Announcement

Collapse
No announcement yet.

VRM question

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

    VRM question

    I am testing a 3-phase VRM. Appears to have a short

    I req. some advice on replacement (if possible)

    The VRM has an Intersil ISL6556B controller
    This can drive up to 4 phases - this mobo has 3
    The MOSFET drivers are a;
    1 An Intersil HIP6601B and
    2 An Intersil HIP6602B

    The former package can drive 1 phase and the latter
    can drive the other 2 phases.

    I have 9 On Semiconductor NTD60N02R Power MOSFETs.
    The Upper Gate on the HIPs drives three for the three phases
    The Lower Gate on the HIPs drives six (from my inspection)

    So this is an asymmetric topology for obvious reasons such as;
    1. Reduces switching losses in the high-side FET and
    2. Conduction losses in the low-side FET
    at ratio of 2:1 (Low side to high side)

    I did a quick check (in cct) of the FETs for shorts and the upper
    side appear to be common to the gate and the sources are all
    common to the phase on the HIP6601B.
    As the FETs appear to be in good condition, I checked the drivers
    to see if phase and the upper gate on each device for each phase were s/c.

    I found on the HIP6601B, the upper gate drive was s/c to
    the phase pin - this is the third phase. The lower gate was o/c.
    On the HIP6602B, the upper gate 2 (phase 2) was s/c to the
    the phase 2 pin. Lower gate 2 was o/c.
    With the first phase on the HIP6602B, upper/lower were both o/c.

    Am I right in stating the phase pin for each phase should be
    common to the sources of the upper and lower FETs as is
    typical with a multi-phase VRM?

    As this was a quick analysis, and this board did get to POST
    first time I tried it, would the fact, two of the upper phases
    appear s/c impact this, or would, the apparent loss of two
    phases of the upper FETs cause a failure to POST?

    Of course, if I am correct, it may be either the driver or the
    FETs (or both) are at fault. Would the replacement of the FET
    drivers and the FETs be worthwhile?

    A little history of this mobo tells a well known story (at least
    in this forum). The system was powered by a Bestec and
    the mobo was an apparent casualty.

    The lytics appear ok. Have tested with an ESR meter.

    Any help and insight appreciated.

    #2
    Re: VRM question

    Do you hear any weird sounds, like squeaking or squealing coming from the VRM?? Or do any parts get hot?? If you have a VRM short, the VRM controller chip, (the Intersil 6556) should be signalling a fault via its Power OK pin, which could also be labelled as PWR_Good.. Typically it's an active high signal that indicates the VRM is outputting the correct voltage..


    As for checking the FET's, the FET drivers should be connected to the gates, not the sources..

    From what i remember of the 660X series of FET drivers, , the Top FET and Bottom FETs of each VRM phase each have their own connection to the FET driver.. THe 6602 is basically just two 6601's in the same package, with common power and ground connections..

    As for checking VRM FET's for shorts, make sure the board is fully powered off, then go from source to drain, and they should all measure open.. Also try measuring Vcore at the output side of one or each of the VRM phase inductors, with the power and a test CPU in the socket, to see if Vcore is actually present, or not.. The status pins of the 6556 are useful too..

    They should all read the same.. I assume o/c means open circuit, and s/c means short circuited.. With everything powered off, you should be able to

    Comment


      #3
      Re: VRM question

      Bestec, you say?
      The 250W model has a bad habit of frying southbridge chips. If that's the case, the board is basically just a bunch of parts soldered on a PCB with no functionality. Check if the southbridge chip/heatsink gets extremely hot or does not warm up at all.

      Comment


        #4
        Re: VRM question

        The original post is at
        https://www.badcaps.net/forum/showthread.php?t=4479

        I forced the PSU on as it would not start.
        So... to recap (scuse the pun);
        the mobo did post but progressively got degraded so far
        as the ticking sound was obviously a start into a short or was
        an xtal.
        Second time, after this the southbridge started to get hot.

        Hence, I suspect all (possibly) is not lost with this mobo.
        Alternatively, it may not be worth the bother repairing it!

        The VRM did not make any odd sounds.

        Your observations re - the FETs are correct, so I believe either
        the upper FETs have gone and this would obviously mean,
        very probably, the driver ICs also. As I measured these from the
        driver IC into the FETs and read s/c as with the phase pin s/c to the
        gate drive, it would imply the FETs are not open between gate
        and source.

        Thanks for your insight
        Last edited by JEWilson; 07-28-2008, 06:16 PM.

        Comment


          #5
          Re: VRM question

          Can anyone give guides on basic motherboard testing and fault finding

          Comment


            #6
            Re: VRM question

            Search the forums and all will be revealed

            There are a number of tutorials on basic fault finding

            These can be used as a start

            HTH

            Comment


              #7
              Re: VRM question

              @ coolday

              You can start from here:

              https://www.badcaps.net/forum/showthread.php?t=600

              and searching trough the forum as suggested by JEWilson

              Ciao
              Gianni
              "In the confrontation between the stream and the rock, the stream always wins...Not through strength, but through persistence."
              H. J. Brown

              Comment

              Working...