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

Re: Dead 800

Expand Messages
  • selloutdesigns
    I just got my scope and hooked it up last night. The first odd thing I noticed was that CPU RES 7.5 (interrupt oscillator that drives the display and EG/MG
    Message 1 of 21 , Mar 3, 2009
    • 0 Attachment
      I just got my scope and hooked it up last night. The first odd thing I noticed was that CPU RES 7.5 (interrupt oscillator that drives the display and EG/MG among other things) was running at about .34ms per cycle, which is about a 1000 times slower than what it should be (.34usec). Because of the exact 1000x multiplier, I thought maybe I was misreading the scope, but I do not believe I am. For example, when I measure the CPU Clock Out it measures at .3 usec per cycle, but as soon as I switch the probe to CPU RES 7.5 I have to crank up the Time/Div setting. All components check out in that part of the circuit and everything is connected and powered as it should. Any insight as to what might cause this? It does seem if this oscillator is running slow that might be the cause of my display issue.

      Other oddities which may be related: CPU ALE seems to be slower than normal, at about 1.375usec per cycle vs. .18usec where it is supposed to be. IO/M is always low. On M81C55, Timer Out is always high.

      Thanks is advance for any guidance anyone may have!

      <<aaRoN>>

      --- In korgpolyex@yahoogroups.com, Michael Hawkins <korgpolyex800@...> wrote:
      >
      > If you have goop from what might have been a battery leak then things can look OK but turn out to be alot worse than you think.
      >
      > Nevertheless, you might be OK. A scope is going to help alot.
      >
      > For example, IO/M signal will spend most of it's time low because it is only high for IO. In the Poly, all of the hardware is mapped into normal memory except for the 81C55. So the IO/M signal is almost 100% low. But a scope will show you thos tiny little 20uS high periods. That would tell you that the CPU was actually doing something as opposed to completely comatose.
      >
      > Mike.
      >
      >
      >
      >
      >
      >
      > ________________________________
      > From: selloutdesigns <aah_public@...>
      > To: korgpolyex@yahoogroups.com
      > Sent: Wednesday, February 25, 2009 10:57:05 AM
      > Subject: [korgpolyex] Re: Dead 800
      >
      >
      > Took another look internally last night. There was some sticky gunk on
      > some of the boards which may have been from a battery compartment leak
      > at some point, but it cleaned up fine and there appears to be no trace
      > or component damage. As best I can tell, all the rails are showing
      > correct voltages. Strangely I found that the rails on the main board
      > differed slightly from the smaller board (KLM-598), being about 4.6V.
      > However, adjusting VR1 slightly boosted this voltage to 5V, while
      > causing voltage rails on the main board to be slightly above 5V. I am
      > guessing this is all within tolerable range though.
      >
      > I also took voltage measurements on all the pins on IC24 (M80C85A) and
      > IC25 (M81C55). I see a voltage on all the output pins of IC24
      > connecting to the input pins of IC25. However, the output pins of IC25
      > (which feed the display driver circuit) all show 0V. IC24 and 25 both
      > are powered. I did notice that IO/M (both), RESOUT (IC24) and RES
      > (IC25) were showing 0V. Is it correct to assume that if IO/M is always
      > low, that IC25 will never output any data? What would be keeping IC24
      > from toggling IO/M to high?
      >
      > Thanks in advance for any insight/guidance. Anyone have scans of a
      > Poly 800 Service Manual they can share? :)
      >
      > --- In korgpolyex@yahoogro ups.com, Michael Hawkins <korgpolyex800@ ...>
      > wrote:
      > >
      > > Aaron,
      > >
      > > you should check all voltage rails. especially +5 and -5 volt rails.
      > >
      > > The power supply of the Poly is somewhat reliable. Note the "somewhat".
      > >
      > > Sometimes it does fail or drift enough to cause Poly's to die.
      > >
      > > If you don't have a display showing then tape and patches etc are 99
      > percent NOT going to work.
      > >
      > > Even when memory is corrupted due to battery failure, those
      > display's will at least show something.
      > >
      > > Also, I would give the thing a complete internal thorough check out
      > looking for battery acid damage, broken wires, misaligned or misplaced
      > connectors etc.
      > >
      > > Sometimes it's the simple things.
      > >
      > > Mike.
      > >
      > >
      > >
      > >
      > >
      > > ____________ _________ _________ __
      > > From: selloutdesigns <aah_public@ ...>
      > > To: korgpolyex@yahoogro ups.com
      > > Sent: Tuesday, February 24, 2009 9:48:45 AM
      > > Subject: [korgpolyex] Re: Dead 800
      > >
      > >
      > > I gave this a try on my dead Poly, but no luck. It seems that VR1
      > > adjusts the +5V rail voltage and subsequently the supply voltage to
      > > the CPU, but in my case it seems like all the chips are getting
      > > adequate power.
      > >
      > > <<aaRoN>>
      > > --- In korgpolyex@yahoogro ups.com, Russ <russdaren@ ..> wrote:
      > > >
      > > > I had this problem a while ago and got her back upby fiddling with
      > > one of the VR pots. Cant recal which one did it. But one of them
      > > seemed to control voltage for the CPU board. It wnet from no display,
      > > to garbled, to functioning just fine ever since.
      > > >
      > >
      >
    • narfman96
      Definitely timing issues with IC36b the interrupt oscillator. Try disconnecting the output at R107 and see if it runs at the correct frequency. If it does the
      Message 2 of 21 , Mar 4, 2009
      • 0 Attachment
        Definitely timing issues with IC36b the interrupt oscillator. Try disconnecting the output at R107 and see if it runs at the correct frequency. If it does the CPU is loading it down somehow. If not then suspect the timing caps C76 or C79. The resistors R102-R104 need to be verified and then the Tl072 op amp itself. This circuit needs to run in the mhz or no go for the LED display and envelope generators.

        --- In korgpolyex@yahoogroups.com, "selloutdesigns" <aah_public@...> wrote:
        >
        > I just got my scope and hooked it up last night. The first odd thing I noticed was that CPU RES 7.5 (interrupt oscillator that drives the display and EG/MG among other things) was running at about .34ms per cycle, which is about a 1000 times slower than what it should be (.34usec). Because of the exact 1000x multiplier, I thought maybe I was misreading the scope, but I do not believe I am. For example, when I measure the CPU Clock Out it measures at .3 usec per cycle, but as soon as I switch the probe to CPU RES 7.5 I have to crank up the Time/Div setting. All components check out in that part of the circuit and everything is connected and powered as it should. Any insight as to what might cause this? It does seem if this oscillator is running slow that might be the cause of my display issue.
        >
        > Other oddities which may be related: CPU ALE seems to be slower than normal, at about 1.375usec per cycle vs. .18usec where it is supposed to be. IO/M is always low. On M81C55, Timer Out is always high.
        >
        > Thanks is advance for any guidance anyone may have!
        >
        > <<aaRoN>>
        >
        > --- In korgpolyex@yahoogroups.com, Michael Hawkins <korgpolyex800@> wrote:
        > >
        > > If you have goop from what might have been a battery leak then things can look OK but turn out to be alot worse than you think.
        > >
        > > Nevertheless, you might be OK. A scope is going to help alot.
        > >
        > > For example, IO/M signal will spend most of it's time low because it is only high for IO. In the Poly, all of the hardware is mapped into normal memory except for the 81C55. So the IO/M signal is almost 100% low. But a scope will show you thos tiny little 20uS high periods. That would tell you that the CPU was actually doing something as opposed to completely comatose.
        > >
        > > Mike.
        > >
        > >
        > >
        > >
        > >
        > >
        > > ________________________________
        > > From: selloutdesigns <aah_public@>
        > > To: korgpolyex@yahoogroups.com
        > > Sent: Wednesday, February 25, 2009 10:57:05 AM
        > > Subject: [korgpolyex] Re: Dead 800
        > >
        > >
        > > Took another look internally last night. There was some sticky gunk on
        > > some of the boards which may have been from a battery compartment leak
        > > at some point, but it cleaned up fine and there appears to be no trace
        > > or component damage. As best I can tell, all the rails are showing
        > > correct voltages. Strangely I found that the rails on the main board
        > > differed slightly from the smaller board (KLM-598), being about 4.6V.
        > > However, adjusting VR1 slightly boosted this voltage to 5V, while
        > > causing voltage rails on the main board to be slightly above 5V. I am
        > > guessing this is all within tolerable range though.
        > >
        > > I also took voltage measurements on all the pins on IC24 (M80C85A) and
        > > IC25 (M81C55). I see a voltage on all the output pins of IC24
        > > connecting to the input pins of IC25. However, the output pins of IC25
        > > (which feed the display driver circuit) all show 0V. IC24 and 25 both
        > > are powered. I did notice that IO/M (both), RESOUT (IC24) and RES
        > > (IC25) were showing 0V. Is it correct to assume that if IO/M is always
        > > low, that IC25 will never output any data? What would be keeping IC24
        > > from toggling IO/M to high?
        > >
        > > Thanks in advance for any insight/guidance. Anyone have scans of a
        > > Poly 800 Service Manual they can share? :)
        > >
        > > --- In korgpolyex@yahoogro ups.com, Michael Hawkins <korgpolyex800@ ...>
        > > wrote:
        > > >
        > > > Aaron,
        > > >
        > > > you should check all voltage rails. especially +5 and -5 volt rails.
        > > >
        > > > The power supply of the Poly is somewhat reliable. Note the "somewhat".
        > > >
        > > > Sometimes it does fail or drift enough to cause Poly's to die.
        > > >
        > > > If you don't have a display showing then tape and patches etc are 99
        > > percent NOT going to work.
        > > >
        > > > Even when memory is corrupted due to battery failure, those
        > > display's will at least show something.
        > > >
        > > > Also, I would give the thing a complete internal thorough check out
        > > looking for battery acid damage, broken wires, misaligned or misplaced
        > > connectors etc.
        > > >
        > > > Sometimes it's the simple things.
        > > >
        > > > Mike.
        > > >
        > > >
        > > >
        > > >
        > > >
        > > > ____________ _________ _________ __
        > > > From: selloutdesigns <aah_public@ ...>
        > > > To: korgpolyex@yahoogro ups.com
        > > > Sent: Tuesday, February 24, 2009 9:48:45 AM
        > > > Subject: [korgpolyex] Re: Dead 800
        > > >
        > > >
        > > > I gave this a try on my dead Poly, but no luck. It seems that VR1
        > > > adjusts the +5V rail voltage and subsequently the supply voltage to
        > > > the CPU, but in my case it seems like all the chips are getting
        > > > adequate power.
        > > >
        > > > <<aaRoN>>
        > > > --- In korgpolyex@yahoogro ups.com, Russ <russdaren@ ..> wrote:
        > > > >
        > > > > I had this problem a while ago and got her back upby fiddling with
        > > > one of the VR pots. Cant recal which one did it. But one of them
        > > > seemed to control voltage for the CPU board. It wnet from no display,
        > > > to garbled, to functioning just fine ever since.
        > > > >
        > > >
        > >
        >
      • Gordon JC Pearce MM3YEQ
        ... Bah. Just change the TL072 on spec. Horrible things. Gordon
        Message 3 of 21 , Mar 4, 2009
        • 0 Attachment
          On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
          > Definitely timing issues with IC36b the interrupt oscillator. Try
          > disconnecting the output at R107 and see if it runs at the correct
          > frequency. If it does the CPU is loading it down somehow. If not then
          > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
          > verified and then the Tl072 op amp itself. This circuit needs to run
          > in the mhz or no go for the LED display and envelope generators.

          Bah. Just change the TL072 on spec. Horrible things.

          Gordon
        • selloutdesigns
          Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible. Thanks,
          Message 4 of 21 , Mar 4, 2009
          • 0 Attachment
            Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible.

            Thanks,

            <<aaRoN>>

            --- In korgpolyex@yahoogroups.com, Gordon JC Pearce MM3YEQ <gordon@...> wrote:
            >
            > On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
            > > Definitely timing issues with IC36b the interrupt oscillator. Try
            > > disconnecting the output at R107 and see if it runs at the correct
            > > frequency. If it does the CPU is loading it down somehow. If not then
            > > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
            > > verified and then the Tl072 op amp itself. This circuit needs to run
            > > in the mhz or no go for the LED display and envelope generators.
            >
            > Bah. Just change the TL072 on spec. Horrible things.
            >
            > Gordon
            >
          • narfman96
            Should work fine. Install a good socket if you want. Then you can use a TL072 when you get one.
            Message 5 of 21 , Mar 4, 2009
            • 0 Attachment
              Should work fine. Install a good socket if you want. Then you can use a TL072 when you get one.

              --- In korgpolyex@yahoogroups.com, "selloutdesigns" <aah_public@...> wrote:
              >
              > Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible.
              >
              > Thanks,
              >
              > <<aaRoN>>
              >
              > --- In korgpolyex@yahoogroups.com, Gordon JC Pearce MM3YEQ <gordon@> wrote:
              > >
              > > On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
              > > > Definitely timing issues with IC36b the interrupt oscillator. Try
              > > > disconnecting the output at R107 and see if it runs at the correct
              > > > frequency. If it does the CPU is loading it down somehow. If not then
              > > > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
              > > > verified and then the Tl072 op amp itself. This circuit needs to run
              > > > in the mhz or no go for the LED display and envelope generators.
              > >
              > > Bah. Just change the TL072 on spec. Horrible things.
              > >
              > > Gordon
              > >
              >
            • selloutdesigns
              Hmm, now I am starting to think I m chasing my tail. In the schematic (pg 6 in Service Manual) for the main board, it shows CPU RES 7.5 as running at .34usec
              Message 6 of 21 , Mar 4, 2009
              • 0 Attachment
                Hmm, now I am starting to think I'm chasing my tail. In the schematic (pg 6 in Service Manual) for the main board, it shows CPU RES 7.5 as running at .34usec per cycle, or about 2.94 Mhz, which is what got me thinking I had a problem there. However, in the description of the Interrupt Oscillator circuit in the Service Guide (pg 18) it reads:

                "This oscillator cycle is used for the EG, MG, LED displays, and S/H time division processing. Oscillator frequency is 24OOHz-360OHz. Interrupt order is by priority. If this circuit fails, EG operation and LED indication may become erratic."

                I measured the oscillator at about .34msec per cycle, or 2941Hz, which seems normal according to that.

                So which is correct? The schematic or the circuit description? I'm starting to think the circuit description. I've tested all the components except the TL072, but all seem fine.

                <<aaRoN>>

                6) -Address Decoder:

                --- In korgpolyex@yahoogroups.com, "narfman96" <narfman96@...> wrote:
                >
                > Should work fine. Install a good socket if you want. Then you can use a TL072 when you get one.
                >
                > --- In korgpolyex@yahoogroups.com, "selloutdesigns" <aah_public@> wrote:
                > >
                > > Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible.
                > >
                > > Thanks,
                > >
                > > <<aaRoN>>
                > >
                > > --- In korgpolyex@yahoogroups.com, Gordon JC Pearce MM3YEQ <gordon@> wrote:
                > > >
                > > > On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
                > > > > Definitely timing issues with IC36b the interrupt oscillator. Try
                > > > > disconnecting the output at R107 and see if it runs at the correct
                > > > > frequency. If it does the CPU is loading it down somehow. If not then
                > > > > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
                > > > > verified and then the Tl072 op amp itself. This circuit needs to run
                > > > > in the mhz or no go for the LED display and envelope generators.
                > > >
                > > > Bah. Just change the TL072 on spec. Horrible things.
                > > >
                > > > Gordon
                > > >
                > >
                >
              • Michael Hawkins
                The circuit description is correct. Interrupt RES7.5 runs at around 2400 to 3600 hz. It can t run too fast otherwise the CPU won t do anything else except
                Message 7 of 21 , Mar 4, 2009
                • 0 Attachment
                  The circuit description is correct.

                  Interrupt RES7.5 runs at around 2400 to 3600 hz.

                  It can't run too fast otherwise the CPU won't do anything else except service this hardware interrupt.

                  A rate of 2.94Mhz is only half the full clock rate of the CPU. So at that speed, the CPU would not even get to fetch one instruction before being interrupted.

                  Mike.



                  From: selloutdesigns <aah_public@...>
                  To: korgpolyex@yahoogroups.com
                  Sent: Wednesday, March 4, 2009 4:22:11 PM
                  Subject: [korgpolyex] Re: Dead 800

                  Hmm, now I am starting to think I'm chasing my tail. In the schematic (pg 6 in Service Manual) for the main board, it shows CPU RES 7.5 as running at .34usec per cycle, or about 2.94 Mhz, which is what got me thinking I had a problem there. However, in the description of the Interrupt Oscillator circuit in the Service Guide (pg 18) it reads:

                  "This oscillator cycle is used for the EG, MG, LED displays, and S/H time division processing. Oscillator frequency is 24OOHz-360OHz. Interrupt order is by priority. If this circuit fails, EG operation and LED indication may become erratic."

                  I measured the oscillator at about .34msec per cycle, or 2941Hz, which seems normal according to that.

                  So which is correct? The schematic or the circuit description? I'm starting to think the circuit description. I've tested all the components except the TL072, but all seem fine.

                  <<aaRoN>>

                  6) -Address Decoder:

                  --- In korgpolyex@yahoogro ups.com, "narfman96" <narfman96@. ..> wrote:
                  >
                  > Should work fine. Install a good socket if you want. Then you can use a TL072 when you get one.
                  >
                  > --- In korgpolyex@yahoogro ups.com, "selloutdesigns" <aah_public@ > wrote:
                  > >
                  > > Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible.
                  > >
                  > > Thanks,
                  > >
                  > > <<aaRoN>>
                  > >
                  > > --- In korgpolyex@yahoogro ups.com, Gordon JC Pearce MM3YEQ <gordon@> wrote:
                  > > >
                  > > > On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
                  > > > > Definitely timing issues with IC36b the interrupt oscillator. Try
                  > > > > disconnecting the output at R107 and see if it runs at the correct
                  > > > > frequency. If it does the CPU is loading it down somehow. If not then
                  > > > > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
                  > > > > verified and then the Tl072 op amp itself. This circuit needs to run
                  > > > > in the mhz or no go for the LED display and envelope generators.
                  > > >
                  > > > Bah. Just change the TL072 on spec. Horrible things.
                  > > >
                  > > > Gordon
                  > > >
                  > >
                  >


                • zoinky420
                  ... tl072 is the low noise version of tl082, but I wonder if tl062, which is the battery power version would extend Poly800 battery life?? 1mA for the
                  Message 8 of 21 , Mar 5, 2009
                  • 0 Attachment
                    --- In korgpolyex@yahoogroups.com, "selloutdesigns" <aah_public@...> wrote:
                    >
                    > Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible.
                    >

                    tl072 is the 'low noise' version of tl082, but I wonder if tl062, which is the 'battery power' version would extend Poly800 battery life?? "1mA for the TL072 against less than 500uA for the TL062"


                    > Thanks,
                    >
                    > <<aaRoN>>
                    >
                    > --- In korgpolyex@yahoogroups.com, Gordon JC Pearce MM3YEQ <gordon@> wrote:
                    > >
                    > > On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
                    > > > Definitely timing issues with IC36b the interrupt oscillator. Try
                    > > > disconnecting the output at R107 and see if it runs at the correct
                    > > > frequency. If it does the CPU is loading it down somehow. If not then
                    > > > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
                    > > > verified and then the Tl072 op amp itself. This circuit needs to run
                    > > > in the mhz or no go for the LED display and envelope generators.
                    > >
                    > > Bah. Just change the TL072 on spec. Horrible things.
                    > >
                    > > Gordon
                    > >
                    >
                  • selloutdesigns
                    Okay, I m totally stumped at this point. The main issue I see now with the display is that the IO/M pin on the CPU is low 100% of the time, as confirmed with a
                    Message 9 of 21 , Mar 6, 2009
                    • 0 Attachment
                      Okay, I'm totally stumped at this point. The main issue I see now with the display is that the IO/M pin on the CPU is low 100% of the time, as confirmed with a scope reading. However, I have no idea why. SOD is also low all the time as well, which is probably related. The CPU clock is fine and I see RD,WR,S0,S1 going high and low on a regular basis so it seems memory reads and writes are occurring fine, but no I/O whatsoever. I'm wondering if there is an issue with the CPU receiving instructions from memory, which is causing the I/O cycle to never occur.

                      In any case, I have just about given up in trying to figure out what's going on. I think this unit is destined to land on someone else's bench. If it does get fixed I'll be sure to find out what the cause was.

                      Thanks to all for the input.

                      <<aaRoN>>

                      --- In korgpolyex@yahoogroups.com, Michael Hawkins <korgpolyex800@...> wrote:
                      >
                      > The circuit description is correct.
                      >
                      > Interrupt RES7.5 runs at around 2400 to 3600 hz.
                      >
                      > It can't run too fast otherwise the CPU won't do anything else except service this hardware interrupt.
                      >
                      > A rate of 2.94Mhz is only half the full clock rate of the CPU. So at that speed, the CPU would not even get to fetch one instruction before being interrupted.
                      >
                      > Mike.
                      >
                      >
                      >
                      >
                      >
                      > ________________________________
                      > From: selloutdesigns <aah_public@...>
                      > To: korgpolyex@yahoogroups.com
                      > Sent: Wednesday, March 4, 2009 4:22:11 PM
                      > Subject: [korgpolyex] Re: Dead 800
                      >
                      >
                      > Hmm, now I am starting to think I'm chasing my tail. In the schematic (pg 6 in Service Manual) for the main board, it shows CPU RES 7.5 as running at .34usec per cycle, or about 2.94 Mhz, which is what got me thinking I had a problem there. However, in the description of the Interrupt Oscillator circuit in the Service Guide (pg 18) it reads:
                      >
                      > "This oscillator cycle is used for the EG, MG, LED displays, and S/H time division processing. Oscillator frequency is 24OOHz-360OHz. Interrupt order is by priority. If this circuit fails, EG operation and LED indication may become erratic."
                      >
                      > I measured the oscillator at about .34msec per cycle, or 2941Hz, which seems normal according to that.
                      >
                      > So which is correct? The schematic or the circuit description? I'm starting to think the circuit description. I've tested all the components except the TL072, but all seem fine.
                      >
                      > <<aaRoN>>
                      >
                      > 6) -Address Decoder:
                      >
                      > --- In korgpolyex@yahoogro ups.com, "narfman96" <narfman96@ ..> wrote:
                      > >
                      > > Should work fine. Install a good socket if you want. Then you can use a TL072 when you get one.
                      > >
                      > > --- In korgpolyex@yahoogro ups.com, "selloutdesigns" <aah_public@ > wrote:
                      > > >
                      > > > Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible.
                      > > >
                      > > > Thanks,
                      > > >
                      > > > <<aaRoN>>
                      > > >
                      > > > --- In korgpolyex@yahoogro ups.com, Gordon JC Pearce MM3YEQ <gordon@> wrote:
                      > > > >
                      > > > > On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
                      > > > > > Definitely timing issues with IC36b the interrupt oscillator. Try
                      > > > > > disconnecting the output at R107 and see if it runs at the correct
                      > > > > > frequency. If it does the CPU is loading it down somehow. If not then
                      > > > > > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
                      > > > > > verified and then the Tl072 op amp itself. This circuit needs to run
                      > > > > > in the mhz or no go for the LED display and envelope generators.
                      > > > >
                      > > > > Bah. Just change the TL072 on spec. Horrible things.
                      > > > >
                      > > > > Gordon
                      > > > >
                      > > >
                      > >
                      >
                    • Michael Hawkins
                      Well, you might be giving up a little too early don t you think? Remind me again but didn t you say it looked like it had some battery leak on it? If that is
                      Message 10 of 21 , Mar 6, 2009
                      • 0 Attachment
                        Well, you might be giving up a little too early don't you think?

                        Remind me again but didn't you say it looked like it had some battery leak on it?

                        If that is the case, then I would take a long, hard look at those spots that you think had those leaks on them.

                        Because that acid can do it's corrosive thing without showing much damage after you clean it up.

                        There is an EPROM in the EPROM socket right? ;-)

                        Mike.

                        From: selloutdesigns <aah_public@...>
                        To: korgpolyex@yahoogroups.com
                        Sent: Friday, March 6, 2009 9:53:52 AM
                        Subject: [korgpolyex] Re: Dead 800

                        Okay, I'm totally stumped at this point. The main issue I see now with the display is that the IO/M pin on the CPU is low 100% of the time, as confirmed with a scope reading. However, I have no idea why. SOD is also low all the time as well, which is probably related. The CPU clock is fine and I see RD,WR,S0,S1 going high and low on a regular basis so it seems memory reads and writes are occurring fine, but no I/O whatsoever. I'm wondering if there is an issue with the CPU receiving instructions from memory, which is causing the I/O cycle to never occur.

                        In any case, I have just about given up in trying to figure out what's going on. I think this unit is destined to land on someone else's bench. If it does get fixed I'll be sure to find out what the cause was.

                        Thanks to all for the input.

                        <<aaRoN>>

                        --- In korgpolyex@yahoogro ups.com, Michael Hawkins <korgpolyex800@ ...> wrote:
                        >
                        > The circuit description is correct.
                        >
                        > Interrupt RES7.5 runs at around 2400 to 3600 hz.
                        >
                        > It can't run too fast otherwise the CPU won't do anything else except service this hardware interrupt.
                        >
                        > A rate of 2.94Mhz is only half the full clock rate of the CPU. So at that speed, the CPU would not even get to fetch one instruction before being interrupted.
                        >
                        > Mike.
                        >
                        >
                        >
                        >
                        >
                        > ____________ _________ _________ __
                        > From: selloutdesigns <aah_public@ ...>
                        > To: korgpolyex@yahoogro ups.com
                        > Sent: Wednesday, March 4, 2009 4:22:11 PM
                        > Subject: [korgpolyex] Re: Dead 800
                        >
                        >
                        > Hmm, now I am starting to think I'm chasing my tail. In the schematic (pg 6 in Service Manual) for the main board, it shows CPU RES 7.5 as running at .34usec per cycle, or about 2.94 Mhz, which is what got me thinking I had a problem there. However, in the description of the Interrupt Oscillator circuit in the Service Guide (pg 18) it reads:
                        >
                        > "This oscillator cycle is used for the EG, MG, LED displays, and S/H time division processing. Oscillator frequency is 24OOHz-360OHz. Interrupt order is by priority. If this circuit fails, EG operation and LED indication may become erratic."
                        >
                        > I measured the oscillator at about .34msec per cycle, or 2941Hz, which seems normal according to that.
                        >
                        > So which is correct? The schematic or the circuit description? I'm starting to think the circuit description. I've tested all the components except the TL072, but all seem fine.
                        >
                        > <<aaRoN>>
                        >
                        > 6) -Address Decoder:
                        >
                        > --- In korgpolyex@yahoogro ups.com, "narfman96" <narfman96@ ..> wrote:
                        > >
                        > > Should work fine. Install a good socket if you want. Then you can use a TL072 when you get one.
                        > >
                        > > --- In korgpolyex@yahoogro ups.com, "selloutdesigns" <aah_public@ > wrote:
                        > > >
                        > > > Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible.
                        > > >
                        > > > Thanks,
                        > > >
                        > > > <<aaRoN>>
                        > > >
                        > > > --- In korgpolyex@yahoogro ups.com, Gordon JC Pearce MM3YEQ <gordon@> wrote:
                        > > > >
                        > > > > On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
                        > > > > > Definitely timing issues with IC36b the interrupt oscillator. Try
                        > > > > > disconnecting the output at R107 and see if it runs at the correct
                        > > > > > frequency. If it does the CPU is loading it down somehow. If not then
                        > > > > > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
                        > > > > > verified and then the Tl072 op amp itself. This circuit needs to run
                        > > > > > in the mhz or no go for the LED display and envelope generators.
                        > > > >
                        > > > > Bah. Just change the TL072 on spec. Horrible things.
                        > > > >
                        > > > > Gordon
                        > > > >
                        > > >
                        > >
                        >


                      • selloutdesigns
                        I haven t given up on the Poly, just my own ability to get it working in a timely fashion. Possibly if I spent a lot more time I would eventually stumble upon
                        Message 11 of 21 , Mar 6, 2009
                        • 0 Attachment
                          I haven't given up on the Poly, just my own ability to get it working in a timely fashion. Possibly if I spent a lot more time I would eventually stumble upon the fix, but fortunately I have a local resource that does vintage synth repair who is taking a look at it now. I'm hoping to have the diagnosis soon, and hopefully a functioning unit. Then I can get back to the important business of modding!

                          <<aaRoN>>

                          --- In korgpolyex@yahoogroups.com, Michael Hawkins <korgpolyex800@...> wrote:
                          >
                          > Well, you might be giving up a little too early don't you think?
                          >
                          > Remind me again but didn't you say it looked like it had some battery leak on it?
                          >
                          > If that is the case, then I would take a long, hard look at those spots that you think had those leaks on them.
                          >
                          > Because that acid can do it's corrosive thing without showing much damage after you clean it up.
                          >
                          > There is an EPROM in the EPROM socket right? ;-)
                          >
                          > Mike.
                          >
                          >
                          > ________________________________
                          > From: selloutdesigns <aah_public@...>
                          > To: korgpolyex@yahoogroups.com
                          > Sent: Friday, March 6, 2009 9:53:52 AM
                          > Subject: [korgpolyex] Re: Dead 800
                          >
                          >
                          > Okay, I'm totally stumped at this point. The main issue I see now with the display is that the IO/M pin on the CPU is low 100% of the time, as confirmed with a scope reading. However, I have no idea why. SOD is also low all the time as well, which is probably related. The CPU clock is fine and I see RD,WR,S0,S1 going high and low on a regular basis so it seems memory reads and writes are occurring fine, but no I/O whatsoever. I'm wondering if there is an issue with the CPU receiving instructions from memory, which is causing the I/O cycle to never occur.
                          >
                          > In any case, I have just about given up in trying to figure out what's going on. I think this unit is destined to land on someone else's bench. If it does get fixed I'll be sure to find out what the cause was.
                          >
                          > Thanks to all for the input.
                          >
                          > <<aaRoN>>
                          >
                          > --- In korgpolyex@yahoogro ups.com, Michael Hawkins <korgpolyex800@ ...> wrote:
                          > >
                          > > The circuit description is correct.
                          > >
                          > > Interrupt RES7.5 runs at around 2400 to 3600 hz.
                          > >
                          > > It can't run too fast otherwise the CPU won't do anything else except service this hardware interrupt.
                          > >
                          > > A rate of 2.94Mhz is only half the full clock rate of the CPU. So at that speed, the CPU would not even get to fetch one instruction before being interrupted.
                          > >
                          > > Mike.
                          > >
                          > >
                          > >
                          > >
                          > >
                          > > ____________ _________ _________ __
                          > > From: selloutdesigns <aah_public@ ...>
                          > > To: korgpolyex@yahoogro ups.com
                          > > Sent: Wednesday, March 4, 2009 4:22:11 PM
                          > > Subject: [korgpolyex] Re: Dead 800
                          > >
                          > >
                          > > Hmm, now I am starting to think I'm chasing my tail. In the schematic (pg 6 in Service Manual) for the main board, it shows CPU RES 7.5 as running at .34usec per cycle, or about 2.94 Mhz, which is what got me thinking I had a problem there. However, in the description of the Interrupt Oscillator circuit in the Service Guide (pg 18) it reads:
                          > >
                          > > "This oscillator cycle is used for the EG, MG, LED displays, and S/H time division processing. Oscillator frequency is 24OOHz-360OHz. Interrupt order is by priority. If this circuit fails, EG operation and LED indication may become erratic."
                          > >
                          > > I measured the oscillator at about .34msec per cycle, or 2941Hz, which seems normal according to that.
                          > >
                          > > So which is correct? The schematic or the circuit description? I'm starting to think the circuit description. I've tested all the components except the TL072, but all seem fine.
                          > >
                          > > <<aaRoN>>
                          > >
                          > > 6) -Address Decoder:
                          > >
                          > > --- In korgpolyex@yahoogro ups.com, "narfman96" <narfman96@ ..> wrote:
                          > > >
                          > > > Should work fine. Install a good socket if you want. Then you can use a TL072 when you get one.
                          > > >
                          > > > --- In korgpolyex@yahoogro ups.com, "selloutdesigns" <aah_public@ > wrote:
                          > > > >
                          > > > > Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible.
                          > > > >
                          > > > > Thanks,
                          > > > >
                          > > > > <<aaRoN>>
                          > > > >
                          > > > > --- In korgpolyex@yahoogro ups.com, Gordon JC Pearce MM3YEQ <gordon@> wrote:
                          > > > > >
                          > > > > > On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
                          > > > > > > Definitely timing issues with IC36b the interrupt oscillator. Try
                          > > > > > > disconnecting the output at R107 and see if it runs at the correct
                          > > > > > > frequency. If it does the CPU is loading it down somehow. If not then
                          > > > > > > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
                          > > > > > > verified and then the Tl072 op amp itself. This circuit needs to run
                          > > > > > > in the mhz or no go for the LED display and envelope generators.
                          > > > > >
                          > > > > > Bah. Just change the TL072 on spec. Horrible things.
                          > > > > >
                          > > > > > Gordon
                          > > > > >
                          > > > >
                          > > >
                          > >
                          >
                        • selloutdesigns
                          So I got my previously dead Poly back last night in fully working order! It turns out the CPU (IC24 80085)had failed somehow and it had shorted and burned out
                          Message 12 of 21 , Mar 26, 2009
                          • 0 Attachment
                            So I got my previously dead Poly back last night in fully working order! It turns out the CPU (IC24 80085)had failed somehow and it had shorted and burned out one of the D Flip Flops connected to it (IC 28). Replacing just the flip flop did not work and just immediately burned out the new chip. However, with both of those components replaced everything is back up and running.

                            <<aaRoN>>

                            --- In korgpolyex@yahoogroups.com, "selloutdesigns" <aah_public@...> wrote:
                            >
                            > I haven't given up on the Poly, just my own ability to get it working in a timely fashion. Possibly if I spent a lot more time I would eventually stumble upon the fix, but fortunately I have a local resource that does vintage synth repair who is taking a look at it now. I'm hoping to have the diagnosis soon, and hopefully a functioning unit. Then I can get back to the important business of modding!
                            >
                            > <<aaRoN>>
                            >
                            > --- In korgpolyex@yahoogroups.com, Michael Hawkins <korgpolyex800@> wrote:
                            > >
                            > > Well, you might be giving up a little too early don't you think?
                            > >
                            > > Remind me again but didn't you say it looked like it had some battery leak on it?
                            > >
                            > > If that is the case, then I would take a long, hard look at those spots that you think had those leaks on them.
                            > >
                            > > Because that acid can do it's corrosive thing without showing much damage after you clean it up.
                            > >
                            > > There is an EPROM in the EPROM socket right? ;-)
                            > >
                            > > Mike.
                            > >
                            > >
                            > > ________________________________
                            > > From: selloutdesigns <aah_public@>
                            > > To: korgpolyex@yahoogroups.com
                            > > Sent: Friday, March 6, 2009 9:53:52 AM
                            > > Subject: [korgpolyex] Re: Dead 800
                            > >
                            > >
                            > > Okay, I'm totally stumped at this point. The main issue I see now with the display is that the IO/M pin on the CPU is low 100% of the time, as confirmed with a scope reading. However, I have no idea why. SOD is also low all the time as well, which is probably related. The CPU clock is fine and I see RD,WR,S0,S1 going high and low on a regular basis so it seems memory reads and writes are occurring fine, but no I/O whatsoever. I'm wondering if there is an issue with the CPU receiving instructions from memory, which is causing the I/O cycle to never occur.
                            > >
                            > > In any case, I have just about given up in trying to figure out what's going on. I think this unit is destined to land on someone else's bench. If it does get fixed I'll be sure to find out what the cause was.
                            > >
                            > > Thanks to all for the input.
                            > >
                            > > <<aaRoN>>
                            > >
                            > > --- In korgpolyex@yahoogro ups.com, Michael Hawkins <korgpolyex800@ ...> wrote:
                            > > >
                            > > > The circuit description is correct.
                            > > >
                            > > > Interrupt RES7.5 runs at around 2400 to 3600 hz.
                            > > >
                            > > > It can't run too fast otherwise the CPU won't do anything else except service this hardware interrupt.
                            > > >
                            > > > A rate of 2.94Mhz is only half the full clock rate of the CPU. So at that speed, the CPU would not even get to fetch one instruction before being interrupted.
                            > > >
                            > > > Mike.
                            > > >
                            > > >
                            > > >
                            > > >
                            > > >
                            > > > ____________ _________ _________ __
                            > > > From: selloutdesigns <aah_public@ ...>
                            > > > To: korgpolyex@yahoogro ups.com
                            > > > Sent: Wednesday, March 4, 2009 4:22:11 PM
                            > > > Subject: [korgpolyex] Re: Dead 800
                            > > >
                            > > >
                            > > > Hmm, now I am starting to think I'm chasing my tail. In the schematic (pg 6 in Service Manual) for the main board, it shows CPU RES 7.5 as running at .34usec per cycle, or about 2.94 Mhz, which is what got me thinking I had a problem there. However, in the description of the Interrupt Oscillator circuit in the Service Guide (pg 18) it reads:
                            > > >
                            > > > "This oscillator cycle is used for the EG, MG, LED displays, and S/H time division processing. Oscillator frequency is 24OOHz-360OHz. Interrupt order is by priority. If this circuit fails, EG operation and LED indication may become erratic."
                            > > >
                            > > > I measured the oscillator at about .34msec per cycle, or 2941Hz, which seems normal according to that.
                            > > >
                            > > > So which is correct? The schematic or the circuit description? I'm starting to think the circuit description. I've tested all the components except the TL072, but all seem fine.
                            > > >
                            > > > <<aaRoN>>
                            > > >
                            > > > 6) -Address Decoder:
                            > > >
                            > > > --- In korgpolyex@yahoogro ups.com, "narfman96" <narfman96@ ..> wrote:
                            > > > >
                            > > > > Should work fine. Install a good socket if you want. Then you can use a TL072 when you get one.
                            > > > >
                            > > > > --- In korgpolyex@yahoogro ups.com, "selloutdesigns" <aah_public@ > wrote:
                            > > > > >
                            > > > > > Anyone know offhand if a tl082 can safely replace a tl072? They do appear compatible.
                            > > > > >
                            > > > > > Thanks,
                            > > > > >
                            > > > > > <<aaRoN>>
                            > > > > >
                            > > > > > --- In korgpolyex@yahoogro ups.com, Gordon JC Pearce MM3YEQ <gordon@> wrote:
                            > > > > > >
                            > > > > > > On Wed, 2009-03-04 at 13:34 +0000, narfman96 wrote:
                            > > > > > > > Definitely timing issues with IC36b the interrupt oscillator. Try
                            > > > > > > > disconnecting the output at R107 and see if it runs at the correct
                            > > > > > > > frequency. If it does the CPU is loading it down somehow. If not then
                            > > > > > > > suspect the timing caps C76 or C79. The resistors R102-R104 need to be
                            > > > > > > > verified and then the Tl072 op amp itself. This circuit needs to run
                            > > > > > > > in the mhz or no go for the LED display and envelope generators.
                            > > > > > >
                            > > > > > > Bah. Just change the TL072 on spec. Horrible things.
                            > > > > > >
                            > > > > > > Gordon
                            > > > > > >
                            > > > > >
                            > > > >
                            > > >
                            > >
                            >
                          Your message has been successfully submitted and would be delivered to recipients shortly.