Announcement

Collapse
No announcement yet.

Crack & Stink Gigabyte GA7N400 Pro 2

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

    #41
    Re: Crack & Stink Gigabyte GA7N400 Pro 2

    Originally posted by socketa View Post
    So was thinking that this "short" could be on the motherboard, which, in-turn , could upset the PSU.
    Unlikely it happened that way around as most PC PSUs just shut down when their output is shorted, so the only way to kill a PSU with a shorted motherboard is if the PSU itself is a totally trash design that has no short circuit protection - which would be a PSU not worth using in the first place, and its death would be of little importance...

    (Note, I am ONLY talking about PC\ATX PSUs here, MANY other PSUs (including SMPS and linear) do NOT have short-circuit protection and shutdown)

    The fact you already have bad capacitors shows it's most likely the PSU that went bad first. momaka already explained the most logical scenario in post #34.

    The bad capacitors in your PSU caused instability and\or other improper operation and likely resulted in the resistor burning out in the PSU.

    Failure of the PSU likely led to the incorrect voltage going to the motherboard on one or more rails, which killed it.
    "Tantalum for the brave, Solid Aluminium for the wise, Wet Electrolytic for the adventurous"
    -David VanHorn

    Comment


      #42
      Re: Crack & Stink Gigabyte GA7N400 Pro 2

      Yes,
      just found out that another psu has short circuit protection, when i accidentally shorted
      it - thought that it would have blown a fuse, but it started back up again , no problem.

      so it seems that there is some component/circuit that automatically resets itself after a short circuit condition

      Failure of the PSU likely led to the incorrect voltage
      referring to post #34:
      what is meant by "secondary-side auxiliary supply"?

      does that just mean that a greater voltage is now on the secondary-side of the 5vsb transformer because the pwm is not controlling the voltages correctly as a result of a bad output cap?
      i.e.,
      bad output caps >> incorrect operation of pwm controller chip >> excessive voltage across the 5vsb secondary transformer winding

      Comment


        #43
        Re: Crack & Stink Gigabyte GA7N400 Pro 2

        Originally posted by socketa View Post
        bad output caps >> incorrect operation of pwm controller chip >> excessive voltage across the 5vsb secondary transformer winding
        Almost .
        more like, bad output 5VSB caps >> excessive voltage across the 5vsb secondary-side auxiliary winding >> burnt resistor >> erratic operation of PWM controller

        Better yet, let's have us some PC PSU "101 basics"...

        Some common terms:
        I/P - short for Input; likewise O/P is same as output.
        When referring to a voltage that the PSU produces, the following terms are sometimes used interchangeably: rail, supply, line (or supply line, and output (IIRC, there was some confusion about that either in this thread or in the other, so that's why I added this).

        To keep things simple, I will not get into the circuits of the power supply too much here, just the general concepts.

        In general, there are actually two power supplies in a PC PSU: the stand-by supply and the main power supply.
        The stand-by supply is always turned ON whenever the PSU is plugged in the wall. It is meant to provide only a small ammount of power to the motherboard and the PSU's chips (i.e. PWM controller and other ICs). The stand-by supply generates the 5VSB line and the so-called (by me, at least) "secondary-side auxiliary supply/rail". 5VSB is usually only used by the motherboard. On the other hand, the secondary-side auxiliary rail is not available to the motherboard at all and is entirely dedicated to providing power to the PSU's ICs. So, in other words, when the computer is OFF but plugged in the wall, the motherboard is provided with 5V via the 5VSB line and the PSU's IC's are provided a voltage from the secondary-side auxiliary rail; that way, they are ready and awaiting a "turn-ON" signal (usually known as PS-ON). When the signal is given by the motherboard (that is, when you press the power switch on your PC), the PSU's ICs "wake up" and turn ON and control the main power suppy.
        The main power supply is meant for providing power to the PC whenever it is ON (and sometimes in S1 stand-by). It generates 3.3V, 5V, 12V, -12V, and sometimes -5V (usually older PSU designs - though I should note that a lack of -5V does not mean the design of the PSU is newer).
        Last edited by momaka; 07-17-2014, 10:23 PM.

        Comment


          #44
          Re: Crack & Stink Gigabyte GA7N400 Pro 2

          For the record, ATX and similar supplies have standby power. Older PSUs like AT and XT don't.
          "Tantalum for the brave, Solid Aluminium for the wise, Wet Electrolytic for the adventurous"
          -David VanHorn

          Comment


            #45
            Re: Crack & Stink Gigabyte GA7N400 Pro 2

            ^ Excellent point.

            Comment


              #46
              Re: Crack & Stink Gigabyte GA7N400 Pro 2

              Thanks momaka,
              Just to make sure i've got this straight...

              Looking at the example schematic (in attachment below),
              the 'circuit' (orange) that goes to the ps-on terminal, is teed-off from the 5vsb output.
              Also there is an independent 'circuit' (purple) that originates from the transformer, that feeds one of the pwm controller (TL494) terminals.

              Am picking that the "secondary-side auxiliary rail" is the purple path - since the ps-on output is already live, along with the 5vsb output.

              i can see that the orange circuit also powers a transistor, so am guessing that when the PC power button is pressed, that that transistor must liven up the TL494 controller.

              Does all of that sound correct?

              i also checked this iso psu pcb with a continuity meter, and see that the 5vsb output comes from a inductor and then a diode, and then from the transformer.

              there is continuity from 5vsb output to one side of the diode, but no continuity across the diode.
              forward or reverse bias, multimeter displays 50ohms

              Should this diode have continuity in one direction, or is it faulty?
              (photo of this diode in attachment below)
              Attached Files
              Last edited by socketa; 07-18-2014, 06:05 PM.

              Comment


                #47
                Re: Crack & Stink Gigabyte GA7N400 Pro 2

                Originally posted by socketa View Post
                Just to make sure i've got this straight...

                Looking at the example schematic (in attachment below),
                the 'circuit' (orange) that goes to the ps-on terminal, is teed-off from the 5vsb output.
                Also there is an independent 'circuit' (purple) that originates from the transformer, that feeds one of the pwm controller (TL494) terminals.

                Am picking that the "secondary-side auxiliary rail" is the purple path - since the ps-on output is already live, along with the 5vsb output.

                i can see that the orange circuit also powers a transistor, so am guessing that when the PC power button is pressed, that that transistor must liven up the TL494 controller.

                Does all of that sound correct?
                Yes, all of that is correct.

                Originally posted by socketa View Post
                there is continuity from 5vsb output to one side of the diode, but no continuity across the diode.
                forward or reverse bias, multimeter displays 50ohms

                Should this diode have continuity in one direction, or is it faulty?
                (photo of this diode in attachment below)
                If your multimeter has a diode test function (usually, but not always, combined with continuity on most multimeters), then the diode should show a reading one way but not the other. HOWEVER, most 2-transistor 5VSB circuits tend to have a load resistor to keep them stable when there is not much load on it from the motherboard. 47, 51, and 100 Ohms are very common resistor values here. If your PSU has one, that can easily mess up your readings.

                Looking at the last picture you uploaded, your load resistor seems to be 100 Ohms. Remove it and double check that diode again (that would be D23). If you still get 50 Ohms, then the diode could as well be bad.
                ... Or, remove the diode out of the circuit and measure it. It's almost the same work as removing the resistor (though, the diode has thicker leads, so naturally, it is easier to remove the resistor).
                Last edited by momaka; 07-23-2014, 09:34 PM.

                Comment


                  #48
                  Re: Crack & Stink Gigabyte GA7N400 Pro 2

                  Good day to you.

                  i had trouble seeing those resistor colors with a magnifying glass and a torch!
                  (wouldn't it be easier for all if they just wrote some numbers on it?)
                  It looks like black brown black gold
                  But since it tests as 50 ohms, i guess the first black must be "green"

                  Yes, the multimeter has diode test mode
                  What does that do, that the continuity tester doesn't?
                  What does the voltage reading indicate?

                  i cut the resistor
                  now readings across diode are:
                  forward bias = 0.154v
                  reverse bias = 0.588v

                  Also the resistor that is between the emitter of C945, and the emitter of C5353, had a hard short
                  (diagram attached)

                  As can be seen, for educational purposes, i spent the day tracing out some of the primary side circuit, after being inspired by everell's thread on the 311 mod:
                  https://www.badcaps.net/forum/showthread.php?t=16071

                  There could possibly be a mistake in it, even though i went over it twice.
                  Does it look right?
                  Both of the windings are on the same transformer.
                  i think the drawing could be tidied up by removing the 2 long bottom lines, and replacing them with symbols for common, but i'm not sure if they are actually "common"

                  i have another iso psu, similar to this one, and also a dm311 chip out of a dead psu that i found.
                  Finding this quite intriguing, and am keen to give the mod a shot, sooner or later.
                  Attached Files
                  Last edited by socketa; 07-25-2014, 02:43 AM.

                  Comment


                    #49
                    Re: Crack & Stink Gigabyte GA7N400 Pro 2

                    Yes, the multimeter has diode test mode
                    What does that do, that the continuity tester doesn't?
                    What does the voltage reading indicate?
                    Continuity tester measures continuity - at a resistance below a certain threshold (some have 20 ohm, some less) the meter will beep and later (if you keep probes connected) it may show the actual resistance.
                    If the resistance is above a threshold, it won't say there's continuity (try putting probes on the leads of a 100k resistor and see if it says continuity.

                    Diode test sends current from one lead to the other and measures the voltage drop on the diode. A good diode will show a voltage drop with the leads in one position and no voltage drop if you reverse the leads (the diode blocks current flow).

                    Exceptions are zener diodes and I think some more obscure diode types (but I can't say one off the top of my head). Zener diodes conduct when a particular voltage is reached, so the way to test them is to put a resistor in series to limit the current to a few mA and then use a power supply with current limit to slowly raise the voltage - when the voltage zener works at is reached, it will conduct and you'll have a short and you'll have those mA flowing through the diode and the power supply will enter current limit mode.

                    Resistance and continuity and diode test use different voltages sometimes. My ut61e uses -0.35v for resistance, -2.5v for continuity, 2.8v for diodes. The current can also vary - you may be able to light up leds using diode test, but the resistance modes may be limited to a few mA, a value below the amount a led would turn on.

                    Comment


                      #50
                      Re: Crack & Stink Gigabyte GA7N400 Pro 2

                      Originally posted by socketa View Post
                      i had trouble seeing those resistor colors with a magnifying glass and a torch!
                      (wouldn't it be easier for all if they just wrote some numbers on it?)
                      It looks like black brown black gold
                      But since it tests as 50 ohms, i guess the first black must be "green"
                      Good point.
                      Probably a 51 Ohm resistor then as that's a standard value (i.e. Green Brown Black Gold).

                      Originally posted by socketa View Post
                      i cut the resistor
                      now readings across diode are:
                      forward bias = 0.154v
                      reverse bias = 0.588v
                      Not quite the results I expected, but I will check my ISO PSUs and let you know. Did you make sure to leave the multimeter for a few seconds before nothing down the measurements. The capacitors on the output of the 5VSB can take some time to charge so that can mess up the results.
                      I'd expect to see about 0.5 to 0.7V in forward bias and higher (if any at all) reverse bias results. But again, I can check.

                      Originally posted by socketa View Post
                      Also the resistor that is between the emitter of C945, and the emitter of C5353, had a hard short
                      (diagram attached)
                      Could be just a low-value resistor. Did you check its color bands? Resistors almost always fail open circuit. I haven't actually seen one fail as a short circuit. With diodes, it's the other way around.

                      Originally posted by socketa View Post
                      As can be seen, for educational purposes, i spent the day tracing out some of the primary side circuit, after being inspired by everell's thread on the 311 mod:
                      https://www.badcaps.net/forum/showthread.php?t=16071

                      There could possibly be a mistake in it, even though i went over it twice.
                      Does it look right?
                      Hard to tell without the resistor values, but many parts of it look about right. I'll trace mine when I can and try to post it here.

                      Originally posted by socketa View Post
                      Both of the windings are on the same transformer.
                      i think the drawing could be tidied up by removing the 2 long bottom lines, and replacing them with symbols for common, but i'm not sure if they are actually "common"
                      They could be "common", but it would be the primary side common (which could be 160V lower than ground). Secondary side common, on the other hand, is ground (or should be unless you are running ungrounded outlets).

                      Originally posted by socketa View Post
                      Finding this quite intriguing, and am keen to give the mod a shot, sooner or later.
                      Yeah, everell's 311 mod is neat! I bought two 311 chips but haven't used them yet. A lot safer and a little bit more efficient than 2-transistor design. Should be easy for the most part, though.

                      Comment

                      Working...
                      X