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

What is "Runout"?

Expand Messages
  • josh6502
    Hi, I m reading up the Siemens Floppy Disk Drive FDD 100-8. (FDD100-8_V1.pdf). On Page 1-4, they say For minimum runout the floppy disk should be loaded while
    Message 1 of 11 , Mar 20, 2012
    • 0 Attachment
      Hi,

      I'm reading up the Siemens Floppy Disk Drive FDD 100-8.
      (FDD100-8_V1.pdf).

      On Page 1-4, they say "For minimum runout the floppy disk should be loaded while the spindle is turning."

      What is runout?

      My floppy drive is not seeking track 0 on reset. The controller is a WD1771, the chip is sending out the stepping pulses, the logic at the FDC is low for Drive Select and the stepping pulses are there coming out of the FDC. I don't care much about direction for now, but from what I've read, this should be enough to step the head. I've manually moved the head to the middle.

      I'm going to next check the signals reaching the drive, then follow them on the Drive board, but if anyone knows if I'm missing something obvious, please let me know.

      Thanks!
      Josh
    • Dave McGuire
      ... Runout is the amount of lateral displacement of the floppy with respect to the spindle. It s basically how far off-center the disk is when it s clamped
      Message 2 of 11 , Mar 20, 2012
      • 0 Attachment
        On 03/21/2012 01:41 AM, josh6502 wrote:
        > I'm reading up the Siemens Floppy Disk Drive FDD 100-8.
        > (FDD100-8_V1.pdf).
        >
        > On Page 1-4, they say "For minimum runout the floppy disk should be
        > loaded while the spindle is turning."
        >
        > What is runout?

        Runout is the amount of lateral displacement of the floppy with
        respect to the spindle. It's basically how far off-center the disk is
        when it's clamped onto the spindle.

        > My floppy drive is not seeking track 0 on reset. The controller is a
        > WD1771, the chip is sending out the stepping pulses, the logic at the
        > FDC is low for Drive Select and the stepping pulses are there coming
        > out of the FDC. I don't care much about direction for now, but from
        > what I've read, this should be enough to step the head. I've
        > manually moved the head to the middle.
        >
        > I'm going to next check the signals reaching the drive, then follow
        > them on the Drive board, but if anyone knows if I'm missing something
        > obvious, please let me know.

        Is the drive being selected? Is the stepper controller on the drive
        ok? Have you looked at its outputs with a scope?

        -Dave

        --
        Dave McGuire, AK4HZ
        New Kensington, PA
      • josh6502
        ... That makes sense. As I was thinking about it, what else could clamping the disk while the spindle was turning accomplish. But it s always good to ask
        Message 3 of 11 , Mar 21, 2012
        • 0 Attachment
          --- In midatlanticretro@yahoogroups.com, Dave McGuire <Mcguire@...> wrote:
          >
          > On 03/21/2012 01:41 AM, josh6502 wrote:
          > > I'm reading up the Siemens Floppy Disk Drive FDD 100-8.
          > > (FDD100-8_V1.pdf).
          > > On Page 1-4, they say "For minimum runout the floppy disk should be
          > > loaded while the spindle is turning."
          > > What is runout?
          >
          > Runout is the amount of lateral displacement of the floppy with
          > respect to the spindle. It's basically how far off-center the disk is
          > when it's clamped onto the spindle.

          That makes sense. As I was thinking about it, what else could clamping the disk while the spindle was turning accomplish. But it's always good to ask someone who knows more about this.


          > > My floppy drive is not seeking track 0 on reset. The controller is a
          > > WD1771, the chip is sending out the stepping pulses, the logic at the
          > > FDC is low for Drive Select and the stepping pulses are there coming
          > > out of the FDC. I don't care much about direction for now, but from
          > > what I've read, this should be enough to step the head. I've
          > > manually moved the head to the middle.
          > > I'm going to next check the signals reaching the drive, then follow
          > > them on the Drive board, but if anyone knows if I'm missing something
          > > obvious, please let me know.
          >
          > Is the drive being selected? Is the stepper controller on the drive
          > ok? Have you looked at its outputs with a scope?
          > -Dave

          I was actually just looking for some more basic steps...

          According to the FDC, the drive is selected. I came up with 5 more simple steps to perform before removing the drive from the chasis.
          Reseat connector(s) to Floppy disk -no change.
          Test 24V -OK.
          Disconnect drive 2 from cable -no change.
          Replace Floppy cable -Don't have a 2nd cable.
          Use drive 2 connector for drive 1 -no change.

          I unscrewed the drives, checked for drive being selected right on the jumper block of the drive -That's ok. Checked for the step pulses (using a logic probe at the moment), now there's no step pulses... traced that back to the fact that Track00 line is active (even though the head isn't at Track00).

          Decided to test the other drive... reconfigured it for drive 0, it gets selected and the step pulses but it's not stepping either!

          Time to haul it over to the work bench and scope out the circiuts.
          The stepper motor on Drive 0 was getting warm... I might have a shorted transistor (TIP31A's).

          I love this stuff, but it's just that sometimes I do stupid things and forget to check the basics before spending a lot of time looking into something that isn't broken. I am a little suspicious that neither drive would step... perhaps my pulse width's are too short?(I'll have to scope that later).
        • Dave McGuire
          ... TIP31As are pretty easy to find. Also the stepper controller (assuming your drive uses one, SAA1027 or similar?) might be toast. ... I did that all day
          Message 4 of 11 , Mar 24, 2012
          • 0 Attachment
            On 03/21/2012 10:34 AM, josh6502 wrote:
            > I unscrewed the drives, checked for drive being selected right on the
            > jumper block of the drive -That's ok. Checked for the step pulses
            > (using a logic probe at the moment), now there's no step pulses...
            > traced that back to the fact that Track00 line is active (even though
            > the head isn't at Track00).
            >
            > Decided to test the other drive... reconfigured it for drive 0, it
            > gets selected and the step pulses but it's not stepping either!
            >
            > Time to haul it over to the work bench and scope out the circiuts.
            > The stepper motor on Drive 0 was getting warm... I might have a
            > shorted transistor (TIP31A's).

            TIP31As are pretty easy to find. Also the stepper controller
            (assuming your drive uses one, SAA1027 or similar?) might be toast.

            > I love this stuff, but it's just that sometimes I do stupid things
            > and forget to check the basics before spending a lot of time looking
            > into something that isn't broken.

            I did that all day yesterday and all day today. I'm working on a new
            controller design; my processor has eight power and eight ground pins.
            I was getting dead air on the JTAG port with no other stuff populated on
            the (first prototype) PCB.

            I went over the design a dozen times, perhaps more. I dug through
            the (SEVEN HUNDRED PAGE!) manual for the chip. I went through the
            errata. I scoped the lines on a functioning JTAG interface using a
            similar, but not identical chip, and compared all the signaling.

            Nothing!

            I was totally at my wits' end, really getting stressed out because
            I'm under a big deadline for this one, when Dan walks up and asked, "hey
            did you check the Vdd/GND pins around the chip?"

            One of the Vdd pins wasn't getting 3.3V. It turns out that there was
            a minor imaging issue at the PCB fab, and where there was a tiny sliver
            of copper (my Vdd plane) on the gerbers, it wasn't on the board. I
            can't fault them for that...it was only about 5mils wide (I had missed
            that due to it being late and being in a rush) and was narrowed by my
            having moved a nearby trace just before finishing.

            The PCB layout software didn't catch it because it was connected from
            its point of view, and I had mistyped the DRC value as 5mil instead of
            6...otherwise it would've caught it.

            > I am a little suspicious that
            > neither drive would step... perhaps my pulse width's are too
            > short?(I'll have to scope that later).

            That's certainly possible. That's programmable in most controller
            chips. The WD177x has two bits in the "step" command byte that sets the
            step rate, for example. Do you know the step delay requirements for
            your drives?

            -Dave

            --
            Dave McGuire, AK4HZ
            New Kensington, PA
          • Dan Roganti
            ... yes, my work is never done, saving one s arse is a never ending proposition ;) The universe has returned back to a state of equilibrium as a result.
            Message 5 of 11 , Mar 24, 2012
            • 0 Attachment
              On 03/25/2012 01:39 AM, Dave McGuire wrote:
              >
              > I did that all day yesterday and all day today. I'm working on a new
              > controller design; my processor has eight power and eight ground pins.
              > I was getting dead air on the JTAG port with no other stuff populated on
              > the (first prototype) PCB.
              >
              > I went over the design a dozen times, perhaps more. I dug through
              > the (SEVEN HUNDRED PAGE!) manual for the chip. I went through the
              > errata. I scoped the lines on a functioning JTAG interface using a
              > similar, but not identical chip, and compared all the signaling.
              >
              > Nothing!
              >
              > I was totally at my wits' end, really getting stressed out because
              > I'm under a big deadline for this one, when Dan walks up and asked, "hey
              > did you check the Vdd/GND pins around the chip?"

              yes, my work is never done, saving one's arse is a never ending
              proposition ;)
              The universe has returned back to a state of equilibrium as a result.
            • Dave McGuire
              ... Indeed. ;) ... If that s true, then why does it smell funny in here? ;) -Dave -- Dave McGuire, AK4HZ New Kensington, PA
              Message 6 of 11 , Mar 24, 2012
              • 0 Attachment
                On 03/25/2012 01:51 AM, Dan Roganti wrote:
                >> I did that all day yesterday and all day today. I'm working on a new
                >> controller design; my processor has eight power and eight ground pins.
                >> I was getting dead air on the JTAG port with no other stuff populated on
                >> the (first prototype) PCB.
                >>
                >> I went over the design a dozen times, perhaps more. I dug through
                >> the (SEVEN HUNDRED PAGE!) manual for the chip. I went through the
                >> errata. I scoped the lines on a functioning JTAG interface using a
                >> similar, but not identical chip, and compared all the signaling.
                >>
                >> Nothing!
                >>
                >> I was totally at my wits' end, really getting stressed out because
                >> I'm under a big deadline for this one, when Dan walks up and asked, "hey
                >> did you check the Vdd/GND pins around the chip?"
                >
                > yes, my work is never done, saving one's arse is a never ending
                > proposition ;)

                Indeed. ;)

                > The universe has returned back to a state of equilibrium as a result.

                If that's true, then why does it smell funny in here? ;)

                -Dave

                --
                Dave McGuire, AK4HZ
                New Kensington, PA
              • josh6502
                ... What ever controller it has, it s built into an LSI chip. The TIP s are just wired with a 2N3904 in Darlington configuration. I just finished making the
                Message 7 of 11 , Mar 25, 2012
                • 0 Attachment
                  > > The stepper motor on Drive 0 was getting warm... I might have a
                  > > shorted transistor (TIP31A's).
                  >
                  > TIP31As are pretty easy to find. Also the stepper controller
                  > (assuming your drive uses one, SAA1027 or similar?) might be toast.

                  What ever controller it has, it's built into an LSI chip. The TIP's are just wired with a 2N3904 in Darlington configuration. I just finished making the schematic for the drive. I probably could trouble shoot without the schematic, but it's my pleasure to trace and create a schematic. I hope to get it working by today.


                  > mistyped the DRC value as 5mil instead of
                  > 6...otherwise it would've caught it.

                  6 mil traces??? I've seen electrons bigger than that!

                  Good going, and don't be hard on yourself, that's sometimes not such an easy problem to figure out. Oh sure, we can all laugh about it after the fact with the old saying "Did you turn it on?". I had a similar problem, but with a twist, my circuit would work under some conditions. Turned out, the chip was getting it's power through the internal clamping diodes. When most of the inputs were low, the circuit would stop... drove me nuts!

                  > The WD177x has two bits in the "step" command byte that sets the
                  > step rate, for example. Do you know the step delay requirements for
                  > your drives?
                  >
                  > -Dave


                  The WD1771 puts out a seek to track00 on reset. I saw the bits you are talking about, and the WD1771 gives a fixed width for the reset seek (20mSec). But, the Tarbell conroller negates this with a 1 shot. I've measured the pulse and it's 1mSec. The line levels (on the ribbon cable) are +3V/0V, but this must be enough because the correct 1mSec pulse is reaching the LSI chip on the drive. Well, that's as far as I am for the moment... hope to get back on it later today.
                • josh6502
                  Hi Guys, thanks for all being here, even as a sounding board. It s always great to throw ideas around like minded techies. It didn t take me long to find the
                  Message 8 of 11 , Mar 25, 2012
                  • 0 Attachment
                    Hi Guys, thanks for all being here, even as a sounding board. It's always great to throw ideas around like minded techies.

                    It didn't take me long to find the Opto Coupler for Track00 detect was blown. The LSI chip wouldn't seek to track 0 because it thought it was already there. The Write Protect sensor is also blown. I don't know why, but they used 130 ohm current limiting resistor in series with the LED. We all know it typically should be 330!

                    As it turns out, the Optocoupler for my 2nd drive is also blown!!!
                    Probably a design flaw?

                    I've only got 2 of them in my junk box... time to buy some!

                    Now, I replaced the sensor careful not to move the alignment bracket... but since it's another sensor, I'm almost certain track 00 will be off.

                    Where can I get an allignment disk? All I have are new 3M disks, and I am guessing they are not formated.

                    Cheers,
                    Josh
                  • Dave McGuire
                    ... Ah-HA! ... Hmm, dunno about that...that depends entirely on the LED s V(f) figure, which for an IR led (which is what most opto-interrupters use) might
                    Message 9 of 11 , Mar 25, 2012
                    • 0 Attachment
                      On 03/25/2012 03:10 PM, josh6502 wrote:
                      > Hi Guys, thanks for all being here, even as a sounding board. It's
                      > always great to throw ideas around like minded techies.

                      :-)

                      > It didn't take me long to find the Opto Coupler for Track00 detect
                      > was blown. The LSI chip wouldn't seek to track 0 because it thought
                      > it was already there. The Write Protect sensor is also blown.

                      Ah-HA!

                      > I
                      > don't know why, but they used 130 ohm current limiting resistor in
                      > series with the LED. We all know it typically should be 330!

                      Hmm, dunno about that...that depends entirely on the LED's V(f)
                      figure, which for an IR led (which is what most opto-interrupters use)
                      might very well be higher than your standard red LED. (~1.2V) Just sayin'..

                      > As it turns out, the Optocoupler for my 2nd drive is also blown!!!
                      > Probably a design flaw?

                      Could be. It might have been a tradeoff between light intensity (and
                      thus resilience to the caked dust that'll inevitably show up in that
                      environment) and longevity.

                      > I've only got 2 of them in my junk box... time to buy some!
                      >
                      > Now, I replaced the sensor careful not to move the alignment
                      > bracket... but since it's another sensor, I'm almost certain track 00
                      > will be off.

                      Ahh. Poop.

                      > Where can I get an allignment disk? All I have are new 3M disks, and
                      > I am guessing they are not formated.

                      This is an 8" drive, yes? I have an 8" alignment disk that I can
                      send to you as long as you SWEAR to be very careful with it. At the
                      moment I'm not 100% certain of where it is (I'm still in the process of
                      moving) but I will find it. Where are you located?

                      -Dave

                      --
                      Dave McGuire, AK4HZ
                      New Kensington, PA
                    • josh6502
                      ... I suspect you are right. This machine was in production for many years, so it must have worked then. ... Thanks for your offer Dave, it s very much
                      Message 10 of 11 , Mar 26, 2012
                      • 0 Attachment
                        > > I
                        > > don't know why, but they used 130 ohm current limiting resistor in
                        > > series with the LED. We all know it typically should be 330!
                        >
                        > Hmm, dunno about that...that depends entirely on the LED's V(f)
                        > figure, which for an IR led (which is what most opto-interrupters use)
                        > might very well be higher than your standard red LED. (~1.2V) Just sayin'..

                        I suspect you are right. This machine was in production for many years, so it must have worked then.


                        > > I'm almost certain track 00 will be off.
                        >
                        > Ahh. Poop.
                        >
                        > > Where can I get an allignment disk? All I have are new 3M disks, and
                        > > I am guessing they are not formated.
                        >
                        > This is an 8" drive, yes? I have an 8" alignment disk that I can
                        > send to you as long as you SWEAR to be very careful with it. At the
                        > moment I'm not 100% certain of where it is (I'm still in the process of
                        > moving) but I will find it. Where are you located?
                        >
                        > -Dave


                        Thanks for your offer Dave, it's very much apreciated. However, I don't really think it's necessary. I forgot to consider that the LSI chip only uses this sensor to know if it's near track 00. It won't register as track 00 until stepper motor is in phase A. It's a 3 phase stepper motor. I did not touch the mechanics of the worm gear, so I don't think it will be out of alignment. Ultimately, I don't really care how the alignment is, as long as both drives read the same disks. I don't know of another 8" drive within humdreds of miles of where I live. I don't think I'll need to swap software etc.
                        PS. I live in Toronto. It would not be smart to mail your alignment disk anywhere... what if it got demagnetized by some scanner? It would be better if the drive(s) were shipped to you and you did the work (for a nominal fee of course).

                        Good luck on your move, I hope it's to a bigger house with lots of space for all your interests.

                        Cheers,
                        Josh
                      • Dave McGuire
                        ... Most of my boards are laid out on 6x6 rules, 6mil trace width and 6mil spacing. The processors I usually work with (in this case a Philips...erm, NXP
                        Message 11 of 11 , Mar 29, 2012
                        • 0 Attachment
                          On 03/25/2012 09:27 AM, josh6502 wrote:
                          >> mistyped the DRC value as 5mil instead of 6...otherwise it would've
                          >> caught it.
                          >
                          > 6 mil traces??? I've seen electrons bigger than that!

                          Most of my boards are laid out on 6x6 rules, 6mil trace width and
                          6mil spacing. The processors I usually work with (in this case a
                          Philips...erm, "NXP" LPC2387 ARM7 MCU) are typically in LQFP packages
                          with 0.5mm spacing. That's about as tight as I like to go, but I do
                          have a leadless chip (QFN package) with 0.4mm pad spacing that I'll need
                          to solder soon. I'll get a good night's sleep before that has to go on
                          the board.

                          > Good going, and don't be hard on yourself, that's sometimes not such
                          > an easy problem to figure out. Oh sure, we can all laugh about it
                          > after the fact with the old saying "Did you turn it on?". I had a
                          > similar problem, but with a twist, my circuit would work under some
                          > conditions. Turned out, the chip was getting it's power through the
                          > internal clamping diodes. When most of the inputs were low, the
                          > circuit would stop... drove me nuts!

                          Ugh!!! Circuits getting power through sneak paths SUCKS. At work
                          recently we had a bus populated with bill validators and coin acceptors
                          (an "MDB" bus), and the controller had to be able to power-cycle the bus
                          when it hung due to EMI issues. (I know, treating the symptom...the
                          real problem is being addressed now) The MDB bus was low-side switched
                          (HORRIBLY BAD IDEA, but thankfully it pre-dated my arrival!) and the
                          microcontrollers in the units on the MDB bus were never able to be fully
                          reset because they were...you guessed it, getting power from the serial
                          lines! Dammit.

                          -Dave

                          --
                          Dave McGuire, AK4HZ
                          New Kensington, PA
                        Your message has been successfully submitted and would be delivered to recipients shortly.