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

8800b front panel glitch

Expand Messages
  • mfeberhard
    I recently debugged an 8800b front panel, and I thought I d share what I learned with those of you who might also be working on 8800b s. The conclusion first:
    Message 1 of 6 , Aug 23, 2013
    • 0 Attachment
      I recently debugged an 8800b front panel, and I thought I'd share what I learned with those of you who might also be working on 8800b's.

      The conclusion first: I strongly recommend replacing IC 'R' on the D/C Interface board with a 7414. The manual calls for a 74LS04 here, which is not reliable.

      This 8800b of mine was flaky - it would occasionally be able to boot Basic from disk. But normally it would fail with an 'M' error, indicating that the monitor did not read back the same data that it wrote to memory.

      The most repeatable symptom that I saw was using TURMON. Using the 'M' command, it would write successfully to memory for addresses 000000 through 000003, and would normally fail (and crash) when it tried to write to address 000004. Sometimes it would fail at other addresses. Occasionally, it would work fine at address 000004. When I looked in memory (with the front panel), the 'failed' write had actually written the correct data!

      The worst part was that nothing was repeatable - the bad behavior seemed to morph as I chased the problem.

      Perhaps you have seen the same misbehavior on your 8800b...

      I have another (working) 8800b, so I was able to narrow the problem down to the D/C Interface board by swapping boards.

      It seemed like the ribbon cables to the front panel were not constantly making good contact, so I replaced them - no easy task! Then I one-by-one replaced nearly every single chip on the board, with no improvement. And none of these chips was socketed :-(

      Finally I noticed that IC 'R' on my good D/C Interface board was not a 74LS04 as the assembly instructions indicate. Instead, it was a plain 7404. So I substituted a 7404 onto the troublesome board, and the board suddenly works perfectly.

      This was a bit puzzling to me, because none of the inverters in IC 'R' are involved in *any* switching during normal operation. 2 of the inverters are in the PROT (memory protect) circuitry, which doesn't do anything. One is in the PRESET path. One drives Single Step onto the bus, which goes to nothing. One inverts the -FRDY signal, which is used to stop/start the processor when the front panel is in the STOP mode, and is always high when in RUN mode.

      So I put the 74LS04 back, and set my 'scope to trigger at 1.25V, and looked at the inputs to IC 'R' when running. And sure enough, the -FRDY signal was occasionally experiencing a >1.25V glitch, induced by the adjacent signal on the ribbon cable to the front panel.

      The adjacent ribbon cable signal is -DO5 - the only data-out signal sent to the front panel. With the right combination of data bytes written by the processor, the processor was getting a random -RDYIN glitch.

      Just for fun, I swapped out the 7404 on the good board for a 74LS04. Sure enough - it failed too. This made me conclude that the problem is more universal than just my bad boy 8800b - you might have seen the same problem.

      The 7404 works because it draws 4 times the current on its input as does the 74LS04 - and so it "eats" the glitch, reducing the glitch amplitude to around 0.75V.

      For extra margin, I replaced IC 'R' with a 7414, which has the higher input current, but is also a Schmitt trigger, so that it could tolerate a higher glitch voltage.

      So as I say, I recommend replacing IC 'R' on the D/C Interface board with a 7414.

      -Martin E
    • deramp5113
      Was this the same, original 74LS04 moved between the two computers, or were you able to make them both fail with a new 74LS04 as well? Mike
      Message 2 of 6 , Aug 23, 2013
      • 0 Attachment
        Was this the same, original 74LS04 moved between the two computers, or were you able to make them both fail with a "new" 74LS04 as well?

        Mike

        --- In altaircomputerclub@yahoogroups.com, "mfeberhard" <eberhard@...> wrote:
        >
        > I recently debugged an 8800b front panel, and I thought I'd share what I learned with those of you who might also be working on 8800b's.
        >
        > The conclusion first: I strongly recommend replacing IC 'R' on the D/C Interface board with a 7414. The manual calls for a 74LS04 here, which is not reliable.
        >
        > This 8800b of mine was flaky - it would occasionally be able to boot Basic from disk. But normally it would fail with an 'M' error, indicating that the monitor did not read back the same data that it wrote to memory.
        >
        > The most repeatable symptom that I saw was using TURMON. Using the 'M' command, it would write successfully to memory for addresses 000000 through 000003, and would normally fail (and crash) when it tried to write to address 000004. Sometimes it would fail at other addresses. Occasionally, it would work fine at address 000004. When I looked in memory (with the front panel), the 'failed' write had actually written the correct data!
        >
        > The worst part was that nothing was repeatable - the bad behavior seemed to morph as I chased the problem.
        >
        > Perhaps you have seen the same misbehavior on your 8800b...
        >
        > I have another (working) 8800b, so I was able to narrow the problem down to the D/C Interface board by swapping boards.
        >
        > It seemed like the ribbon cables to the front panel were not constantly making good contact, so I replaced them - no easy task! Then I one-by-one replaced nearly every single chip on the board, with no improvement. And none of these chips was socketed :-(
        >
        > Finally I noticed that IC 'R' on my good D/C Interface board was not a 74LS04 as the assembly instructions indicate. Instead, it was a plain 7404. So I substituted a 7404 onto the troublesome board, and the board suddenly works perfectly.
        >
        > This was a bit puzzling to me, because none of the inverters in IC 'R' are involved in *any* switching during normal operation. 2 of the inverters are in the PROT (memory protect) circuitry, which doesn't do anything. One is in the PRESET path. One drives Single Step onto the bus, which goes to nothing. One inverts the -FRDY signal, which is used to stop/start the processor when the front panel is in the STOP mode, and is always high when in RUN mode.
        >
        > So I put the 74LS04 back, and set my 'scope to trigger at 1.25V, and looked at the inputs to IC 'R' when running. And sure enough, the -FRDY signal was occasionally experiencing a >1.25V glitch, induced by the adjacent signal on the ribbon cable to the front panel.
        >
        > The adjacent ribbon cable signal is -DO5 - the only data-out signal sent to the front panel. With the right combination of data bytes written by the processor, the processor was getting a random -RDYIN glitch.
        >
        > Just for fun, I swapped out the 7404 on the good board for a 74LS04. Sure enough - it failed too. This made me conclude that the problem is more universal than just my bad boy 8800b - you might have seen the same problem.
        >
        > The 7404 works because it draws 4 times the current on its input as does the 74LS04 - and so it "eats" the glitch, reducing the glitch amplitude to around 0.75V.
        >
        > For extra margin, I replaced IC 'R' with a 7414, which has the higher input current, but is also a Schmitt trigger, so that it could tolerate a higher glitch voltage.
        >
        > So as I say, I recommend replacing IC 'R' on the D/C Interface board with a 7414.
        >
        > -Martin E
        >
      • mfeberhard
        I tried several 74LS04 s in this position, from several manufacturers. All failed. I tried several 7404 s, all worked fine. Same with 7414 s. The particular
        Message 3 of 6 , Aug 23, 2013
        • 0 Attachment
          I tried several 74LS04's in this position, from several manufacturers. All failed. I tried several 7404's, all worked fine. Same with 7414's.

          The particular chip in position 'Y' made a difference on how bad the failure was with the 74LS04. IC 'Y' is the driver for the adjacent signal on the ribbon cable. I would get a variety of flaky behaviors depending on which particular chip I put in IC 'Y', so long as a 74LS04 was in 'R'.



          --- In altaircomputerclub@yahoogroups.com, "deramp5113" <deramp5113@...> wrote:
          >
          > Was this the same, original 74LS04 moved between the two computers, or were you able to make them both fail with a "new" 74LS04 as well?
          >
          > Mike
          >
          > --- In altaircomputerclub@yahoogroups.com, "mfeberhard" <eberhard@> wrote:
          > >
          > > I recently debugged an 8800b front panel, and I thought I'd share what I learned with those of you who might also be working on 8800b's.
          > >
          > > The conclusion first: I strongly recommend replacing IC 'R' on the D/C Interface board with a 7414. The manual calls for a 74LS04 here, which is not reliable.
          > >
          > > This 8800b of mine was flaky - it would occasionally be able to boot Basic from disk. But normally it would fail with an 'M' error, indicating that the monitor did not read back the same data that it wrote to memory.
          > >
          > > The most repeatable symptom that I saw was using TURMON. Using the 'M' command, it would write successfully to memory for addresses 000000 through 000003, and would normally fail (and crash) when it tried to write to address 000004. Sometimes it would fail at other addresses. Occasionally, it would work fine at address 000004. When I looked in memory (with the front panel), the 'failed' write had actually written the correct data!
          > >
          > > The worst part was that nothing was repeatable - the bad behavior seemed to morph as I chased the problem.
          > >
          > > Perhaps you have seen the same misbehavior on your 8800b...
          > >
          > > I have another (working) 8800b, so I was able to narrow the problem down to the D/C Interface board by swapping boards.
          > >
          > > It seemed like the ribbon cables to the front panel were not constantly making good contact, so I replaced them - no easy task! Then I one-by-one replaced nearly every single chip on the board, with no improvement. And none of these chips was socketed :-(
          > >
          > > Finally I noticed that IC 'R' on my good D/C Interface board was not a 74LS04 as the assembly instructions indicate. Instead, it was a plain 7404. So I substituted a 7404 onto the troublesome board, and the board suddenly works perfectly.
          > >
          > > This was a bit puzzling to me, because none of the inverters in IC 'R' are involved in *any* switching during normal operation. 2 of the inverters are in the PROT (memory protect) circuitry, which doesn't do anything. One is in the PRESET path. One drives Single Step onto the bus, which goes to nothing. One inverts the -FRDY signal, which is used to stop/start the processor when the front panel is in the STOP mode, and is always high when in RUN mode.
          > >
          > > So I put the 74LS04 back, and set my 'scope to trigger at 1.25V, and looked at the inputs to IC 'R' when running. And sure enough, the -FRDY signal was occasionally experiencing a >1.25V glitch, induced by the adjacent signal on the ribbon cable to the front panel.
          > >
          > > The adjacent ribbon cable signal is -DO5 - the only data-out signal sent to the front panel. With the right combination of data bytes written by the processor, the processor was getting a random -RDYIN glitch.
          > >
          > > Just for fun, I swapped out the 7404 on the good board for a 74LS04. Sure enough - it failed too. This made me conclude that the problem is more universal than just my bad boy 8800b - you might have seen the same problem.
          > >
          > > The 7404 works because it draws 4 times the current on its input as does the 74LS04 - and so it "eats" the glitch, reducing the glitch amplitude to around 0.75V.
          > >
          > > For extra margin, I replaced IC 'R' with a 7414, which has the higher input current, but is also a Schmitt trigger, so that it could tolerate a higher glitch voltage.
          > >
          > > So as I say, I recommend replacing IC 'R' on the D/C Interface board with a 7414.
          > >
          > > -Martin E
          > >
          >
        • deramp5113
          Interesting. My board has an LS04 in position R but the computer has been pretty solid for quite a while now. I ll have to pull out a scope and see if my
          Message 4 of 6 , Aug 24, 2013
          • 0 Attachment
            Interesting. My board has an 'LS04 in position 'R' but the computer has been pretty solid for quite a while now. I'll have to pull out a scope and see if my hardware is coming close to experiencing the glitch you noticed. One subtle difference -- all of the chips on my DC interface board are socketed.

            Soon I'll be working on this 8800b extensively for testing of a floppy disk controller I'm building. I definitely don't want random variables like this problem thrown into the middle of hardware debug of the new board, so I'm going to dig into this one. Thanks for the heads-up!

            Mike


            --- In altaircomputerclub@yahoogroups.com, "mfeberhard" <eberhard@...> wrote:
            >
            > I tried several 74LS04's in this position, from several manufacturers. All failed. I tried several 7404's, all worked fine. Same with 7414's.
            >
            > The particular chip in position 'Y' made a difference on how bad the failure was with the 74LS04. IC 'Y' is the driver for the adjacent signal on the ribbon cable. I would get a variety of flaky behaviors depending on which particular chip I put in IC 'Y', so long as a 74LS04 was in 'R'.
            >
            >
            >
            > --- In altaircomputerclub@yahoogroups.com, "deramp5113" <deramp5113@> wrote:
            > >
            > > Was this the same, original 74LS04 moved between the two computers, or were you able to make them both fail with a "new" 74LS04 as well?
            > >
            > > Mike
            > >
            > > --- In altaircomputerclub@yahoogroups.com, "mfeberhard" <eberhard@> wrote:
            > > >
            > > > I recently debugged an 8800b front panel, and I thought I'd share what I learned with those of you who might also be working on 8800b's.
            > > >
            > > > The conclusion first: I strongly recommend replacing IC 'R' on the D/C Interface board with a 7414. The manual calls for a 74LS04 here, which is not reliable.
            > > >
            > > > This 8800b of mine was flaky - it would occasionally be able to boot Basic from disk. But normally it would fail with an 'M' error, indicating that the monitor did not read back the same data that it wrote to memory.
            > > >
            > > > The most repeatable symptom that I saw was using TURMON. Using the 'M' command, it would write successfully to memory for addresses 000000 through 000003, and would normally fail (and crash) when it tried to write to address 000004. Sometimes it would fail at other addresses. Occasionally, it would work fine at address 000004. When I looked in memory (with the front panel), the 'failed' write had actually written the correct data!
            > > >
            > > > The worst part was that nothing was repeatable - the bad behavior seemed to morph as I chased the problem.
            > > >
            > > > Perhaps you have seen the same misbehavior on your 8800b...
            > > >
            > > > I have another (working) 8800b, so I was able to narrow the problem down to the D/C Interface board by swapping boards.
            > > >
            > > > It seemed like the ribbon cables to the front panel were not constantly making good contact, so I replaced them - no easy task! Then I one-by-one replaced nearly every single chip on the board, with no improvement. And none of these chips was socketed :-(
            > > >
            > > > Finally I noticed that IC 'R' on my good D/C Interface board was not a 74LS04 as the assembly instructions indicate. Instead, it was a plain 7404. So I substituted a 7404 onto the troublesome board, and the board suddenly works perfectly.
            > > >
            > > > This was a bit puzzling to me, because none of the inverters in IC 'R' are involved in *any* switching during normal operation. 2 of the inverters are in the PROT (memory protect) circuitry, which doesn't do anything. One is in the PRESET path. One drives Single Step onto the bus, which goes to nothing. One inverts the -FRDY signal, which is used to stop/start the processor when the front panel is in the STOP mode, and is always high when in RUN mode.
            > > >
            > > > So I put the 74LS04 back, and set my 'scope to trigger at 1.25V, and looked at the inputs to IC 'R' when running. And sure enough, the -FRDY signal was occasionally experiencing a >1.25V glitch, induced by the adjacent signal on the ribbon cable to the front panel.
            > > >
            > > > The adjacent ribbon cable signal is -DO5 - the only data-out signal sent to the front panel. With the right combination of data bytes written by the processor, the processor was getting a random -RDYIN glitch.
            > > >
            > > > Just for fun, I swapped out the 7404 on the good board for a 74LS04. Sure enough - it failed too. This made me conclude that the problem is more universal than just my bad boy 8800b - you might have seen the same problem.
            > > >
            > > > The 7404 works because it draws 4 times the current on its input as does the 74LS04 - and so it "eats" the glitch, reducing the glitch amplitude to around 0.75V.
            > > >
            > > > For extra margin, I replaced IC 'R' with a 7414, which has the higher input current, but is also a Schmitt trigger, so that it could tolerate a higher glitch voltage.
            > > >
            > > > So as I say, I recommend replacing IC 'R' on the D/C Interface board with a 7414.
            > > >
            > > > -Martin E
            > > >
            > >
            >
          • Kip Koon
            Hi Guys! I have an Altair 8800 with 3 - 4 slot motherboards wired together into one motherboard to give you some idea of the type of system it is. The
            Message 5 of 6 , Aug 24, 2013
            • 0 Attachment

              Hi Guys!

              I have an Altair 8800 with 3 – 4 slot motherboards wired together into one motherboard to give you some idea of the type of system it is.  The previous owner said that it did not work, that the front panel did not work.   After following what you guys have been saying about the R IC being a 74LS04 and needing to be a 7414,  I wonder if that is the problem with mine?  I will have to check this out.  If it turns out that this IS true, you all will have solved a problem that I had no idea how to troubleshoot much less fix!  Once I verify this and if true fix it, I’ll definitely have to check out the power supply to see if it will work.  It has not been turned on in literally more than 3 decades, maybe more!  Talking about resurrecting a computer!  It Will be interesting.  I’ll keep you guys updated as to my progress as soon as I can get to it.  My computer repair business is growing finally after 5 years of limping along on two broken legs, J so it looks like I will be able to eventually restore my not so little Altair 8800!  Take care guys!

              Kip

               

              From: altaircomputerclub@yahoogroups.com [mailto:altaircomputerclub@yahoogroups.com] On Behalf Of deramp5113
              Sent: Saturday, August 24, 2013 9:52 AM
              To: altaircomputerclub@yahoogroups.com
              Subject: [Altair Computer Club] Re: 8800b front panel glitch

               

               

              Interesting. My board has an 'LS04 in position 'R' but the computer has been pretty solid for quite a while now. I'll have to pull out a scope and see if my hardware is coming close to experiencing the glitch you noticed. One subtle difference -- all of the chips on my DC interface board are socketed.

              Soon I'll be working on this 8800b extensively for testing of a floppy disk controller I'm building. I definitely don't want random variables like this problem thrown into the middle of hardware debug of the new board, so I'm going to dig into this one. Thanks for the heads-up!

              Mike

              --- In altaircomputerclub@yahoogroups.com, "mfeberhard" <eberhard@...> wrote:
              >
              > I tried several 74LS04's in this position, from several manufacturers. All failed. I tried several 7404's, all worked fine. Same with 7414's.
              >
              > The particular chip in position 'Y' made a difference on how bad the failure was with the 74LS04. IC 'Y' is the driver for the adjacent signal on the ribbon cable. I would get a variety of flaky behaviors depending on which particular chip I put in IC 'Y', so long as a 74LS04 was in 'R'.
              >
              >
              >
              > --- In altaircomputerclub@yahoogroups.com, "deramp5113" <deramp5113@> wrote:
              > >
              > > Was this the same, original 74LS04 moved between the two computers, or were you able to make them both fail with a "new" 74LS04 as well?
              > >
              > > Mike
              > >
              > > --- In altaircomputerclub@yahoogroups.com, "mfeberhard" <eberhard@> wrote:
              > > >
              > > > I recently debugged an 8800b front panel, and I thought I'd share what I learned with those of you who might also be working on 8800b's.
              > > >
              > > > The conclusion first: I strongly recommend replacing IC 'R' on the D/C Interface board with a 7414. The manual calls for a 74LS04 here, which is not reliable.
              > > >
              > > > This 8800b of mine was flaky - it would occasionally be able to boot Basic from disk. But normally it would fail with an 'M' error, indicating that the monitor did not read back the same data that it wrote to memory.
              > > >
              > > > The most repeatable symptom that I saw was using TURMON. Using the 'M' command, it would write successfully to memory for addresses 000000 through 000003, and would normally fail (and crash) when it tried to write to address 000004. Sometimes it would fail at other addresses. Occasionally, it would work fine at address 000004. When I looked in memory (with the front panel), the 'failed' write had actually written the correct data!
              > > >
              > > > The worst part was that nothing was repeatable - the bad behavior seemed to morph as I chased the problem.
              > > >
              > > > Perhaps you have seen the same misbehavior on your 8800b...
              > > >
              > > > I have another (working) 8800b, so I was able to narrow the problem down to the D/C Interface board by swapping boards.
              > > >
              > > > It seemed like the ribbon cables to the front panel were not constantly making good contact, so I replaced them - no easy task! Then I one-by-one replaced nearly every single chip on the board, with no improvement. And none of these chips was socketed :-(
              > > >
              > > > Finally I noticed that IC 'R' on my good D/C Interface board was not a 74LS04 as the assembly instructions indicate. Instead, it was a plain 7404. So I substituted a 7404 onto the troublesome board, and the board suddenly works perfectly.
              > > >
              > > > This was a bit puzzling to me, because none of the inverters in IC 'R' are involved in *any* switching during normal operation. 2 of the inverters are in the PROT (memory protect) circuitry, which doesn't do anything. One is in the PRESET path. One drives Single Step onto the bus, which goes to nothing. One inverts the -FRDY signal, which is used to stop/start the processor when the front panel is in the STOP mode, and is always high when in RUN mode.
              > > >
              > > > So I put the 74LS04 back, and set my 'scope to trigger at 1.25V, and looked at the inputs to IC 'R' when running. And sure enough, the -FRDY signal was occasionally experiencing a >1.25V glitch, induced by the adjacent signal on the ribbon cable to the front panel.
              > > >
              > > > The adjacent ribbon cable signal is -DO5 - the only data-out signal sent to the front panel. With the right combination of data bytes written by the processor, the processor was getting a random -RDYIN glitch.
              > > >
              > > > Just for fun, I swapped out the 7404 on the good board for a 74LS04. Sure enough - it failed too. This made me conclude that the problem is more universal than just my bad boy 8800b - you might have seen the same problem.
              > > >
              > > > The 7404 works because it draws 4 times the current on its input as does the 74LS04 - and so it "eats" the glitch, reducing the glitch amplitude to around 0.75V.
              > > >
              > > > For extra margin, I replaced IC 'R' with a 7414, which has the higher input current, but is also a Schmitt trigger, so that it could tolerate a higher glitch voltage.
              > > >
              > > > So as I say, I recommend replacing IC 'R' on the D/C Interface board with a 7414.
              > > >
              > > > -Martin E
              > > >
              > >
              >

            • steve
              Kip, The 8800 and 8800a don t have a D/C Interface board, so your problems lie elsewhere. A common cause of trouble with your machine is the one-shot devices
              Message 6 of 6 , Aug 25, 2013
              • 0 Attachment
                Kip,

                The 8800 and 8800a don't have a D/C Interface board, so your problems lie elsewhere. A common cause of trouble with your machine is the one-shot devices on the front panel. Their timing can drift.

                Steve
                ----------

                --- In altaircomputerclub@yahoogroups.com, "Kip Koon" <computerdoc@...> wrote:
                >
                > Hi Guys!
                >
                > I have an Altair 8800 with 3 - 4 slot motherboards wired together into one
                > motherboard to give you some idea of the type of system it is. The previous
                > owner said that it did not work, that the front panel did not work. After
                > following what you guys have been saying about the R IC being a 74LS04 and
                > needing to be a 7414, I wonder if that is the problem with mine? I will
                > have to check this out. If it turns out that this IS true, you all will
                > have solved a problem that I had no idea how to troubleshoot much less fix!
                > Once I verify this and if true fix it, I'll definitely have to check out the
                > power supply to see if it will work. It has not been turned on in literally
                > more than 3 decades, maybe more! Talking about resurrecting a computer! It
                > Will be interesting. I'll keep you guys updated as to my progress as soon
                > as I can get to it. My computer repair business is growing finally after 5
                > years of limping along on two broken legs, J so it looks like I will be able
                > to eventually restore my not so little Altair 8800! Take care guys!
                >
                > Kip
                >
                >
                >
                > From: altaircomputerclub@yahoogroups.com
                > [mailto:altaircomputerclub@yahoogroups.com] On Behalf Of deramp5113
                > Sent: Saturday, August 24, 2013 9:52 AM
                > To: altaircomputerclub@yahoogroups.com
                > Subject: [Altair Computer Club] Re: 8800b front panel glitch
                >
                >
                >
                >
                >
                > Interesting. My board has an 'LS04 in position 'R' but the computer has been
                > pretty solid for quite a while now. I'll have to pull out a scope and see if
                > my hardware is coming close to experiencing the glitch you noticed. One
                > subtle difference -- all of the chips on my DC interface board are socketed.
                >
                > Soon I'll be working on this 8800b extensively for testing of a floppy disk
                > controller I'm building. I definitely don't want random variables like this
                > problem thrown into the middle of hardware debug of the new board, so I'm
                > going to dig into this one. Thanks for the heads-up!
                >
                > Mike
                >
                > --- In altaircomputerclub@yahoogroups.com
                > <mailto:altaircomputerclub%40yahoogroups.com> , "mfeberhard" <eberhard@>
                > wrote:
                > >
                > > I tried several 74LS04's in this position, from several manufacturers. All
                > failed. I tried several 7404's, all worked fine. Same with 7414's.
                > >
                > > The particular chip in position 'Y' made a difference on how bad the
                > failure was with the 74LS04. IC 'Y' is the driver for the adjacent signal on
                > the ribbon cable. I would get a variety of flaky behaviors depending on
                > which particular chip I put in IC 'Y', so long as a 74LS04 was in 'R'.
                > >
                > >
                > >
                > > --- In altaircomputerclub@yahoogroups.com
                > <mailto:altaircomputerclub%40yahoogroups.com> , "deramp5113" <deramp5113@>
                > wrote:
                > > >
                > > > Was this the same, original 74LS04 moved between the two computers, or
                > were you able to make them both fail with a "new" 74LS04 as well?
                > > >
                > > > Mike
                > > >
                > > > --- In altaircomputerclub@yahoogroups.com
                > <mailto:altaircomputerclub%40yahoogroups.com> , "mfeberhard" <eberhard@>
                > wrote:
                > > > >
                > > > > I recently debugged an 8800b front panel, and I thought I'd share what
                > I learned with those of you who might also be working on 8800b's.
                > > > >
                > > > > The conclusion first: I strongly recommend replacing IC 'R' on the D/C
                > Interface board with a 7414. The manual calls for a 74LS04 here, which is
                > not reliable.
                > > > >
                > > > > This 8800b of mine was flaky - it would occasionally be able to boot
                > Basic from disk. But normally it would fail with an 'M' error, indicating
                > that the monitor did not read back the same data that it wrote to memory.
                > > > >
                > > > > The most repeatable symptom that I saw was using TURMON. Using the 'M'
                > command, it would write successfully to memory for addresses 000000 through
                > 000003, and would normally fail (and crash) when it tried to write to
                > address 000004. Sometimes it would fail at other addresses. Occasionally, it
                > would work fine at address 000004. When I looked in memory (with the front
                > panel), the 'failed' write had actually written the correct data!
                > > > >
                > > > > The worst part was that nothing was repeatable - the bad behavior
                > seemed to morph as I chased the problem.
                > > > >
                > > > > Perhaps you have seen the same misbehavior on your 8800b...
                > > > >
                > > > > I have another (working) 8800b, so I was able to narrow the problem
                > down to the D/C Interface board by swapping boards.
                > > > >
                > > > > It seemed like the ribbon cables to the front panel were not
                > constantly making good contact, so I replaced them - no easy task! Then I
                > one-by-one replaced nearly every single chip on the board, with no
                > improvement. And none of these chips was socketed :-(
                > > > >
                > > > > Finally I noticed that IC 'R' on my good D/C Interface board was not a
                > 74LS04 as the assembly instructions indicate. Instead, it was a plain 7404.
                > So I substituted a 7404 onto the troublesome board, and the board suddenly
                > works perfectly.
                > > > >
                > > > > This was a bit puzzling to me, because none of the inverters in IC 'R'
                > are involved in *any* switching during normal operation. 2 of the inverters
                > are in the PROT (memory protect) circuitry, which doesn't do anything. One
                > is in the PRESET path. One drives Single Step onto the bus, which goes to
                > nothing. One inverts the -FRDY signal, which is used to stop/start the
                > processor when the front panel is in the STOP mode, and is always high when
                > in RUN mode.
                > > > >
                > > > > So I put the 74LS04 back, and set my 'scope to trigger at 1.25V, and
                > looked at the inputs to IC 'R' when running. And sure enough, the -FRDY
                > signal was occasionally experiencing a >1.25V glitch, induced by the
                > adjacent signal on the ribbon cable to the front panel.
                > > > >
                > > > > The adjacent ribbon cable signal is -DO5 - the only data-out signal
                > sent to the front panel. With the right combination of data bytes written by
                > the processor, the processor was getting a random -RDYIN glitch.
                > > > >
                > > > > Just for fun, I swapped out the 7404 on the good board for a 74LS04.
                > Sure enough - it failed too. This made me conclude that the problem is more
                > universal than just my bad boy 8800b - you might have seen the same problem.
                > > > >
                > > > > The 7404 works because it draws 4 times the current on its input as
                > does the 74LS04 - and so it "eats" the glitch, reducing the glitch amplitude
                > to around 0.75V.
                > > > >
                > > > > For extra margin, I replaced IC 'R' with a 7414, which has the higher
                > input current, but is also a Schmitt trigger, so that it could tolerate a
                > higher glitch voltage.
                > > > >
                > > > > So as I say, I recommend replacing IC 'R' on the D/C Interface board
                > with a 7414.
                > > > >
                > > > > -Martin E
                > > > >
                > > >
                > >
                >
              Your message has been successfully submitted and would be delivered to recipients shortly.