESR to the value of the resistor (it was a 0.036 Ohm resistor, if I am not mistaken). In any case, had the caps been connected directly without it, there is a chance they could have taken the hit instead of the MOSFETs and possibly saved everything else. Judging by the blown resistor, looks like they did short out from the over-voltage too, which is why R15 opened. However, my multimeter indicated they were fine - and they were, indeed.

    As such, I replaced R15 with a jumper wire (figured it wouldn't hurt anything) and also the rest of the blown components: that is, four MOSFETs (two 2SK4067 N-channel FETs and two 2SJ646 P-channel FETs) and the controller. Miraculously, everything else was fine, including the Source resistors for the FETs. For the N-ch FETs, I used two 4813NH units from a Xbox 360 motherboard (they look burned because they were removed with a blow torch as an experiment ) and for the P-ch FETs, I used P4051L - not exact replacements, but close enough.

    Here are the results:
    https://www.badcaps.net/forum/attach...1&d=1500759537
    https://www.badcaps.net/forum/attach...1&d=1500759537
    https://www.badcaps.net/forum/attach...1&d=1500759537
    https://www.badcaps.net/forum/attach...1&d=1500759537
    That one FET burned so bad that even the trace for its Source pad was burned. As such, I had to bridge it with some wire and solder. It is a miracle the Source resistors survived.
    It also took a bit of eye-strain to hand-solder that controller, as this was my second SSOP IC I soldered by hand (SOIC are much easier). I do have hot air as well, but I never got to like using hot air. When doing these SSOP/TSOP packages, make sure your iron has almost no solder on its tip. Even a tiny bit too much can cause a bridge. Drag soldering was not possible here, but I will go over the details of why a little later below.

    After all of this, I put the fan back together (but without gluing the stator to the cage hub). Unfortunately, during disassembly a long time ago, I had damaged the metal c-clip that held the rotor's shaft to the fan. So I used a good old friend here: Mr. Paper Clip.
    https://www.badcaps.net/forum/attach...1&d=1500759537
    (The bearing shield also had some damage, but the bearing was okay.)

    Finally, the moment I've been waiting for: I plugged the fan into 12V with the PWM wire grounded and measured the current – it was drawing just a few mA and not spinning, as it should. I then connected the PWM wire to a PWM circuit based on a 555-timer and started cranking up. Result: IT WORKS!

    On that note, I strongly discourage anyone from running these fans at full speed when they are removed from their PS3 housing. The rotor spins very fast at full speed and can hurt you or damage itself if it gets in contact with something. Always control the speed through a PWM controller on these fans!
    Attached Files

    #2
    Re: Some PlayStation 3 fan repairs, just for fun

    Naturally, this is where the above repair should have ended. And it did… until one day I decided to play around with the fan a bit. However, instead of using my usual ATX PSU, I decided to use a linear 12V power brick (wall wart). The linear brick output about 16V unloaded and I knew about it. The fan controller was rated for 18V, and the FETs much higher, so I thought that wouldn't hurt anything… but I thought wrong. As soon as I connected the fan to the linear brick, it spun a few times and then stopped. Connecting and disconnecting the adapter made large sparks on the connector leads – uh-oh, looks like something got fried again , and this time it was me that did it.

    I opened the fan again (good that I didn't glue it back) and checked the FETs, but they were fine. I checked all the other resistors and whatnot, but everything else was fine too. Then I plugged the fan board briefly without the rotor. It still drew a huge amount of current. And sure enough, the controller became hot instantly. I knew what that meant - dead controller again.

    Luckily, I had purchased two of these controllers: one for this fan and one for the other fan (which I will post about in just a minute). Turned out the other fan did not need a new controller, so I got luckily and used the second controller for this fan again.

    But things didn't go as smoothly as I hoped. Back when I changed the FETs on the fan, I could hear the PCB making crackling noises when I heated it with my iron, most likely due to moisture trapped in the layers – after all, cheap brown phenolic paper boards are known to do this.

    So when I tried to remove the controller the second time, guess what happened?
    … snap, crackle, and pop – the board popcorned on one side of the controller and nearly ripped all of the pads out.

    With lots of care, however, I was able to remove the (now shorted) controller and clean up the board for the new one (again). I soldered the new one and tested the fan with PWM pulled to ground, as before: the fan drew just a few mA as it should. However, when I tried to spin it up, it wouldn't start every time. I noticed that it started only at certain positions. Thus, I opened the fan again and checked everything. I couldn't find anything wrong, though. Almost thought the hall sensor must have gone bad. But then one final inspection under good light revealed the problem:
    https://www.badcaps.net/forum/attach...1&d=1500760207
    See the last pin on the right side of the controller? –Yup, not soldered. And all of this was due to the popcorned board making the controller not sit evenly on the pads. I actually had a ton of trouble getting the (new) new controller soldered because of the damage some of the pads took from the popcorned PCB. This is why I am glad I did not attempt to drag-solder it. Had I done that, who knows how much worse the PCB would have gotten. Phenolic boards that have picked up moisture are extremely sensitive to high heat, so it is best to solder things as quickly as possible.

    After re-soldering that unconnected pin, the fan worked fine again. But this whole accident got me thinking: maybe I should have a series resistor and a Zener diode to limit the voltage going to the controller. Given that the controller only drew a few mA, I thought that a resistor of a few hundred Ohms would do the trick. So I cut the positive (+) bus trace going to the controller's VCC pin and soldered an SMD 220 Ohm resistor, as I had one handy.
    https://www.badcaps.net/forum/attach...1&d=1500760207

    That however, turned out to be a NOT-so-good idea. Even though the controller only drew a few mA, the resistance multiplied by the current draw made the resistor drop a little over 1.5V. If the controller was designed to drive all N-channel FETs, that likely wouldn't have been a problem. But because it wasn't, the P-channel FETs would automatically become turned ON as soon as the fan was plugged in. With both conducting, clearly there would be no way the fan would work.

    Thus, I removed that resistor and replaced it with a 0-Ohm one. Perhaps in the future, I can replace that with an SMD fuse and have a Zener diode to clamp the voltage in case of OV. With that, the fuse should blow and save the controller… or so I hope. But I haven't done that yet.

    However, I did do another mod… or two of them, actually. The first one was to make the fan turn at a minimum speed, even when PWM pin was disconnected or grounded. This was done by populating the spots for resistors R12 and R13. I used 10-KOhm and 4.7-KOhm resistors, respectively. I forgot what speed these values set, but I remember spending some time reading into the controller's datasheet and trying to figure it out, as the datasheet did not mention anything specific.

    The second mod was RPM monitoring. As per the controller's datasheet, pin 9 (FG) can be connected to a 10-KOhm resistor and then out for RPM monitoring. I haven't tried to see if I have successfully implemented that feature yet, but I did add the resistor and wire. So with any luck, this fan may be used as a 4-pin fan in a PC.

    Here is a picture of both mods:
    https://www.badcaps.net/forum/attach...1&d=1500760207

    And that folks, concludes this repair. I imagine this is probably the most detailed post on the Internet about this particular PS3 fan. But I had fun fixing it and even learned a few things along the way. So I think it was time well spent.
    Attached Files

    Comment


      #3
      Re: Some PlayStation 3 fan repairs, just for fun

      They say, if you don't succeed, you should try again.

      After said technician above managed to blow up the first fan, he went for anther round and took a different PS3 fan – a Nidec model G14t12bs1af-56j14. But this time, he used one of these cheap light dimmers in hopes of “regulating the voltage speed” (his words ). Unsurprisingly, the results were the same, of course: magic smoke and probably some fireworks again.

      Due to the (improperly wired) AC light dimmer, this fan didn't seem to have gone out in such a spectacular fashion as the first one. But it still had quite a few dead components. I'll get to that in a bit.

      First, here is a fan shot:

      This one is the 15-blade version, as opposed to the other one, which has 19 blades.

      And the label:

      Also not as powerful, rated for *only* 1.75 Amps at 12V. That's still a decent amount for a 140 mm fan, though.

      Here's also a back shot:

      Note here that the only way to take the fan apart to get to the circuit board was to cut the plastic tabs (already cut in the picture) that hold the stator hub and rotor to the metal cage. If there is another way to take apart these fans, I am all ears. At least it was much easier than the NBM fan.

      Removing the rotor and stator assemblies revealed the PCB. Here it is:

      As you can see, nothing looked visibly bad. However, both MOSFETs (SP8M2, complimentary MOSFET pair in SOIC8 package) were shorted or partially-shorted, as were their Source resistors
      https://www.badcaps.net/forum/attach...1&d=1500761204

      This fan also had a fuse labeled “T” (which implies this should be a 5 Amp fuse). It was blown as well.
      https://www.badcaps.net/forum/attach...1&d=1500761204

      And to make matters even worse, the transistor responsible for sending PWM signals to the controller was also open.
      https://www.badcaps.net/forum/attach...1&d=1500761204

      After removing all blown parts, and installing a short thin wire in place of the fuse, I plugged the fan to a 12V, 1A power adapter and measured the current to see if there would be more surprises. But the only surprise was that despite all of the bad components mentioned above, the controller (again, a Sanyo LB11867FV) was fine here and drew normal current (just a few mA again). This was great news, because I had ordered two controllers. A spare of a hard-to-find part is always good, as you saw why in the previous post.

      With all of that, I started replacing components. I changed the FETs with exact replacements. Unfortunately, I didn't get anything for the Source resistors, so I just used a piece of wire in place of those. The blown transistor for the PWM speed control, I replaced with a diode for the time being (just for testing), as the fan already had minimum speed resistors. This essentially converted the PWM input pin to a linear voltage control input (if I remember correctly, the controller varies the fan speed according to a voltage it reads between around 1V and 3V). And finally, just like the opened R15 resistor on the NBM fan, there was a similar resistor on this fan as well (up on the stator winding side). I replaced that with a jumper, too. So here are the pictures of the repair:
      https://www.badcaps.net/forum/attach...1&d=1500761266
      https://www.badcaps.net/forum/attach...1&d=1500761266
      https://www.badcaps.net/forum/attach...1&d=1500761266
      https://www.badcaps.net/forum/attach...1&d=1500761266
      https://www.badcaps.net/forum/attach...1&d=1500761266

      After all of this, I plugged the fan in to a 12V source with the PWM pin disconnected (floating). Fan spun at its set minimum speed, which was good to see. I ran it for about 20 minutes and pulled apart to check if the FETs or something else was hot. But as my current measurements indicated, the fan was drawing normal current the whole time, so nothing was excessively warm. I then varied the voltage on the PWM pin between 1 and 3.3V, and the fan speed responded accordingly. So looks like everything worked correctly.

      This concludes the repair for both fans… for now. I might open them again someday and mod them with the protection Zener diodes and fuses. After all, it is always worthwhile to have a few spare fans, especially weird ones. Who knows how much longer until you all see me make a post in the ghetto mod thread with one of these. Though at this moment, I don't have any use for them, so they are back in storage.
      Attached Files

      Comment


        #4
        Re: Some PlayStation 3 fan repairs, just for fun

        Originally posted by momaka View Post
        Hopefully he remembered. Glad I wasn't a person to give him my computers for repair. But after all, everyone has to start somewhere, right.
        Gee I don't know man.
        Some things you just have to know before you start.
        Otherwise you are a danger to yourself and those around you.

        But as always momaka a great read and cool fix, well done!
        And I guess we have that "tech" to thank for it,
        Just like I have the cheapass capacitor manufacturers to thank for allot of the equipment I have
        "The one who says it cannot be done should never interrupt the one who is doing it."

        Comment


          #5
          Re: Some PlayStation 3 fan repairs, just for fun

          Originally posted by Per Hansson View Post
          Gee I don't know man.
          Some things you just have to know before you start.
          Otherwise you are a danger to yourself and those around you.
          Yeah, that's what I thought as well.
          However, a few months afterwards, he got a lot better - at least in terms of computer repair. His AC power electronics knowledge was seriously lacking, but he wasn't doing anything with that anymore. Instead, he was doing mostly repairs on laptops. And surprisingly, he could actually read some board schematics and figure out how to find short circuits on buck rails and bad charger chips on some boards.

          Originally posted by Per Hansson View Post
          But as always momaka a great read and cool fix, well done!
          Thanks!

          I know it was a long read, but sometimes I forget things and need to read my old posts to re-educate myself. Also, a few years back when Th3_uN1Qu3 was more active here, he made some very long posts with similar repairs, and I've learned quite a bit from them. So I figure it still may be worth posting, if someone had the time to read these.

          Originally posted by Per Hansson View Post
          Just like I have the cheapass capacitor manufacturers to thank for allot of the equipment I have
          Yes!
          Thank you CapXon, Teapo, OST, and GSC/Evercon/Sacon!
          These cap brands (along with Chemicon KZG) have given me a lot of good cheap PC hardware scores.
          Last edited by momaka; 07-23-2017, 05:43 PM.

          Comment


            #6
            Re: Some PlayStation 3 fan repairs, just for fun

            I have a feeling your tech comes from the same class of engineers that build cheap Chinese 240V to 120V "step-down transformers" using a single diode. Because diode means half the voltage, on AC, right?
            Please do not PM me with questions! Questions via PM will not be answered. Post on the forums instead!
            For service manual, schematic, boardview (board view), datasheet, cad - use our search.

            Comment


              #7
              Re: Some PlayStation 3 fan repairs, just for fun

              Lol.

              No, the guy actually did not have any degree in electronics. Just fixing computers out of necessity to support himself and his family. I've lost contacts with him. But last I saw him, he had learned quite a few things by himself by reading.

              But yeah, I've seen those tiny step-down "transformers" / AC travel "converters". Funny at best. Dangerous, at least.

              Comment

              Working...
              X