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

RE: [midatlanticretro] FD 1771 - Floppy Drive Controller TAX?

Expand Messages
  • Bill Sudbrink
    Hi, Have you looked at the manual scan on Harte s site: http://www.hartetechnologies.com/manuals/Tarbell/Tarbell%20Single%20Density% 20Disk%20Controller.pdf It
    Message 1 of 29 , Apr 3, 2012
      Hi,

      Have you looked at the manual scan on Harte's site:


      http://www.hartetechnologies.com/manuals/Tarbell/Tarbell%20Single%20Density%
      20Disk%20Controller.pdf

      It has a lot of notes and addenda appended to it.

      Bill S.
    • Bill Sudbrink
      Some things I would check: Make sure FD1771 pin 25 is truly grounded. It is socketed, right? Did you reseat the 1771? Likewise, make sure you have a good
      Message 2 of 29 , Apr 3, 2012
        Some things I would check:

        Make sure FD1771 pin 25 is truly grounded. It is
        socketed, right? Did you reseat the 1771?

        Likewise, make sure you have a good clock signal on
        1771 pin 24.

        What does the 74ls175 (u33) look like? Is it socketed
        or soldered? Are its pins tarnished? Does the clock
        on pin 4 of the 175 look good?

        Basically, if any of the chips on the board are TI,
        socketed and have pins tarnished to black or almost
        black, it would probably be a good idea to gently lift
        and reseat them.

        Are the gate voltages on the 74ls08 (u36) good? If they
        are marginal, you might be getting noise in the signal.

        Bill S.
      • joshbensadon
        Hi Bill, Thanks for your ideas. ... Yes, it s 100% grounded. I measured less than .05 Ohms to ground on this pin. I ve reseated, and replaced the 1771 with 2
        Message 3 of 29 , Apr 3, 2012
          Hi Bill,

          Thanks for your ideas.

          > Make sure FD1771 pin 25 is truly grounded. It is
          > socketed, right? Did you reseat the 1771?

          Yes, it's 100% grounded. I measured less than .05 Ohms to ground on this pin.

          I've reseated, and replaced the 1771 with 2 other chips. Same results every time.

          > Likewise, make sure you have a good clock signal on
          > 1771 pin 24.

          Clock is good, 4v p-p, the low is about .2V there is a little ringing but I see that all the time (probably my probe), it's nothing severe.

          > What does the 74ls175 (u33) look like? Is it socketed
          > or soldered? Are its pins tarnished? Does the clock
          > on pin 4 of the 175 look good?

          On a socket, pins were tarnished, but I cleaned them with an eraser many steps ago. Data on pin 4 looks good, Clock on pin 9 looks a little slanted, I mean it has a slow rise/drop time. I would say 10% rise time, 10% drop, 80% at the high/low levels. 4V high, .2V low.

          > Basically, if any of the chips on the board are TI,
          > socketed and have pins tarnished to black or almost
          > black, it would probably be a good idea to gently lift
          > and reseat them.

          Great idea, did this on all chips a couple of days ago.

          > Are the gate voltages on the 74ls08 (u36) good? If they
          > are marginal, you might be getting noise in the signal.

          Yes, they are .2V low, 4V high coming in. 0V/4.75V going out because I replaced this chip with a 74HCT08. The 74LS08 was putting out the standard .2V / 4V levels.
          But thank you, you gave me another idea... add more decoupling capacitors! And double check the voltages around the 1771. The +12, +5 and -5 are all present and good.

          Here's the funny part, I tried grounding the DATA input, this chip is getting NO DATA, yet out comes a bunch of FF's ????
          The only time it gives something other than FF's is when the input pattern is AA's and the data seperator reverses the clock and data pulses. If 2 pulses come in on the data line then 1 pulse on the clock line, the chip reads all AA's (or 55's) to RAM.

          The routine I'm using to read the chip is right out of the Manual you described, yes, it's a very good manual, saved me days of work already. I've looked at that code 5 times already. I'm even scoping the data lines at strobe time and I see the FF's (inverted as 00's) on the 1771 data outputs. My latest pattern input is C0's, 2 data pulses appear very neat and sharp for every 8 clock's, but out comes FF!

          I'm running out of ideas.... next I think I'm going to replace all the chips around the Read/Write of this chip.

          Also, I found C7 missing from the 4Mhz clock, but the parts list doesn't list C7. I think it's only used for certain crystals.

          sigh.

          I am begining to think we should stop using floppy disks... not so reliable!
        • Bill Sudbrink
          Ok, here are a couple of weird things that have bit me in the past: Cracked pins. I have had TI pins crack in such a way that they looked good in place,
          Message 4 of 29 , Apr 4, 2012
            Ok, here are a couple of weird things that have "bit" me
            in the past:

            Cracked pins. I have had TI pins crack in such a way that
            they looked good in place, would test good with a probe but
            were bad during a run. I've seen two kinds of cracks.
            Warning, bad ASCII art follows:


            | | <-1
            | |
            \ /
            \ /
            || <-2
            ||
            \/

            I have seen cracks at position 1, right where the pin makes
            the bend to go into the chip body. Straight across the bend
            on a soldered chip. Almost invisible to the naked eye.
            Pressure from a probe on top of the pin closed the crack and
            made the pin look good under test.

            I have seen cracks at position 2, at the bottom of the taper
            where the pin narrows to its insertion width. I have seen this
            in both soldered and socketed chips. In the socketed case, I
            stupidly didn't notice the missing piece of pin through several
            removal/insertions. Under test, it can behave just as above,
            with the pressure on the pin closing the circuit.

            On other thing I have seen exactly once was a case where the
            pin from the socket was crumpled under the socket. This board
            was probably machine assembled. The crumpled pin was making
            intermittent contact with the solder pad that it had been
            intended to be soldered to. It was one of those tall, solid
            body sockets making the problem very difficult to see.

            Bill S.
          • Evan Koblentz
            ... That might be the post of the year.
            Message 5 of 29 , Apr 4, 2012
              >> I've seen two kinds of cracks. Warning, bad ASCII art follows

              That might be the post of the year.
            • joshbensadon
              Hi Bill, Thanks, that s good solid advice. I found a #2 break on one of the front panel chips. Always good to check the pins of the sockets too. I ll give
              Message 6 of 29 , Apr 4, 2012
                Hi Bill,

                Thanks, that's good solid advice. I found a #2 break on one of the front panel chips. Always good to check the pins of the sockets too. I'll give the board a good visual before moving the problem to a bread board (what would be my next step without your input).

                I am also curious about the -5V. I saw some chips don't require it. But I think I have chips that do.

                While fixing an Ohio Scientific Superboard, the extra pressure I applied to the pin when scoping it had it make contact to the socket. I caught on to this pretty quickly but it still tricked me at first.

                Again, my biggest stumbling block on my 1771 problem is how it produces valid numbers when there are 2 data pulses for every clock pulse. This leads me to believe the CPU decoding/interfacing is correct, along with my code. I don't believe the data sheet and circuit could be wrong either. So there's something here that I've never seen before... but with a little patience and persistence I might have a good story to tell!

                Cheers,
                Josh


                --- In midatlanticretro@yahoogroups.com, "Bill Sudbrink" <wh.sudbrink@...> wrote:
                >
                > Ok, here are a couple of weird things that have "bit" me
                > in the past:
                >
                > Cracked pins. I have had TI pins crack in such a way that
                > they looked good in place, would test good with a probe but
                > were bad during a run. I've seen two kinds of cracks.
                > Warning, bad ASCII art follows:
                >
                >
                > | | <-1
                > | |
                > \ /
                > \ /
                > || <-2
                > ||
                > \/
                >
                > I have seen cracks at position 1, right where the pin makes
                > the bend to go into the chip body. Straight across the bend
                > on a soldered chip. Almost invisible to the naked eye.
                > Pressure from a probe on top of the pin closed the crack and
                > made the pin look good under test.
                >
                > I have seen cracks at position 2, at the bottom of the taper
                > where the pin narrows to its insertion width. I have seen this
                > in both soldered and socketed chips. In the socketed case, I
                > stupidly didn't notice the missing piece of pin through several
                > removal/insertions. Under test, it can behave just as above,
                > with the pressure on the pin closing the circuit.
                >
                > On other thing I have seen exactly once was a case where the
                > pin from the socket was crumpled under the socket. This board
                > was probably machine assembled. The crumpled pin was making
                > intermittent contact with the solder pad that it had been
                > intended to be soldered to. It was one of those tall, solid
                > body sockets making the problem very difficult to see.
                >
                > Bill S.
                >
              • joshbensadon
                Yeah! Thanks for the Bad Ascii art warning! lol.
                Message 7 of 29 , Apr 4, 2012
                  Yeah! Thanks for the Bad Ascii art warning! lol.



                  --- In midatlanticretro@yahoogroups.com, "Evan Koblentz" <evan@...> wrote:
                  >
                  > >> I've seen two kinds of cracks. Warning, bad ASCII art follows
                  >
                  > That might be the post of the year.
                  >
                • Bill Sudbrink
                  ... Actually, that s old usenet code for be sure your viewer is set to a fixed width font or this won t make any sense . Bill S.
                  Message 8 of 29 , Apr 4, 2012
                    Joshbensadon wrote:
                    >
                    > Yeah! Thanks for the Bad Ascii art warning! lol.
                    >

                    Actually, that's old usenet code for "be sure your viewer is
                    set to a fixed width font or this won't make any sense".

                    Bill S.
                  • joshbensadon
                    ... I never joined usenet. How long ago was usenet popular? Is it still around? I recall the days of BBS s and all the cool ANSI art. There wasn t a need to
                    Message 9 of 29 , Apr 4, 2012
                      > Actually, that's old usenet code for "be sure your viewer is
                      > set to a fixed width font or this won't make any sense".
                      >
                      > Bill S.
                      >

                      I never joined usenet. How long ago was usenet popular? Is it still around? I recall the days of BBS's and all the cool ANSI art. There wasn't a need to worry about fixed width then, it was all 80 column.
                    • Dave McGuire
                      ... It s still around, and it s still active. I was shocked out of my gourd to find this out about a year ago. I thought it was completely dead. I was (among
                      Message 10 of 29 , Apr 4, 2012
                        On 04/04/2012 07:22 PM, joshbensadon wrote:
                        > I never joined usenet. How long ago was usenet popular? Is it still
                        > around?

                        It's still around, and it's still active. I was shocked out of my
                        gourd to find this out about a year ago. I thought it was completely dead.

                        I was (among other things) a news admin at a large ISP for a long
                        time. We bought one of the first SGI Origin 2000 systems to build a
                        news server; that was fun.

                        When I learned that it was still active, I built a news server here
                        and finagled my way into getting a news feed. I get all the major
                        hierarchies but no binaries. (I don't care about the binaries) The
                        technical groups, in particular, are wonderful. My news server receives
                        between 30,000 and 40,000 articles per day. (remember, that's not
                        including binaries!) Probably 1/3 of that is spam, but the rest is a
                        lot of good meat.

                        -Dave

                        --
                        Dave McGuire, AK4HZ
                        New Kensington, PA
                      • Dave McGuire
                        ... I ve seen this failure mode as well. In one case, microscopic examination revealed that the only thing holding the pin together was the thin layer of
                        Message 11 of 29 , Apr 4, 2012
                          On 04/04/2012 10:12 AM, Bill Sudbrink wrote:
                          > Ok, here are a couple of weird things that have "bit" me
                          > in the past:
                          >
                          > Cracked pins. I have had TI pins crack in such a way that
                          > they looked good in place, would test good with a probe but
                          > were bad during a run. I've seen two kinds of cracks.
                          > Warning, bad ASCII art follows:
                          >
                          >
                          > | | <-1
                          > | |
                          > \ /
                          > \ /
                          > || <-2
                          > ||
                          > \/
                          >
                          > I have seen cracks at position 1, right where the pin makes
                          > the bend to go into the chip body. Straight across the bend
                          > on a soldered chip. Almost invisible to the naked eye.
                          > Pressure from a probe on top of the pin closed the crack and
                          > made the pin look good under test.
                          >
                          > I have seen cracks at position 2, at the bottom of the taper
                          > where the pin narrows to its insertion width. I have seen this
                          > in both soldered and socketed chips. In the socketed case, I
                          > stupidly didn't notice the missing piece of pin through several
                          > removal/insertions. Under test, it can behave just as above,
                          > with the pressure on the pin closing the circuit.

                          I've seen this failure mode as well. In one case, microscopic
                          examination revealed that the only thing holding the pin together was
                          the thin layer of oxide. That was a serious pain to track down.

                          -Dave


                          --
                          Dave McGuire, AK4HZ
                          New Kensington, PA
                        • joshbensadon
                          I ve been struggling with this Tarbell 1011 controller using the FD 1771 controller for a couple of weeks. I kept getting FF s out while I was watching good
                          Message 12 of 29 , Apr 6, 2012
                            I've been struggling with this Tarbell 1011 controller using the FD 1771 controller for a couple of weeks.

                            I kept getting FF's out while I was watching good data go in.
                            The only time I got any other data, was when the (external) Clock/Data seperator swapped the clock & data.

                            I looked everywhere for the problem. I even made my own Clock/Data test pattern generator to give clean pulses. Once again, only when I reverse Clock and Data does it seem to work.

                            So, I swaped the Clock and Data. Curiously, the circuit now works!

                            The first ID field looks like this...
                            FE 00 00 01 00 D2 C3

                            I am assuming the CRC (D2C3) is correct.

                            Does anyone have any ERRATA on the FD 1771 Floppy Drive controller?

                            I am curious if anyone else out there has any schematics or circuits using the FD 1771 with external data seperator and would you confirm Pin 27 for Data, Pin 26 for Clock (as per data sheet) or are these truely reversed?

                            PS. Thank you to everyone for all the help, support and invaluable idea's given. And sorry for the mess of hair on the floor :)
                          • B Degnan
                            Herb Johnson would know a lot about that stuff, but he is not usually checking this list. He can be contacted through his web site retrotechnology.com
                            Message 13 of 29 , Apr 6, 2012
                              Herb Johnson would know a lot about that stuff, but he is not usually
                              checking this list. He can be contacted through his web site
                              retrotechnology.com
                              http://www.retrotechnology.com/herbs_stuff/s100bus.html
                              Bill

                              At 10:47 AM 4/6/2012, you wrote:
                              >I've been struggling with this Tarbell 1011 controller using the FD
                              >1771 controller for a couple of weeks.
                              >
                              >I kept getting FF's out while I was watching good data go in.
                              >The only time I got any other data, was when the (external)
                              >Clock/Data seperator swapped the clock & data.
                              >
                              >I looked everywhere for the problem. I even made my own Clock/Data
                              >test pattern generator to give clean pulses. Once again, only when
                              >I reverse Clock and Data does it seem to work.
                              >
                              >So, I swaped the Clock and Data. Curiously, the circuit now works!
                              >
                              >The first ID field looks like this...
                              >FE 00 00 01 00 D2 C3
                              >
                              >I am assuming the CRC (D2C3) is correct.
                              >
                              >Does anyone have any ERRATA on the FD 1771 Floppy Drive controller?
                              >
                              >I am curious if anyone else out there has any schematics or circuits
                              >using the FD 1771 with external data seperator and would you confirm
                              >Pin 27 for Data, Pin 26 for Clock (as per data sheet) or are these
                              >truely reversed?
                              >
                              >PS. Thank you to everyone for all the help, support and invaluable
                              >idea's given. And sorry for the mess of hair on the floor :)
                              >
                              >
                              >
                              >
                              >
                              >------------------------------------
                              >
                              >Yahoo! Groups Links
                              >
                              >
                              >
                            Your message has been successfully submitted and would be delivered to recipients shortly.