Loading ...
Sorry, an error occurred while loading the content.

Re: Ensemble RX II (HF) U10 voltages; removing SMD chip

Expand Messages
  • bconwell
    Mystery solved: operator error. (Instead of leaving the U12 pads empty and installing the chip at U6, I d mistakenly left the U6 pads empty, and installed the
    Message 1 of 5 , Dec 31, 2010
    • 0 Attachment
      Mystery solved: operator error.

      (Instead of leaving the U12 pads empty and installing the chip at U6, I'd mistakenly left the U6 pads empty, and installed the chip at U12.)

      So... I learned a bit about desoldering SMD ICs.

      I put on a broader soldering iron tip, and wicked most of the solder from each side of the chip using 1.5mm solder wick (wetted, a bit, with flux). The IC was still tight on the board, but most of the solder was gone.

      Researching SMD chip removal on the web, I found a suggestion to run a length of small piano wire between the chip body and the legs, secure one end to the PCB with hemostats, and then pull sideways on the wire while heating pads.

      I had no piano wire. But I had some left-over #30 enameled wire from the toroid winding. I'd earlier been dismayed that the enamel didn't come off with heat. But now that liability turned to an advantage.

      I snaked a length of the number 30 wire between the chip body and under the pins (I wasn't able to do this until I'd wicked-away most of the solder), and secured one end. Then I briefly heated each pad while pulling the wire out between the pin and pad. When the solder melted, the lateral wire force lifted the leg from the pad, allowing the wire to slip out; I then moved quickly to the next leg. Neat. When I repeated the process on the other side of the chip, the chip fell free. Fortunately, the pins were all still essentially co-planar - seemingly none the worse for wear.

      Installed it at U6, and QSD unit now checks out FB.

      Whew!

      Tomorrow the final, op amp, stage...

      73,

      /Bill



      --- In softrock40@yahoogroups.com, "bconwell" <wycpublic@...> wrote:
      >
      > I took a break for dinner, and recognized that the QSD clock signals that appear on pins 2 and 14 of U10 come from U6, which earlier checked out fb.
      >
      > So now re-measuring pin voltages on U6, I find a wild variance from my earlier-recorded voltages. (glad I wrote them down...)
      >
      > On the HF jumper wire, I find a squarish-wave signal, about +/- 1 volt around a nominal 5v. This is the signal that appears at the junction of R10/R11, which the build notes indicate should be 2.5v on a DVM. The DVM measures 4.93v instead.
      >
      > So my sleuthing process continues. (Glad, at least, to think that it may not be a problem with the U10 chip itself...)
      >
      > (I can hear the 28.2 MHz signal in my transceiver S9+, so at least the oscillator, etc., are doing their job...)
      >
      > /Bill
      >
      > --- In softrock40@yahoogroups.com, "bconwell" <wycpublic@> wrote:
      > >
      > > All's gone well up to this point. But my measurements on U10 aren't as stated in the build notes.
      > >
      > > In particular, I get nil voltage at pins 2, 4, 5, 9,10,13,14. On pins 3,6,11,12 I get 2.50 v (instead of the suggest 40%, or 2.1 v). The other pins measure OK.
      > >
      > > I do get 2.5v from the voltage divider circuit (which, thru T3, drives pins 11/12. Interestingly, the voltage drop across the 10 ohms resistors R25/R26 is 0v (as far as my cheap DVM can tell - it nominally reports down to the tenth of a millivolt).
      > >
      > > I've measured both at the U10 terminals, and at the PCB pads - no apparent soldering trouble.
      > >
      > > I've got a scope, but don't see any clock signals on pins 2/14 (or on any other U10 pins). What frequency should I expect to see there?
      > >
      > > I've got CFGSR.EXE v 9.0 running, and the libusb-win32 driver. (No other software is running - is anything else necessary?)
      > >
      > > (I'm hoping I haven't somehow zapped the chip. I've just sort of mastered soldering these little things into place. Getting one out will require untested skills.)
      > >
      > > 'Any guidance will be welcomed.
      > >
      > > 73 & HNY,
      > >
      > > /Bill, K2PO/7
      > > Portland, OR
      > >
      >
    • Philip
      Great de soldering tip Bill... I did the same thing (put the chip on the wrong pads) Must admit though, my de soldering destroyed the chip...
      Message 2 of 5 , Jan 1, 2011
      • 0 Attachment
        Great de soldering tip Bill...

        I did the same thing (put the chip on the wrong pads)
        Must admit though, my de soldering destroyed the chip...



        --- In softrock40@yahoogroups.com, "bconwell" <wycpublic@...> wrote:
        >
        > Mystery solved: operator error.
        >
        > (Instead of leaving the U12 pads empty and installing the chip at U6, I'd mistakenly left the U6 pads empty, and installed the chip at U12.)
        >
        > So... I learned a bit about desoldering SMD ICs.
        >
        > I put on a broader soldering iron tip, and wicked most of the solder from each side of the chip using 1.5mm solder wick (wetted, a bit, with flux). The IC was still tight on the board, but most of the solder was gone.
        >
        > Researching SMD chip removal on the web, I found a suggestion to run a length of small piano wire between the chip body and the legs, secure one end to the PCB with hemostats, and then pull sideways on the wire while heating pads.
        >
        > I had no piano wire. But I had some left-over #30 enameled wire from the toroid winding. I'd earlier been dismayed that the enamel didn't come off with heat. But now that liability turned to an advantage.
        >
        > I snaked a length of the number 30 wire between the chip body and under the pins (I wasn't able to do this until I'd wicked-away most of the solder), and secured one end. Then I briefly heated each pad while pulling the wire out between the pin and pad. When the solder melted, the lateral wire force lifted the leg from the pad, allowing the wire to slip out; I then moved quickly to the next leg. Neat. When I repeated the process on the other side of the chip, the chip fell free. Fortunately, the pins were all still essentially co-planar - seemingly none the worse for wear.
        >
        > Installed it at U6, and QSD unit now checks out FB.
        >
        > Whew!
        >
        > Tomorrow the final, op amp, stage...
        >
        > 73,
        >
        > /Bill
        >
        >
        >
        > --- In softrock40@yahoogroups.com, "bconwell" <wycpublic@> wrote:
        > >
        > > I took a break for dinner, and recognized that the QSD clock signals that appear on pins 2 and 14 of U10 come from U6, which earlier checked out fb.
        > >
        > > So now re-measuring pin voltages on U6, I find a wild variance from my earlier-recorded voltages. (glad I wrote them down...)
        > >
        > > On the HF jumper wire, I find a squarish-wave signal, about +/- 1 volt around a nominal 5v. This is the signal that appears at the junction of R10/R11, which the build notes indicate should be 2.5v on a DVM. The DVM measures 4.93v instead.
        > >
        > > So my sleuthing process continues. (Glad, at least, to think that it may not be a problem with the U10 chip itself...)
        > >
        > > (I can hear the 28.2 MHz signal in my transceiver S9+, so at least the oscillator, etc., are doing their job...)
        > >
        > > /Bill
        > >
        > > --- In softrock40@yahoogroups.com, "bconwell" <wycpublic@> wrote:
        > > >
        > > > All's gone well up to this point. But my measurements on U10 aren't as stated in the build notes.
        > > >
        > > > In particular, I get nil voltage at pins 2, 4, 5, 9,10,13,14. On pins 3,6,11,12 I get 2.50 v (instead of the suggest 40%, or 2.1 v). The other pins measure OK.
        > > >
        > > > I do get 2.5v from the voltage divider circuit (which, thru T3, drives pins 11/12. Interestingly, the voltage drop across the 10 ohms resistors R25/R26 is 0v (as far as my cheap DVM can tell - it nominally reports down to the tenth of a millivolt).
        > > >
        > > > I've measured both at the U10 terminals, and at the PCB pads - no apparent soldering trouble.
        > > >
        > > > I've got a scope, but don't see any clock signals on pins 2/14 (or on any other U10 pins). What frequency should I expect to see there?
        > > >
        > > > I've got CFGSR.EXE v 9.0 running, and the libusb-win32 driver. (No other software is running - is anything else necessary?)
        > > >
        > > > (I'm hoping I haven't somehow zapped the chip. I've just sort of mastered soldering these little things into place. Getting one out will require untested skills.)
        > > >
        > > > 'Any guidance will be welcomed.
        > > >
        > > > 73 & HNY,
        > > >
        > > > /Bill, K2PO/7
        > > > Portland, OR
        > > >
        > >
        >
      • R. R. (Robby) Robson
        Bill: Others on this reflector are more adept at trouble shooting than I. However, for what it s worth, here is my take on your problem. Because you said you
        Message 3 of 5 , Jan 2, 2011
        • 0 Attachment
          Bill:
           
          Others on this reflector are more adept at trouble shooting than I.  However, for what it's worth, here is my take on your problem.
           
          Because you said you had earlier checked out the voltages on U6 and the R1-/R11 voltage divider and they were to spec, it is not likely you have made a not uncommon error of mounting the two resistors (R10/R11) in the wrong orientation (Correct orientation is R10=S-N, R11 = N-S; incorrect orientation is R10=E-W and R11=W-E)
           
          Looks like you are getting the 5V rail voltage at the junction of R10/R11.  Those 10k resistors should divide the rail in half.  Try lifting the hairpin of R10, taking the voltage divider out of the picture.  If you still get 5V at the junction (R11 hairpin), then look for one or more of the following ways for the 5V rail to get into that part of the circuit:
          (1)  a short between U12-4 and U12-3
          (2)  a short between U6-10 and U6-11
          (3)  a short between the top of C48 and the barrel lead of R10
          If this doesn't help, see what happens if you pull the HF Jumper wire.  Does the hairpin of R11 still have 5V
           
          Finally, you could have an issue with R11 or R10.  With R11's hairpin lead out ofd the circuit (and power off), test the resistance of R10 - should be 10k.  Then test the resistance of R11
           
          HTH
           
          On Sat, Jan 1, 2011 at 12:15 AM, bconwell <wycpublic@...> wrote:
           

          I took a break for dinner, and recognized that the QSD clock signals that appear on pins 2 and 14 of U10 come from U6, which earlier checked out fb.

          So now re-measuring pin voltages on U6, I find a wild variance from my earlier-recorded voltages. (glad I wrote them down...)

          On the HF jumper wire, I find a squarish-wave signal, about +/- 1 volt around a nominal 5v. This is the signal that appears at the junction of R10/R11, which the build notes indicate should be 2.5v on a DVM. The DVM measures 4.93v instead.

          So my sleuthing process continues. (Glad, at least, to think that it may not be a problem with the U10 chip itself...)

          (I can hear the 28.2 MHz signal in my transceiver S9+, so at least the oscillator, etc., are doing their job...)

          /Bill



          --- In softrock40@yahoogroups.com, "bconwell" <wycpublic@...> wrote:
          >
          > All's gone well up to this point. But my measurements on U10 aren't as stated in the build notes.
          >
          > In particular, I get nil voltage at pins 2, 4, 5, 9,10,13,14. On pins 3,6,11,12 I get 2.50 v (instead of the suggest 40%, or 2.1 v). The other pins measure OK.
          >
          > I do get 2.5v from the voltage divider circuit (which, thru T3, drives pins 11/12. Interestingly, the voltage drop across the 10 ohms resistors R25/R26 is 0v (as far as my cheap DVM can tell - it nominally reports down to the tenth of a millivolt).
          >
          > I've measured both at the U10 terminals, and at the PCB pads - no apparent soldering trouble.
          >
          > I've got a scope, but don't see any clock signals on pins 2/14 (or on any other U10 pins). What frequency should I expect to see there?
          >
          > I've got CFGSR.EXE v 9.0 running, and the libusb-win32 driver. (No other software is running - is anything else necessary?)
          >
          > (I'm hoping I haven't somehow zapped the chip. I've just sort of mastered soldering these little things into place. Getting one out will require untested skills.)
          >
          > 'Any guidance will be welcomed.
          >
          > 73 & HNY,
          >
          > /Bill, K2PO/7
          > Portland, OR
          >




          --
          Cheers,
          Robby
           
          Richard R. (Robby) Robson
          Amateur Radio Station WB5RVZ

        Your message has been successfully submitted and would be delivered to recipients shortly.