Announcement

Collapse
No announcement yet.

Task TK-940TX-DF Blowing fuse

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

    #21
    Re: Task TK-940TX-DF Blowing fuse

    Also tested across R23A, the small resistor that's in between the primary caps, and got a short.
    One leg of it goes to -ve of primary and the other leg connects to the ferrite beads.
    The resistor tests OK out of circuit.
    Last edited by socketa; 05-06-2020, 05:38 PM.

    Comment


      #22
      Re: Task TK-940TX-DF Blowing fuse

      Originally posted by socketa View Post
      Replaced the 5VSB switching transistor K3067 (which had low resistance between D and S in both directions (i probably destroyed that as well)) with a CEF02N6A (not quite the same, but close enough?).
      Yes, that CEF02N6A is actually a pretty good replacement (slightly better than the original, in fact! )

      Originally posted by socketa View Post
      Powered it up (without the TL3842P, R22, 2 X K2746) with the 60W lightbulb replacing the fuse, and get a steady glow on the light bulb, and a reading of 56W and 0.24A (which is what it also reads with just the light bulb by itself) from my portable power plug meter.

      I also confirmed, with a multimeter, that there was no short between phase and neutral or earth.
      Going by these results, that sounds like either the 5VSB CEF02N6A MOSFET is getting stuck On (not driven off by the C945 BJT transistor) or there is a low resistance across the primary caps... which is unlikely, because you said you tested the MOVs across the caps and found they are not reading short-circuit / low resistance.

      Next up, just to confirm where the steady glow on the bulb is coming... Try removing only the CEF02N6A MOSFET and try the incandescent bulb test again. If should light up for a second or less and then go out as the primary caps charge. If it stays lit, the problem is not the 5VSB. But if not, then we have to see why the CEP02N6A MOSFET is getting stuck On (most likely never driven off by the C945 transistor, which can be due to a number of reasons.)

      Originally posted by socketa View Post
      [P.S. Tangential to this, but wondered why the voltage divided by the resistance (3.6A) of the lightbulb wasn't anywhere near the actual reading (0.24A)
      The resistance of the light bulb varies.
      As the Tungsten element inside heats up, its resistance increases. This actually also applies even to NiChome wire, as well as any conductor really.

      The actualy reading of the current you got agree with the power rating of the light bulb: 0.24 Amps x 240V = 57.6 Watts (which is close to 60W... though the current accuracy measurement of your multimeter may be slightly off.)

      Originally posted by socketa View Post
      Also tested across R23A, the small resistor that's in between the primary caps, and got a short.
      One leg of it goes to -ve of primary and the other leg connects to the ferrite beads.
      The resistor tests OK out of circuit.
      Can you double-check your results and leave the multimeter for at least 2-3 seconds on R23A while it is in circuit to get the resistance? Do you really get a short-circuit (2-3 Ohms or whatever is the lowest your MM can measure) or 47 Ohms? Reason I ask is because I'm trying to figure if there is a possible short-circuit path from the Gates of the main PS MOSFETs to ground.

      You see, R23A (1 KOhm) is connected between -Ve and the Gate of one of the main PS MOSFETs, while R12 (also 1 KOhm) is connected between -Ve and the Gate of the other MOSFET for the main PS. With both MOSFETs removed, the only possible short-circuit path is through the ferrite beads and jumper J1, then out to either R7 (47 Ohms), Zener Diode ZD1, and Diode D1. To isolate which one of these may be causing the "short-circuit", you can remove them one at a time and re-take the measurements across R23A. Please let me know what you find here.

      Additionally, can you remove the two gray parallel diodes in front of the big transformer, labeled with DX10 on the PCB, along with the small blue cap next to them? These form the snubber circuit for the main PS. Please check both the diodes and the small blue cap for short-circuit as well as proper parameters (diodes should have good forward diode reading on your TT meter, and the blue cap should be a few pF in capacitance.)
      Last edited by momaka; 05-10-2020, 08:12 PM.

      Comment


        #23
        Re: Task TK-940TX-DF Blowing fuse

        Thanks for the explanation.
        With the 5VSB MOSFET removed - the bulb still glows bright and stready.
        And, yeah, there was definitely a short measured across the in-circuit resistor R23.
        I removed zener diode, ZD1, and it tests as a shorted; and the short across R23A dropped right off.
        Will it be ok to replace the shorted zener with a same-looking zener from another parts PSU?

        I wont do any more removing of components and testing (that you suggested) at this stage; as this shorted zener could be the cause of the problem.

        What do you reckon?
        Last edited by socketa; 05-14-2020, 02:25 AM.

        Comment


          #24
          Re: Task TK-940TX-DF Blowing fuse

          Originally posted by socketa View Post
          With the 5VSB MOSFET removed - the bulb still glows bright and stready.
          And, yeah, there was definitely a short measured across the in-circuit resistor R23.
          So something is drawing power across the primary cap. Still have to determine if it's the 5VSB or a component from the main PS... So let's dig a little deeper....

          Originally posted by socketa View Post
          I removed zener diode, ZD1, and it tests as a shorted; and the short across R23A dropped right off.
          Will it be ok to replace the shorted zener with a same-looking zener from another parts PSU?
          No, don't replace it. Just leave it off for now. We will determine what value it needs to be later. By the looks of it, it might be a Gate over-voltage protection for the main PS MOSFETs. So most likely it is rated for 20V or less, but we will see (might need another solder-side picture of that area to determine that.)

          Originally posted by socketa View Post
          I wont do any more removing of components and testing (that you suggested) at this stage; as this shorted zener could be the cause of the problem.
          No, the shorted Zener is probably part of the problem. But since it appears connected to the Gate drive of the main PS MOSFETs, and since the main MOSFETs are removed, then there shouldn't be a short to ground. But there must be if the bulb is glowing steadily. Thus, once again, I suggest removing the two paralleled diodes (DX10?) and the blue ceramic cap in front of the big transformer. Remove them and test them out of circuit. If they show good value, remove diode D3 (located next to resistor R7.) Diode D3 takes power from the auxiliary primary rail generated by the 5VSB transformer. If that rail is shorted to ground due to over-current, then that could be causing the 5VSB circuit to not work properly and make the bulb glow too. Thus, removing D3 will isolate the 5VSB circuit on the primary side from the main PS primary-side circuits (which is what we want - we want to see if 5VSB is working alone first.)
          Last edited by momaka; 05-27-2020, 07:00 PM.

          Comment


            #25
            Re: Task TK-940TX-DF Blowing fuse

            Ok
            I just removed those three components (which tested OK) and the lightbulb still glows

            Comment


              #26
              Re: Task TK-940TX-DF Blowing fuse

              Originally posted by socketa View Post
              Ok
              I just removed those three components (which tested OK) and the lightbulb still glows
              So you have the 5VSB MOSFET and these (four!) components removed, and the bulb still glows??

              You should only see the bulb glow for a moment (up to a second or so) until the primary caps charge. After that, it should go out. If not, you have a short-circuit or low-resistance path on the primary somewhere.

              I'm not sure what else to suggest to check, because there are only so many paths between the primary-side DC bus across the two caps. Namely, that is:
              1) A series connection of the main transformer primary side, main PS MOSFET, and Source resistor of main PS MOSFET
              2) A series connection of the 5VSB transformer primary side, 5VSB MOSFET, and Source resistor of 5VSB MOSFET.

              Since you have all of the MOSFETs removed... and in addition to that, you also have the main PS snubber network removed, there really is nothing else that connects the + and - on the primary DC bus. So the bulb should not glow.

              But yet, it does. This means you must be missing something here that I simply possibly cannot see from the pictures.

              The only last thing I can think of: you live in a country with 230V AC mains. Is the selector switch set for 230V input?
              Last edited by momaka; 06-10-2020, 08:03 PM.

              Comment


                #27
                Re: Task TK-940TX-DF Blowing fuse

                No power selector on this PSU - so that rules that out.
                Attached a photo with the components removed, just to confirm that i removed the right ones.
                Additionally, there is 1 Meg ohm resistance measured across phase and neutral - is that OK? And would that cause the light bulb to glow brightly?
                Attached Files

                Comment


                  #28
                  Re: Task TK-940TX-DF Blowing fuse

                  Originally posted by socketa View Post
                  Attached a photo with the components removed, just to confirm that i removed the right ones.
                  Yup, you got them.

                  Originally posted by socketa View Post
                  Additionally, there is 1 Meg ohm resistance measured across phase and neutral - is that OK?
                  Yes, that is normal.
                  Usually, there is a high-value resistor between Live and Neutral to discharge the X2 caps. 1 MOhm is a very typical value.

                  Originally posted by socketa View Post
                  And would that cause the light bulb to glow brightly?
                  Well, there aren't that many paths left, but possibly there is still something shorting on the DC/rectified side of the primary side.

                  Discharge the primary caps (putting your soldering iron's wall plug across each cap's leads for a few seconds will do the trick), then measure resistance between (+) and (-) leads on the bridge rectifier. Keep the multimeter connected for a few seconds, as the capacity of the input caps may show a low resistance at first. With this test done, if the resistance raises and is not low (over 200 Ohms), remove the two orange MOVs. I know you mentioned you measured them and they aren't shorted, but perhaps they could be conducting at live voltage. Now try plugging in the power supply and see if the bulb still glows. If yes, there are only two more possible paths from faulty components that can cause this.

                  The first one is the blue ceramic cap between the 5VSB transformer and 5VSB MOSFET (i.e. next to that big ceramic resistor.) Remove that and try the bulb test again.

                  Bulb still glowing? If yes, the last possible path is... and this is very unlikely... that one or both of the primary caps are bad and arcing internally when charged. Remove both caps and try the bulb test again. If the light bulb still glows, take a good picture of the bottom side and post it here. I know there is one on page 1, and I've been using it extensively to trace the circuit on this PSU. But just to make sure something isn't shorting there or something hasn't changed in an unexpected way since that last image.

                  So again, just a run down of the procedure:
                  1) discharge primary caps and test resistance between + and - of bridge rectifier. Post value here.
                  2) remove MOVs and perform incandescent bulb test again. If NOT glowing, STOP. The MOVs are likely bad. Otherwise continue below.
                  3) Remove blue ceramic cap next to white ceramic resistor near 5VSB MOSFET area. Perform incandescent bulb test after this. If NOT glowing, STOP - ceramic cap is likely bad and failing when charged. Otherwise, continue below.
                  4) Remove both primary caps and perform incandescent bulb test. Post here if bulb still glows or not.

                  Also, between each step above, it probably IS good idea to try and discharge the primary caps as detailed earlier with a soldering iron plug or something similar. That way, if you get to one of the test steps above where the bulb is no longer glowing because the short on the primary is gone, then the PSU will likely end up with its primary caps charged at high voltage (about 170V DC each) - at least for a few moments/minutes until the discharge/balance resistors across the caps discharge them. Thus, for extra safety, I like to make sure those primary caps are discharged.
                  Last edited by momaka; 06-14-2020, 07:36 PM.

                  Comment


                    #29
                    Re: Task TK-940TX-DF Blowing fuse

                    Finally got around to doing this (Johnny-ten-jobs at the moment! - trying to get as much stuff done as possible, just in case TPTB have more tricks up their sleeves):
                    Went through all of your 'flow chart' four steps, and the bulb only stopped glowing (thank goodness) when i removed the primary caps.
                    So it appears that this is the "unlikely" cause of the problem?
                    Both of the capacitors test about 740uf and have and ESR of 0.05
                    With the multi-meter, +ve lead to +ve terminal, -ve lead to -ve terminal, it rises to 37Megohms and then slowly starts to fall, and for the other cap it rises to 40Megohms and then slowly starts to fall.
                    So primary capacitors can test OK and still arc internally?

                    Thanks for the safety info - when i was at school, years ago, we were charging up capacitors, and they guy next to me discharged one on his face or mouth (i cant remember which) and he said that he saw a flash of light.

                    Comment


                      #30
                      Re: Task TK-940TX-DF Blowing fuse

                      Originally posted by socketa View Post
                      Went through all of your 'flow chart' four steps, and the bulb only stopped glowing (thank goodness) when i removed the primary caps.
                      So it appears that this is the "unlikely" cause of the problem?
                      Both of the capacitors test about 740uf and have and ESR of 0.05
                      Try measuring the resistance of both caps with a multimeter instead of ESR meter.

                      If neither shows shorted, check the resistance across each cap spot on the PSU PCB. See if anything there measures low resistance (please post what resistance you get.)

                      If resistance across either spot is several KOhms or more, try putting different caps in the PSU (and mind the correct polarity, because otherwise you may get some magic smoke even with the series light bulb ). The caps don't need to have the same capacitance, just the same or higher voltage rating. Anything in the range of 220 to 1000 uF will do fine, though the caps do need to be a matching pair (as in, both the same capacitance and voltage.) After this, plug in the PSU and see if the bulb glows again.

                      If it does, there is still something shorting across one or both caps and the problem is something on the PSU. Thus, the original caps may be fine, if that is the case. As requested in my last post, again, it would be great if you can provide another PCB shot of the solder side, preferably more focused and looking at a straight angle at the PCB (so I can see if we missed a possible path somewhere that can be causing a short-circuit.)

                      Originally posted by socketa View Post
                      Thanks for the safety info - when i was at school, years ago, we were charging up capacitors, and they guy next to me discharged one on his face or mouth (i cant remember which) and he said that he saw a flash of light.
                      Ya, big HV caps can store quite a bit of energy. If charged and shorted directly with a metal object, they will definitely produce a large spark and loud pop.(anything over 30V DC and 1000 uF starts to get loud very quickly.) The problem with that is not so much that the spark can hurt you, but rather the loud noise catching you off guard - with the danger being pulling your hand/head/body away and hitting/scraping/cutting yourself in the process. You can actually see that on Electroboom's intro video. Of course, that is scripted, as are all of his other videos, but it's actually a pretty realistic outcome.

                      That's why I suggested discharging the caps though the plug on a soldering iron or something similar that can take the high DC voltage but not draw much current, and hence discharge the caps very "gracefully".
                      Last edited by momaka; 07-02-2020, 07:53 PM.

                      Comment


                        #31
                        Re: Task TK-940TX-DF Blowing fuse

                        Try measuring the resistance of both caps with a multimeter instead of ESR meter.
                        After discharging them and putting the multimeter across them, they both quickly increase in resistance and go up to (according to my meter) just over 40MΩ, then back down slowly to 35MΩ and 29MΩ respectively.
                        If neither shows shorted, check the resistance across each cap spot on the PSU PCB. See if anything there measures low resistance (please post what resistance you get.)
                        There's no low resistance (there's actually ∞ resistance). I also made sure that test path wasn't blocked by any flux residue, by checking continuity to another point on the same trace.
                        If resistance across either spot is several KOhms or more, try putting different caps in the PSU
                        Got some 680uf 200V caps from another PSU and soldered them in.
                        But the light bulb still fecken (long live "Father Ted") glows.
                        Perhaps there is a "free radical" inside of the circuitry?
                        Attached Files
                        Last edited by socketa; 07-05-2020, 05:57 PM.

                        Comment


                          #32
                          Re: Task TK-940TX-DF Blowing fuse

                          Originally posted by socketa View Post
                          Perhaps there is a "free radical" inside of the circuitry?
                          Maybe.

                          I think we are just getting closer, though. Hope you still haven't given up yet.

                          So the last test with the new primary caps confirmed the old ones are probably fine.

                          I analyzed the new image you posted (of the solder side) and determined there is only one last possible path: shorted snubber ceramic cap on the 5VSB primary side. Actually, there are two snubbers on this 5VSB, it appears. One is the standard series connection of a diode with a cap and resistor in parallel. The other is just a small (blue) ceramic cap connected between Source and Drain of the (removed) 5VSB MOSFET. It's possible that this cap is shorted, so remove it and test it. Let me know what results you get. Also try the incandescent bulb test again after you remove this component, regardless if it tests OK or not. If the bulb still glows, remove the 5VSB transformer and try the bulb test again. Still glowing or not? If still glowing, "free radicals" and small gremlins inside the circuit are confirmed. I doubt it, though. If you had to remove the 5VSB transformer for the bulb to stop glowing, test resistance on the primary side of the 5VSB traffo. Using your last image above of the solder side as a reference, and starting from the top two pins on the 5VSB transformer, these two should show low resistance to each other. Likewise, the bottom two pins should also show low resistance to each other. However, the top two and bottom two pins should NOT show any resistance to each other (or near infinite, if you're touching the metal parts on the pins or multimeter probes with your fingers, which can happen to show a high resistance instead of open circuit ).

                          So just to recap the steps:

                          1) Remove blue snubber cap connected between Source and Drain pins on 5VSB MOSFET and test it out of circuit. Post results for this step.
                          2) Try incandescent bulb test. If it is NOT glowing, STOP. If it is still glowing continue to step 3.
                          3) Remove 5VSB transformer. Check resistance between pins and post results here.
                          4) Perform bulb test again. If glowing, STOP and clean all PCB tracks thoroughly with IPA and make sure there are no stray solder blobs shorting things. In addition to that, remove heatshirnk-covered resistor next to 5VSB MOSFET and do the incandescent bulb test again (though I doubt this will change anything.)
                          .
                          .
                          .
                          5) You've been defeated by the circuit gremlins
                          .
                          .
                          Or at least for now. Let's see what results you get, as that may give us a different clue.
                          Last edited by momaka; 07-11-2020, 10:29 PM.

                          Comment


                            #33
                            Re: Task TK-940TX-DF Blowing fuse

                            Took an extra day to do this because my computer wouldn't talk to the router, so i found a spare one and couldn't be bothered checking the volts and amps, and used the same power adapter and killed the adapter because it's output was AC and i plugged it into a router that needed DC (Some derivitive of Murphys law, since this was the first time that i've ever noticed AC output on a power adapter) - So it seems like the way to destroy a power adaptor is to make sure that its output is AC then plug it into a DC device. So AC and DC are not compatible unless you are Bonn Scott (or Brian before his voice broke down).
                            After sorting out another router, the same problem occured which pointed to the computer being the source of the problem.
                            So i had to sort out another computer, which i was going to do anyway because either firefox (and it's few forks) or websites these days demand 2 Cores and more than 2GB of RAM (good grief, all that i want to do is read posts and watch internet videos at 240p or 360p) or they will cause the computer to lock up.
                            So i've got something like 30 462/754 boards that are now practically non-internetable.
                            i was happy with dial-up, buffering, and waiting for the video to load.
                            Take me back 20years and push hold.


                            1) 210pF
                            2) Still Glowing
                            3) Removed but didn't test because the bulb still glows after its removal
                            4) Still glowing (why am i not surprised?) after removing the transformer and the resistor(1Ω)

                            Just brought some keyboard cleaner today - so that was good timing.
                            Cleaned it as much as i could, and couldn't spot any blobs or spashes that would have caused a short.
                            How the heck is the bulb glowing when all that i'm getting between phase and neutral (of the power plug socket), with fuse inserted, is 1MΩ (that "discharge resistor")?
                            If i just connect the bulb in series with a 1MΩ resistor, and nothing else, would it glow like that?
                            Attached Files
                            Last edited by socketa; 07-14-2020, 03:41 AM.

                            Comment


                              #34
                              Re: Task TK-940TX-DF Blowing fuse

                              Originally posted by socketa View Post
                              ...
                              4) Still glowing (why am i not surprised?) after removing the transformer and the resistor(1Ω)

                              Just brought some keyboard cleaner today - so that was good timing.
                              Cleaned it as much as i could, and couldn't spot any blobs or spashes that would have caused a short.
                              How the heck is the bulb glowing when all that i'm getting between phase and neutral (of the power plug socket), with fuse inserted, is 1MΩ (that "discharge resistor")?
                              Good question.

                              I double and triple-checked everything from the last photo you uploaded and there are no possible low-resistance/impedance paths left to ground. That is, again, assuming you have removed the MOVs behind the bridge rectifier.

                              This only leave the bridge rectifier and/or possibly something wrong with your incandescent bulb setup.

                              Do this first: take a known good ATX PSU and connect it to your series incandescent bulb. Make sure there is nothing connected to the output of the PSU (i.e. no motherboard or anything.) Now plug it and see what happens. Does the bulb still glow?

                              If not, then remove the bridge rectifier on the Task PSU and replace with another bridge rectifier, if you have a spare one. Otherwise, you can use 4 regular diodes too in place of the bridge rectifier (but they need to be rated for at least 400V DC reverse blocking voltage, or more.)

                              Originally posted by socketa View Post
                              If i just connect the bulb in series with a 1MΩ resistor, and nothing else, would it glow like that?
                              Well, try it and see.

                              Hint: before trying, think about the resistance of the bulb relative to the 1 MOhm resistor and the total current that will flow through this circuit. That might give you a clue of what results to expect.
                              Hint 2: the results will be rather uneventful.

                              Originally posted by socketa View Post
                              So it seems like the way to destroy a power adaptor is to make sure that its output is AC then plug it into a DC device.
                              Yup.
                              If it's an AC adapter, it's probably just a simple line-connected transformer with no diodes or capacitors on the output, hence outputting at 50/60 Hz AC (depending on country line frequency.)

                              Your router probably has a reverse-voltage protection diode inside, which likely shorted the AC adapter on every reverse-polarity cycle it output.

                              When these small transformers are over-loaded, the primary typically tends to open. The better-built ones will have a small fuse built into the transformer's primary (typically under a few layers of tape.) Change that and the adapter/transformer will work fine again.

                              Originally posted by socketa View Post
                              So i had to sort out another computer, which i was going to do anyway because either firefox (and it's few forks) or websites these days demand 2 Cores and more than 2GB of RAM (good grief, all that i want to do is read posts and watch internet videos at 240p or 360p) or they will cause the computer to lock up.
                              Windows XP 32-bit?

                              2 GB of RAM should be enough (though 1 GB isn't), but pretty much all versions after Firefox 32 have this "bug" where at startup and also at random intervals, they like to hog CPU resources for 10-30 seconds. Firefox 24 was the last good version that didn't do this and also wasn't too heavy on resources. Unfortunately, it's too old and won't render correctly on many websites. You really need version 49 or newer to make sure you don't have problems with websites.

                              That said, I'm still using a s478 Pentium 4 Prescott HT as the CPU in my main PC, where I do watch YT @ 360p once in a while (using FF 52 ESR Portable.) It's a pain, though (takes a few seconds for YT page to load fully, and then a few more to start the video buffering.) Adblock + custom hosts file helps quite a bit - at least with cutting out ads from loading and hogging more resources. But all in all, I'm getting very near the point of retiring that PC.

                              Originally posted by socketa View Post
                              So i've got something like 30 462/754 boards that are now practically non-internetable.
                              The socket 462 boards... yeah, unless they have Athlon XP 2500+ or better with at least 1.5 GB of RAM, they won't be much good on the internet, mostly due to websites being too bloated and not optimized to run on the (few) older instruction sets that these CPUs support (SSE at best on Athlon XP.)

                              The socket 754 boards should be a bit better if equipped with Athlon 64 2800+ CPUs or higher. They have SSE2 and clock-for-clock, Athlon 64 should be faster than Athlon XP. Don't expect miracles, of course.

                              Last time I used my OC'ed s939 Athlon 64 3200+ (@ 2.5 GHz) GPU test system on the web (around a year ago), it still managed to (barely) handle YT @ 720p with the help of an nVidia GeForce 8600 GT (for H.264 acceleration.) It ran OK elsewhere on the web, though. So at a minimum, you don't necessarily need 2 CPUs cores, but a little newer CPU will help. For AMD, I'd say socket 939 @ 2.4 GHz (1 core) or 2 GHz (2 cores) would be a lot better. For Intel, Pentium 4 Prescott HT on socket 775 would be minimum. Even though P4 on s775 is not much better than the 32-bit s478 P4s, the change from AGP to PCI-E GPU could allow you to render some websites faster.

                              Also, if you need a PC just for surfing more basic (and sane) websites, like Badcaps and Wikipedia, try the last Opera 12 release (version 12.18 on Presto engine, -not- Chromium garbage). I'm still using that on my bench "helper" laptop and a few other retro XP systems. In the case of my bench laptop, it has a Pentium 3 @ 700 MHz with 512 MB of RAM. I use it now only for viewing/downloading PDFs, surfing BCN occasionally, and once in a while looking info up in Google/Wikipedia.

                              Originally posted by socketa View Post
                              i was happy with dial-up, buffering, and waiting for the video to load.
                              Take me back 20years and push hold.
                              LOL
                              I don't know about dial-up. I get chills when I remember how bad ours was back in late '99 with its dial-up connection re-try noises before finally getting a connection... only to get lost after 5 minutes max.
                              But I'll be pretty happy if time turned back 10-13 years. Back then, websites still weren't anywhere near as bloated as they are now. And 15 years ago, I think the "bottom-less scrollbar" was still not a thing. I despise websites with bottomless scrolling. We have smartphone users to thank for that "feature".
                              Last edited by momaka; 07-19-2020, 09:32 PM.

                              Comment


                                #35
                                Re: Task TK-940TX-DF Blowing fuse

                                Looks like we may have got it!
                                Took out the PBU805 bridge rectifier and replaced it with a RS405L and the bulb didn't come on. It's a downgrade to 600 Peak Reverse Voltage, is that still perfectly acceptable?
                                Back at the start of this process i did remove the bridge rectifier and it tested good (and still does) according to this video:
                                https://www.youtube.com/watch?v=K-oNaQvMQco
                                and i also tested to both AC Legs, which over all equates to 4+4+2=10 tests in total, that all tested good, according to the diagram in the video.
                                So the rectifier tests good but it is bad! (i double checked that the bulb lit, by tacking the original rectifier onto the reverse side of the board before attaching and soldering the replacement in)
                                Looks like it shorts out only when voltage is applied - thought that there was something odd going on when there was no contact between the power plug pins, but the light came on.
                                I've put the MOV's, heatshrinked resistor, small blue cap, and 5VSB transformer back in, and the bulb remains unlit.
                                Should i keep replacing the parts ( in any specific order?) and doing the bulb test, after each component?
                                I've also lost that big pink resistor that you can see in post #5 - searched high and low for it - maybe it self-dissipated back into the quantum vacuum. What value would you say that it is?, as i can recall that one of the coloured bands looks a different colour in the photo than 'reality'.

                                I don't know about dial-up. I get chills when I remember how bad ours was back in late '99 with its dial-up connection re-try noises before finally getting a connection... only to get lost after 5 minutes max
                                That's either an earth contact or a resistance fault, often accompanied with audible line noise, on the copper phone line, or suspicious.
                                My dial-up was great at around 3kb/s, and only occasionally was disconnected when there was a problem with the copper lines or the exchange equipment (and i rang to find out), or my computer. Also, with a 56k modem, but only a few times, it'd reach speeds of close to 20kb/s when i accessed a certain FTP site to download files, or doing bit torrent. So i suspected that dial-up could go a lot faster than what they were giving us.
                                There is others that verify this as well:
                                https://forums.whirlpool.net.au/archive/408223

                                Pretty much all versions after Firefox 32 have this "bug" where at startup and also at random intervals, they like to hog CPU resources for 10-30 seconds. Firefox 24 was the last good version that didn't do this and also wasn't too heavy on resources
                                I'm using FF 48, and the latest versions of Palemoon, MyPal, and Basilisk. After a while the RAM (probably when i have bitchute or screwyoutube window open, but not playing) goes to just over 2GB and the browser wont let it go (even after closing all of the windows) until i restart the browser. And it still does it on a different computer.

                                When these small transformers are over-loaded, the primary typically tends to open. The better-built ones will have a small fuse built into the transformer's primary (typically under a few layers of tape.) Change that and the adapter/transformer will work fine again.
                                Thanks for that info.
                                i threw the adapter into the rubbish bin, which was collected about 2hrs after i read your post - yes, i should have at least opened it up to see why.
                                Should have let it sit for a while longer, until that idea popped into my head.
                                Last edited by socketa; 07-20-2020, 06:55 PM.

                                Comment


                                  #36
                                  Re: Task TK-940TX-DF Blowing fuse

                                  Correction:
                                  I'm using FF 48, and the latest versions of Palemoon, MyPal, and Basilisk. After a while the RAM (probably when i have bitchute or screwyoutube window open, but not playing) goes to just over 2GB and the browser wont let it go
                                  This happens occasionally, and it was the total RAM usage, not just Firefox. I'm now monitoring it more carefully, to see when it happens.
                                  Last edited by socketa; 07-21-2020, 01:36 AM.

                                  Comment


                                    #37
                                    Re: Task TK-940TX-DF Blowing fuse

                                    Originally posted by socketa View Post
                                    Looks like we may have got it!
                                    Took out the PBU805 bridge rectifier and replaced it with a RS405L and the bulb didn't come on.

                                    Yay, about time, lol.
                                    I guess I should have suspected the BR earlier to save on removing all of those components. But it's such a weird failure, as diodes usually fail short-circuit or occasionally open-circuit if enough current is forced into them. Have not seen one, at least on my bench, fail like yours, where it only breaks down under load. But I do know it is a possible failure mode... just extremely rare.

                                    Originally posted by socketa View Post
                                    It's a downgrade to 600 Peak Reverse Voltage, is that still perfectly acceptable?
                                    Yup, 600V is fine.
                                    Even 400V will work, though that would be cutting it very close (could certainly fail with a power surge, as the DC voltage at the caps is nominally ~340V.)

                                    Originally posted by socketa View Post
                                    I've put the MOV's, heatshrinked resistor, small blue cap, and 5VSB transformer back in, and the bulb remains unlit.
                                    Should i keep replacing the parts ( in any specific order?) and doing the bulb test, after each component?
                                    You can place back all of the 5VSB parts (including test 5VSB transistor) and then do the bulb test. If the 5VSB still has the replaced/good caps on its output, we should hopefully see the 5VSB come up now.

                                    Originally posted by socketa View Post
                                    I've also lost that big pink resistor that you can see in post #5 - searched high and low for it - maybe it self-dissipated back into the quantum vacuum. What value would you say that it is?, as i can recall that one of the coloured bands looks a different colour in the photo than 'reality'.
                                    R14? That one is for the main PS.
                                    Looks like 0.1 Ohms (Brown, Black, Silver, Gold). 5th band (green) I think signifies reliability in this case, or something similar. Not to be confused with regular 5-band color-coded resistors, though.

                                    Originally posted by socketa View Post
                                    at the start of this process i did remove the bridge rectifier and it tested good (and still does) according to this video:
                                    https://www.youtube.com/watch?v=K-oNaQvMQco
                                    and i also tested to both AC Legs, which over all equates to 4+4+2=10 tests in total, that all tested good, according to the diagram in the video.
                                    Good that you tested to both AC legs for steps 1 and 3. The video doesn't tell you to do this, but you always should anyways... so you definitely tested the bridge rectifier properly. It just threw a curve ball on us, breaking down only with voltage.

                                    Originally posted by socketa View Post
                                    That's either an earth contact or a resistance fault, often accompanied with audible line noise, on the copper phone line, or suspicious.
                                    Probably.
                                    The building/apartment we lived in back then was in Eastern Europe, built in the 70's during the Soviet Block days. Earth/grounding was not required by wiring code then, AFAIK, so most buildings have no earth. (The best was when someone grounded their water heater to the hot water/radiator pipes for heating, expecting them to provide a ground connection in case of fault. I remember many-a-times when going into other people's apartments and touching the radiators, I'd get a slight "tickly" feeling. On a few, it actually stung quite a bit. Many years later after I studied electronics I realized what this was.)

                                    Also, phones had just moved to tone dialing in the cities back then in the 90's, so POTS wasn't 100% reliable quite yet. Still better than the old Pulse-dialing system in the rest of the country(side), though - on that one, you couldn't even try to use dial up.

                                    Originally posted by socketa View Post
                                    Also, with a 56k modem, but only a few times, it'd reach speeds of close to 20kb/s when i accessed a certain FTP site to download files, or doing bit torrent. So i suspected that dial-up could go a lot faster than what they were giving us.
                                    Eh, that still sounds pretty horribly slow, lol.
                                    We moved to broadband/cable in 2001 and never really looked back.

                                    Price-wise, the difference between dial-up/ADSL and cable was too small to justify sticking to such slow connection. And in the case of Eastern Europe, where my cousins continued to live for a while... dial-up was always too expensive. In the mid-2000's, small "corner-store" ISPs started cropping up and throwing ugly Ethernet wiring between buildings, as the city didn't care to enforce any communication wiring regulations. The benefit of that was that for a fraction of the price of dial-up, people were getting a fast Ethernet connection - typically uncapped 100/100... or as fast as the tubes could get ya ... because it was literally a router or two on every apartment building, wired to a huge switch, and then an Ethernet connection ran to every customer. On the plus side, it was super-fast for torrenting. On the down side, during a thunderstorm, many people lost their computers due to those Ethernet cables running around and between buildings without protection. Add old buildings with no grounding and cheap Deer power supplies, and you can start to imagine how many PCs bit the dust.

                                    Originally posted by socketa View Post
                                    I'm using FF 48, and the latest versions of Palemoon, MyPal, and Basilisk. After a while the RAM (probably when i have bitchute or screwyoutube window open, but not playing) goes to just over 2GB and the browser wont let it go (even after closing all of the windows) until i restart the browser. And it still does it on a different computer.
                                    Ah yes, that's pretty much standard not just for the latest/last Firefox, but Chrome/Chromium and IE/Edge as well. On systems with less than 2 GB of RAM, you just have to close and restart the browser after a bit of use. Heck, even with 4 GB of RAM, I've ran out after several hour-long sessions on YouTube.

                                    Originally posted by socketa View Post
                                    i threw the adapter into the rubbish bin, which was collected about 2hrs after i read your post - yes, i should have at least opened it up to see why.
                                    Should have let it sit for a while longer, until that idea popped into my head.
                                    Ah, sorry I didn't post that earlier then. It's been pretty hot here for the last few weeks, so I've been avoiding the computer during the day or limiting use.

                                    Originally posted by socketa
                                    This happens occasionally, and it was the total RAM usage, not just Firefox. I'm now monitoring it more carefully, to see when it happens.
                                    It's Firefox, I'm pretty sure.
                                    I always keep Task Manager and regularly keep an eye on my resources - especially on my slower systems, like my main PC.
                                    Last edited by momaka; 07-21-2020, 12:21 PM.

                                    Comment


                                      #38
                                      Re: Task TK-940TX-DF Blowing fuse

                                      i put all of the parts back (and replaced the shorted zenner ZD1, that you can see just above the rectifier in post #27) except for that big pink resistor that i lost, and the other two large FETs - turned it on, and the light bulb instantly glows, and no 5VSB.
                                      I'm starting to think that the original rectifier might not be the cause of the problem

                                      When i test, in circuit, across that heat-shrinked resistor, that's next to the 5VSB transformer, i get a hard short - is that a clue?

                                      That "1MΩ discharge resistor" that i previously mentioned, doesn't exist, because when searching for it on the board, and not finding it, and after looking at a few PSU schematics, i eventually realized that it was the resistance of my body as i was holding the probes with my fingers onto the power supply wires.
                                      Doh!

                                      i found the adapter for the modem (i thought i threw it away, but discovered that i didn't)
                                      How do you open them nicely? As the advice that i've read is to cut it open, or start whacking it with a hammer after freezing it.

                                      Ah yes, that's pretty much standard not just for the latest/last Firefox, but Chrome/Chromium and IE/Edge as well. On systems with less than 2 GB of RAM, you just have to close and restart the browser after a bit of use. Heck, even with 4 GB of RAM, I've ran out after several hour-long sessions on YouTube.
                                      Thanks for the confirmation - i spent hours searching for a way to hard limit Firefox RAM usage, to no actual avail.
                                      Last edited by socketa; 07-22-2020, 04:37 PM.

                                      Comment


                                        #39
                                        Re: Task TK-940TX-DF Blowing fuse

                                        When these small transformers are over-loaded, the primary typically tends to open.
                                        Got it open by freezing it and tapping it out with a hammer and a blunt wide chisel.
                                        Moderately heating it might have worked better because some plastic broke off inside the groove. But anyway - yep, no fuse and open circuit on the primary; but it was too heavy anyway, as it kept falling out of the wall socket - maybe heavy because it needed more power for the Wi-fi.
                                        And i just learned why an effective short in the primary of a transformer doesn't blow the switchboard fuse - because of the winding's inductance.

                                        BTW, Firefox RAM usage was, a few minutes ago, up to nearly 1.5G, even after closing the three youtube/bitchute windows.
                                        Then things started to go slow, and then the masked browser screen appears a few times with the blue ring throbber.
                                        Then Firefox crashed.
                                        Last edited by socketa; 07-22-2020, 05:37 PM.

                                        Comment


                                          #40
                                          Re: Task TK-940TX-DF Blowing fuse

                                          Originally posted by socketa View Post
                                          i put all of the parts back (and replaced the shorted zenner ZD1, that you can see just above the rectifier in post #27) except for that big pink resistor that i lost, and the other two large FETs - turned it on, and the light bulb instantly glows, and no 5VSB.
                                          I'm starting to think that the original rectifier might not be the cause of the problem
                                          Well either way, back to removing parts one by one then, as detailed in the previous pages. Cumbersome? -Yes. But you should be able to find the part responsible when following all of those steps again.

                                          Originally posted by socketa View Post
                                          That "1MΩ discharge resistor" that i previously mentioned, doesn't exist, because when searching for it on the board, and not finding it, and after looking at a few PSU schematics, i eventually realized that it was the resistance of my body as i was holding the probes with my fingers onto the power supply wires.
                                          Doh!
                                          It happens.

                                          Another good reason not to hold the probes by the metal parts is if you encounter a charged HV capacitor - could get a nice sting for sure.

                                          Originally posted by socketa View Post
                                          How do you open them nicely? As the advice that i've read is to cut it open, or start whacking it with a hammer after freezing it.
                                          Depends on how well the adapter is glued. Usually I try the hammer and/or clamp method first (hammer or clamp in the corners to break the case seam.) But some adapters are too stubborn for that, so I just hacksaw or dremmel them open instead. They don't look nice afterwards with the latter option, though.

                                          Originally posted by socketa View Post
                                          Got it open by freezing it and tapping it out with a hammer and a blunt wide chisel.
                                          Moderately heating it might have worked better because some plastic broke off inside the groove. But anyway - yep, no fuse and open circuit on the primary;
                                          Did you try removing/unwinding/cutting the the isolating tape on the transformer? A lot of traffos hide the fuse in there. Usually it's a square box fuse, kind of looking like a small X2-class safety cap.

                                          Originally posted by socketa View Post
                                          And i just learned why an effective short in the primary of a transformer doesn't blow the switchboard fuse - because of the winding's inductance.
                                          Actually, shorting the secondary side pretty much "shorts" the primary side inductance too. So then the only thing left is the resistance of the windins... which is too low and allows too much current to pass through, thus burning the windings. However, the current is not high enough to blow the house breaker.

                                          Originally posted by socketa View Post
                                          BTW, Firefox RAM usage was, a few minutes ago, up to nearly 1.5G, even after closing the three youtube/bitchute windows.
                                          Then things started to go slow, and then the masked browser screen appears a few times with the blue ring throbber.
                                          Then Firefox crashed.
                                          Yeah, I just restart FF after I'm done watching YT or between different "sessions".

                                          Comment

                                          Working...
                                          X