Announcement

Collapse
No announcement yet.

Need help diagnose dead Sunshine ATX235 power supply

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

    Need help diagnose dead Sunshine ATX235 power supply

    Got this non working ATX power supply from a garage sale. The owner is not too sure if it is working. I noticed it is label as Sunshine Model ATX235 Series Rev 5. After opening the case,i saw THR 1 is empty;so I soldered in a SCK084 thermistor. Checking other area I don't see any burn marks and the 5A 250V fuse tested good in continuity.

    Before I switched on the ps,I soldered in a 60W light bulb in series in place of the fuse. By shorting the green wire to black ground wire, I switched on ps, the light bulb glows shortly then dims off. I checked the purple wire with a DMM, it reads 4.9VDC meaning there is standby power. But checking on yellow wire or red wire doesn't show any voltage on DMM.
    I also checked on the 16pin chip pwm IC KA7500B, put my red probe to pin 12 while black probe to pin 7 shows 5VDC.
    Then I unsoldered all the secondary caps and I make sure ESR reading on yellow rail caps of 470UF 16V tested good since ESR reads<0.20 ohm.
    The caps on black rail of 680UF16V & 680UF10V ESR reads<0.10 ohm while the caps on red rail of 680UF10V tested good since ESR reads <0.10 ohm.

    I also tested ESR on smaller caps on C27(1UF50V reads 3 ohm),
    C33(0.68UF50V reads 3.5ohm),C40(4UF50V reads 1.4 ohm) & C45(1UF50V reads 3.4 ohm). All ESR reading on these four small caps are within good range.
    By setting my DMM to diode test, off board the diode FR103 on D15&D19 along heat sink2 reads 466 on one side while "1" on the other side. Both diodes are not shorted or open.
    Then I did diode test on 3 transistors on heatsink 1 which are C5027, and 2 units C2335F. The diode test doesn't find any short or beeps.
    I did diode test on schottky rectifier on heatsink2 which are double diode(can't see the type), PSR16C40CT and MOSPEC S10C40C. Black probe to cathode middle pin and red probe to left and right anode pin
    both shows around 79 ohm. The basic result seem to conclude there is no shorts. Refer to img 1897 on back of pcb, I saw a connecting black wire and
    zener diode and wonder if this is supposed to be here?

    But I have no idea how to test the two big main cap and the black square bridge rectifier. FYI, on hand I have a digital multimeter,a 40W soldering gun, a china made ESR tester(M Tester v2.07) and a digital capacitor meter.
    I will appreciate anyone kind and generous to advice on what to do next to revive this ps.
    Attached Files

    #2
    Re: Need help diagnose dead Sunshine ATX235 power supply

    replace the small caps.

    the psu uses them in a way that requires the ripple capacity more than the low esr.

    Comment


      #3
      Re: Need help diagnose dead Sunshine ATX235 power supply

      If there are no shorts, the primary transistors could have failed open. Replace both C2335F transistors and see where that gets you. As replacements you can use 13007 transistors or anything else you may find in the same position on a similar power supply.
      Originally posted by PeteS in CA
      Remember that by the time consequences of a short-sighted decision are experienced, the idiot who made the bad decision may have already been promoted or moved on to a better job at another company.
      A working TV? How boring!

      Comment


        #4
        Re: Need help diagnose dead Sunshine ATX235 power supply

        Hi stj & Th3_uN1Qu3,

        Before I go replace the small caps and primary transistors, I have the following question:

        Since there is 5V STBY(standby) power, do I need to recheck voltage on two big main caps and the black square bridge rectifier to make sure they are both working?

        How do I test using a DMM ? Appreciate your guide and advice. Thanks for reply.

        Comment


          #5
          Re: Need help diagnose dead Sunshine ATX235 power supply

          Hi all,

          Replaced all the small caps and two transistors C2335F with new 13007. Tested voltage
          on purple and black wire,it reads 5.03VDC.(standby power)

          This time after parts replacement, got merely 0.03VDC on red (5V)and black wire.
          Is this some improvement since the ps got a tiny 0.03VDC now?

          Where do I go from here in order to get the full power supply to work fully.
          Any advice please and thank you for reply.

          Comment


            #6
            Re: Need help diagnose dead Sunshine ATX235 power supply

            Hi all,

            Replaced all the small caps and two transistors C2335F with new 13007. Tested voltage on purple and black wire,it reads 5.03VDC.(standby power)

            This time after parts replacement, got merely 0.03VDC on red (5V)and black wire.
            Is this some improvement since the ps got a tiny 0.03VDC now?

            Where do I go from here in order to get the full power supply to work fully.
            Any advice please and thank you for reply.

            Comment


              #7
              Re: Need help diagnose dead Sunshine ATX235 power supply

              Hi,

              Sorry double post No.6

              Attached photos on repair done.
              Attached Files

              Comment


                #8
                Re: Need help diagnose dead Sunshine ATX235 power supply

                In the opening statement you have posted some ESR measurements and concluded that they are all normal, but to me they all seam wrong, for example the 470u/16V on the 12V rail, you sayd you got 0,2 Ohms, but as you probably know a smps filters it's outputs with low esr caps, and 0,2 Ohms is not a low esr standard vallue, you should always want less than 0,1 Ohms for that vallue ( 470u ), preferably under 0,05, especially on the larger capacity ones. Another example is the small caps, over 3 Ohms for 1uF cap does not seam a good vallue to me... I suppose all the measurements have been done with the caps removed from the board. This is what you should do with the large caps on the primary side too, remove and measure them, you should find a capacity vallue close the one written on it ( within the tolerance level ), and the ESR should not be greater than 0,2 Ohm. The bridge rectifier can be measured on the board, you set the DMM on the diode function and you check each rectifier's diode.

                Another thing, you saud you have measured 5V on the 12 pin of 7500 chip, that is way too low for that IC, it is basically TL494 and it's UVLO is at about 8V, you really should first make sure you have the correct voltage at that pin ( about 12V ), without it the main converter will not work because of the UVLO of the controller. That 12V comes from the auxiliary converter ( the same as the 5Vsb ), so you can check it without powering up the main converter, when you have it, them change all the filter caps on the outputs, then try powering up the main converter again and see what you get.

                Comment


                  #9
                  Re: Need help diagnose dead Sunshine ATX235 power supply

                  Hi Marianelforum,

                  Referring to 16pin chip pwm IC KA7500B which you said "That 12V comes from the auxiliary converter ( the same as the 5Vsb ), so you can check it without powering up the main converter,......"

                  Can you guide me on how to check the auxiliary converter; like DMM black probe to ground and red probe to.....(and where is it on my board).
                  Sorry I do not have any idea or knowledge on this but I can do
                  voltage measurement like on the IC KA7500B.

                  Glad you can reply and assist.
                  Thanks.

                  Comment


                    #10
                    Re: Need help diagnose dead Sunshine ATX235 power supply

                    Your pictures are not clear enough to be of much help, but you said that after changinc some caps you found a good 5Vsb voltage, have you checked again the voltage at pin 12 of the 7500 IC after changing those caps?

                    You should follow back the trace for the controller power, put the black wire of the dmm on the output ground, and with the red one, probe at different points of the power trace back to the auxiliary transformer's secondary, you should find a rectifier diode from one of it's pins leading to a filter capacitor wich can be up to 100uF or so... Since you have corrected the 5Vsb voltage, the 12V for the controller should be ok as well, since it's coming from the same transformer and it's regulation is an indirect one.

                    Comment


                      #11
                      Re: Need help diagnose dead Sunshine ATX235 power supply

                      Hi Marianelforum,

                      With mains AC220V switch on, I measured the 16 pin KA7500B IC by putting DMM red probe to pin 12 & black probe to pin 7, DMM reads 23.4VDC. I try not to do this often working on live AC as my fingers are shaking ;and being careful the probe do not touch any other parts as AC is live.

                      Is the reading too high as I read from Fairchild site on the electrical characteristics of VCC on IC KA7500B SMPS Controller is 20V.
                      I rechecked voltage on purple and black wire on white connector,it reads 5.03VDC.(standby power). FYI, I checked the two big main caps, the DMM reads at 155VDC each.
                      Quoting from your earlier post "Since you have corrected the 5Vsb voltage, the 12V for the controller should be ok as well, since it's coming from the same transformer and it's regulation is an indirect one." But I am still puzzle why there is no 12V.
                      Also quoting from your post "That 12V comes from the auxiliary converter ( the same as the 5Vsb ), so you can check it without powering up the main converter, when you have it, them change all the filter caps on the outputs, then try powering up the main converter again and see what you get.

                      Can you explain further on how to check the auxiliary converter without powering up the main converter? Sorry with my limited experience I am not able to identify which one is the auxiliary and which one is main converter.
                      And also how do I check it. I know I am slow in this but I am willing to learn
                      from the expert.

                      Many thanks for your patience and guidance.

                      Comment


                        #12
                        Re: Need help diagnose dead Sunshine ATX235 power supply

                        I am not an expert, just a knowledge seeker as yourself, it's just that i've seeked it maybe a longer time than you...

                        Let's clarify a few thing first. The main converter is the one that powers up when you connect the green wire to a black one, and delivers the main outputs ( +12V, -12V, +5V and +3,3V ). The auxiliary converter is the one that powers up just as you connect the unit to the line voltage, meaning that is always on, and it delivers the 5Vsb output, it also powers up the main converter's controller IC, wich in your case is that 7500B one.

                        Without the main converter activated ( without the green wire connected to a black one ) there is not much damage you could do by probing around in the safe side ( the secondary side ) just because all the existing power comes only from the auxiliary converter, and even the simplest one have a decent current limiting.

                        23V on the 12 pin of 7500 is way to much, either your dmm is faulty, or something is wrong with the auxiliary converter, sadly there isn't much i can help you with, without some high resolution pictures, so you have to work it out by your self, i can only point out basic stuff.

                        Comment


                          #13
                          Re: Need help diagnose dead Sunshine ATX235 power supply

                          23v is fine. Look for the problem elsewhere. Since you say you have replaced primary transistors. Try replacing driver transistors as well. These should be two TO-92 packages physically located in the secondary, next to the small transformer in the middle, and if you look at the traces from their bases they will lead to the outputs of the KA7500B.

                          I would have directed you to check the ramp and outputs of the KA7500B but you won't be able to do that without an oscilloscope. But many times if the primary transistors blow, the drivers go as well.
                          Originally posted by PeteS in CA
                          Remember that by the time consequences of a short-sighted decision are experienced, the idiot who made the bad decision may have already been promoted or moved on to a better job at another company.
                          A working TV? How boring!

                          Comment


                            #14
                            Re: Need help diagnose dead Sunshine ATX235 power supply

                            Originally posted by Th3_uN1Qu3 View Post
                            23v is fine...
                            Based on what?

                            Comment


                              #15
                              Re: Need help diagnose dead Sunshine ATX235 power supply

                              The main power supply does not run so there is no power used by the driver transistors. These pull more than the KA7500B/TL494 IC so when the PSU is fully up and running, that may well bring the voltage down.

                              Either way, the maximum allowed supply voltage for the KA7500B is 42 volts, so i don't see why that voltage should be 12 volts and why 23v would not be fine. Read the datasheet. I don't recall that many power supplies based on TL494/KA7500B with a controller Vcc of 12v. 15v and up would be the norm usually.
                              Originally posted by PeteS in CA
                              Remember that by the time consequences of a short-sighted decision are experienced, the idiot who made the bad decision may have already been promoted or moved on to a better job at another company.
                              A working TV? How boring!

                              Comment


                                #16
                                Re: Need help diagnose dead Sunshine ATX235 power supply

                                I disagree, because i did fixed many ( many ) ATX psu's of many types, i do not remember finding one with 23V or even close on the pin 12 of 494/7500, the most i found was almost 14V on one case, the majority had between 12 and 13V, all of them measured with the main converter off. As i have said the power for the IC comes from an auxiliary secondary of the auxiliary converter and it is regulated, indirectly yes, but still regulated by the turn ratio of the 5Vsb secondary to this one, the regulation depends on the leakage inductance, but it cannot be so bad that it can drop from over 20V down to less than 15V, i do not think it would do that even unregulated, the driver transistors still do not take that much power.

                                I did read the datasheet of this IC long before i even heard of this forum, so thanks for the advice but it is not needed. I am not saying that it could not work with 23V, but that it is not needed, and trust that i do know what i am saying. 12-14V maximum is all it needs no matter what it drives. In this case it drives a couple of small bjt transistors, and it does so with common emitters so the drivers take their power directly from the pin 12 ( or rather from the power delivered there ). The drivers could do just fine with even 2.5-3V in the base, having 23V drive voltage only serves to heat up real bad the limiting resistors in the bases, and more so with the main converter off, when at least 20V drop on those resistors continuously because the IC's outputs are off ( meaning that with the main converter off, both driver transistors are always ON ). So again it would work with 23V too, but it's more than needed so i am sure something is not ok there.

                                PS: Please don't be like Leco... You do not need to bold up the text for someone to see it, i can read it just fine as stadard. Thank you
                                Last edited by Marianelforum; 08-19-2016, 04:39 AM.

                                Comment


                                  #17
                                  Re: Need help diagnose dead Sunshine ATX235 power supply

                                  I was just pointing out that 23 volts on TL494 Vcc is not an indication of a fault. Let's get back on topic.

                                  OP please report back after replacing driver transistors.
                                  Originally posted by PeteS in CA
                                  Remember that by the time consequences of a short-sighted decision are experienced, the idiot who made the bad decision may have already been promoted or moved on to a better job at another company.
                                  A working TV? How boring!

                                  Comment


                                    #18
                                    Re: Need help diagnose dead Sunshine ATX235 power supply

                                    Hi,

                                    There's like 4 or 5 driver transistors with marking "C945" NPN general purpose
                                    transistor which I need to desolder for off circuit testing. Is the proper way to test these trans by setting DMM to diode mode,red probe to B,black probe to C(middle pin),get the reading,then red still at B,black probe to E. get second reading. Am I right to say the trans is good if both reading is more or less about the same?

                                    Is there any easier way to test for DC voltage on board before I start deolder all the trans.? Like red probing on say anode of diode after the small transformer, or probing on middle pin of primary transistor while black probe on cold ground.. just curious to know instead of doing another round of desoldering parts.

                                    Comment


                                      #19
                                      Re: Need help diagnose dead Sunshine ATX235 power supply

                                      I think the drivers are the 2 above the KA7500B controller in picture #6 of your first post. You can confirm this by tracing with your multimeter, see if the bases are connected (through a resistor) to pins 8 and 11, which are the collector outputs of the KA7500B.
                                      Originally posted by PeteS in CA
                                      Remember that by the time consequences of a short-sighted decision are experienced, the idiot who made the bad decision may have already been promoted or moved on to a better job at another company.
                                      A working TV? How boring!

                                      Comment


                                        #20
                                        Re: Need help diagnose dead Sunshine ATX235 power supply

                                        I desolder the two driver transistor from board position at Q12&Q13-these
                                        are above the KA7500B controller and another one from Q7. The three trans are with
                                        label C945 which are TO-92 NPN Transistor; looking at front side with emitter(E) on left,collector(C) in middle pin and base(B) on right pin.

                                        Using diode test setting, black probe to B while red probe to E, DMM reads 1 (infinite),
                                        while black probe to B while red probe to C, DMM reads 1 (infinite).
                                        Then red probe to B while black probe to E, DMM reads 593; while red probe to B and black probe to C DMM reads 594. All 3 trans read almost similarly between 591 to 594.
                                        The way I see it, the driver transistors are good. Do correct me if I am wrong.

                                        I also desoldered two small caps at C41 (2.2uF 50 V) & at C35 (1UF 50V) and the big coil. refer attached picture. The small caps reads 3 ohm on ESR; are they alright?
                                        The big coil has continuity. I also desolder one end of the two diodes hidden below the big coil; but diode test showed 467 ohm at one side and swapping probes, DMM reads infinite. Then the result mean the diodes are good.

                                        I am stumped now and do not know what test to do. The 5VSB is present but still no 12V and 3.3V rail. Any tip or advice to revive this ps is very appreciated.
                                        Attached Files

                                        Comment

                                        Working...
                                        X