Testing my new seek logic for the Virtual 2315 Cartridge Facility - part 14

ADJUSTING THE DISK DRIVE SEEK SETTINGS

I wired up a number of disk drive signals to the oscilloscope. The amplifier balance and dead band adjustments were previously completed. The operation of the trigger that stops the acceleration of the disk arm so that it comes to a stop .01 or .02 inches from its prior position has to be adjusted. Each of the two step sizes, 10 mil and 20 mil, have their own trigger voltage. The level needed to move farther is higher than for the short step. 

The trigger voltage for the steps - 10 mil or 20 mil - are set to correspond to a velocity of the arm at which the logic stops accelerating and flips over to decelerating the arm to a stop. Later adjustments will tweak this to achieve the most reliable stepping operation, but the initial adjustment voltages get us close. They were pretty close already.

A 10 millisecond timer is adjusted to ensure that the feedback from the disk drive to the controller logic in the 1130 is correct. When a step is requested by dropping the -Access Go signal low, a timer waits 2.6 milliseconds while the detent, a tooth that fits in a rack to hold the arm at a specific cylinder, is retracted allowing free arm movement. After th 2.6 ms, the feedback signal +Access Ready drops low and the arm begins its acceleration. The 10 ms timer determines when the +Access Ready signal returns to high signaling completion of the step.

At the end of these settings, I verified several timings, voltage levels and waveforms. There is a 2.6 millisecond single shot timer that must be correct to give the detents time to disengage from the rack teeth before the arm begins accelerating, for example. I also checked the detent voltages and timing. 

DYNAMIC METHOD OF FINE TUNING THE 10 AND 20 MIL STEP THRESHOLDS

Putting a voltmeter on the threshold potentiometers for the 20 mil step size, I moved the arm continuously in 20 mil steps while rotating the pot. The method involves visually detecting when the detents no longer smoothly engage the rack, perhaps skipping teeth to misposition. The observation has to be over the entire range of cylinders, in both directions.  Recording the low and high voltages where the misbehavior begins, I then set the pot to the 2/3 point between those extremes. 

The idea is that this is going to be the most reliable setting. It is similar to how magnetic core memory is adjusted - finding the extremes where failure begins and then selecting the point in between that is known to work best. For memory, a diagram called a Shmoo Plot is drawn to record the extremes and select the ideal operating point. 

Moving over to the 10 mil step potentiometer and the drive step size to 10 mil, I again continuously move the arm back and forth over the entire range of cylinders while changing the pot until I see misbehavior. Using the low and high voltages discovered by this process gives me the 2/3 point voltage between the extremes and that will be the sweet spot for the mechanism. 

RECONNECTING DRIVE AND USING SOFTWARE TO DRIVE THE ARM BACK AND FORTH

I used a simple hand loop to move the arm rapidly back and forth between two cylinder targets, observing the drive to ensure that it does not stop at the wrong cylinder or otherwise misbehave. I changed the forward and backward seek counts to various numbers, having first moved the arm to a desired starting point. This involved several ranges of seek and performed them across various portions of the rack (ranges of cylinder addresses). The most extreme being 202 cylinders forward and back, the extremes of the legitimate cylinder addresses for this drive. 

RESULTS OF THE HAND LOOP TESTING

The drive is seeking very smoothly and accurately. It always reached the intended cylinder regardless of the size of the movement. I am ready to move on to running the IBM disk diagnostic program to validate that the Virtual 2315 Cartridge Facility (V2315CF) is working as intended and ready to have the system installation completed. 

TWEAKING THE ENCLOSURE THAT WILL HOLD THE V2315CF INSIDE THE IBM 1130

I made some changes to the size of the enclosure that will be mounted on top of the internal disk drive, underneath the 1130 top cover that pivots up. Barry Ward is finalizing his printed drive mechanism and mini 2315 cartridges. I did a test fit of his working version of the drive and the main V2315CF unit installed in the front plate of the enclosure. 

The enclosure is assembled with 1/8" thick black plexiglas and mostly glued together, relying on 1/4" square rods to bolster the joints.

The disk drive fits in the right side of the front panel, but I don't have the current outlines for the holes since Barry is working on that part, so my CAD files for testing the assembly are missing that portion of the design. It was enough to feel confident in cutting the acrylic. 

Work on the internal disk drive of IBM 1130 - part 3

CHECKING 257 TRANSISTOR FROM DONOR SLT CARD

I put the transistor that I pulled off a spare card and tested it with my curve tracer. I compared it to the two transistors on the M2/M3 card which made the weak/failing transistor obvious. Apparently Germanium power transistors can fail gradually with the gain increasing until there is a complete failure. The base junction is thinned out by tin whiskers or holes punched by voltage spikes, causing the current to increase through the base which drives up the beta. 

SWAPPING THE TRANSISTOR AND TESTING THE DRIVE

I removed the failing transistor and installed the donor transistor. The M2/M3 card is now working properly, with no current through the voice coil during idle times. I used the opportunity to adjust the dead band level up until the voice coil began oscillating and then backed if off which is the the procedure for the adjustment. 

I started up the drive and ran a program to seek back and forth 200 cylinders at a time. However, while the drive grunted with the detent cycling 100 times, it didn't move forward from cylinder 0. The second seek in the program, moving backwards 200 cylinders, ends immediately because the drive knows it is already at cylinder 0, the home cylinder. 

DRIVE MOVES FORWARD AND BACK WITH CE SWITCHES BUT NOT BY PROGRAMMING

I had previously moved the arm back and forth using the CE switches - In/Out, 10/20 Step and Single/Continuous movement. When I hooked up the cable from the Virtual 2315 Cartridge Facility (V2315CF) it should disable the CE switches and obey only the signals from the IBM 1130. The arm was not moving forward, yet the signals from the 1130 were correct for forward movement. 

The logic that responds to the -Access Go signal as well as the -Reverse and the -10 Mil control signals does merge the CE switch signals with the signals from the 1130 before activating the disk drive movement. I probed around and realized that the Out/In switch was set to the Out position, which means reversing towards cylinder 0. I noticed that this overrides the signals from the 1130 when the +CE Interlock signal is low. 

MY ERROR - MY HANDLING OF CE INTERLOCK IS BACKWARDS

The intent of the +CE Interlock signal is that when the cable from the 1130 is plugged into the disk drive, the CE switches should not have any effect. Pulling the cable out enables the switches. As I looked closer at the logic diagrams, I realized that I had things backwards. The 1130 delivers +3V on the +CE Interlock wire and the disk drive has a pull down resistor to -3V on the same pin. Thus, when the cable is pulled, the line is logically low and when the cable is attached it is logic high.

I designed for the inverse. I set up the cable to pull the +CE Interlock signal to ground (logic low) thus the signal is always low as far as the disk drive is concerned. Since the Out/In switch was set to Out, it forced the drive into reverse. 

Somehow, every time I tested the V2315CF before this, the Out/In switch was set to In and had no impact. It only overrides the direction when it is set to Out due to the way that Solid Logic Technology gates work. 

REWORKING THE 2310 INTERFACE BOARD TO FIX CE INTERLOCK

The cable from the V2315CF to the disk drive has the wire for +CE Interlock on pin D07 that is hooked to the ground of the 2310 Interface Board. The wire coming from the IBM 1130 to the 2310 Interface Board does does nothing with the incoming signal on pin D07. 

My fix was simple. I removed the disk bound cable wire from D07 and added a bodge wire to tie it to D07 of the cable bound for the IBM 1130. That ensures that the 1130 is driving +3V through the cables which will block the CE switches from working when the cable is attached to the disk drive. Pulling the cable lets the internal pulldown resistor to -3V drive the signal to logic low. 

Grounding signal at board

incoming +CE Interlock

PROGRAM SEEKING BACK AND FORTH 200 CYLINDERS CORRECTLY

With the fix, the program worked properly. The arm moved smoothly back and forth. I will continue with the fine tuning adjustments but everything is very promising. 

FINALIZING THE ENCLOSURE FOR THE V2315CF 

I designed an enclosure for the V2315CF that is bolted to the top plate of the internal disk drive, so that it is hidden under the top cover of the IBM 1130 but accessible by opening that lid. The main unit and most of the power supply and power fail ridethrough equipment is installed inside this enclosure. There is a smart battery trickle charger, a 12V motorcycle battery, and the 2310 Interface Board that are mounted externally, the rest is inside this enclosure.

A friend is building a much more realistic looking mini cartridge mechanism that will also be in this enclosure. This mechanism looks like a miniature version of the internal disk drive as viewed from the front of the IBM 1130. The blue handle will operate to allow a mini cartridge to be inserted, then closing the handle inserts the mini cartridge into the V2315CF circuit. 

design visualization

The mini cartridge is now 2.25" in diameter, a slight increase over my original version but one that permits the cartridge to look much more realistic. I will provide a lot more detail and pictures of this as he completes the work. 

I finalized the laser cutter files to trim the acrylic sheets to make the enclosure. It will be 4 1/4" tall, 13 1/16" wide and 11 5/16" deep, with the main V2315CF unit and the disk drive mounted on the front face. Once I have the acrylic cut, I will glue it up and begin mounting the equipment in it pending only the mini disk drive completion. 

Work on the internal disk drive of IBM 1130 - part 2

INVESTIGATING THE REASON THE VOICE COIL IS ENERGIZED AT ALL TIMES

The internal disk drive of the 1130 has a disk arm that is moved by a linear voice coil, much like an audio speaker. It should only be active when the drive is attempting to move .01 or .02" at a time. The circuit will energize it in the movement direction, then switch the current direction to decelerate it after a tachometer shows it having reached a target velocity. 

This should be turned on with the falling edge of -Access Go or when the drive is first powered up. At power-up, it energizes the reverse direction movement causing the arm to move back towards the home cylinder (0) and that is turned off when the Home microswitch is activated by the arm reaching cylinder 0. Activation during regular operation is by -Access Go which is returned to high when the disk controller sees that the feedback from the disk drive, signal +Access Ready, drops to low. 

Inside the drive, the drop of -Access Go turns on the Go Latch which remains active until the tachometer driven trigger determines that the target velocity was reached. That turns off the Go Latch and allows the tachometer to slow the arm down to zero speed which will complete its movement of .01 or .02 inches (10 or 20 mil). 

I studied the circuit diagrams until I understood how they work and what levels I should see at all the observation points offered by the disk drive circuitry. I annotated the Solid Logic Technology (SLT) card pins that correspond to each observation point, which I used to investigate the reason that the drive was pulling hard in the reverse direction at all times. 

Example - diagram of reverse acceleration

Capacitor C1 in the diagram above is initially charged to +6V and should remain there while the arm is not moving. Once we are in the acceleration phase of a movement, the current from the tachometer amplifier lowers the voltage on the capacitor. When that voltage decreases to a trigger level that is set by circuits below, it turns off the Go Latch and the drive is in the deceleration phase.

During deceleration, the tachometer output drives the power amplifier to energize the voice coil in the opposite direction of which it was powered during acceleration, thus slowing the arm. When the tachometer voltage gets low enough, the voice coil is shut off and the arm comes to rest. The trigger levels were set to move the arm 10 or 20 mils between the acceleration and then slowdown actions. 

These circuits are almost all on card F2/F3, pictured below. The final four transistors are on card M2/M3. Most adjustments for the disk drive are made with potentiometers on F2/F3 and any circuit failure is likely to be on that card.

card in F2/F3

OBSERVATIONS

The tachometer amplifier was driving the voice coil to the rear because it hadn't been properly balanced to zero out when idle. I suspect that the dead band adjustment was also not correct. That sets the minimum differential from the tachometer amplifier before the voice coil is driven in either direction. 

CAUSE AND RESOLUTION

The potentiometers on the card are fine pitched - they have many turns to adjust from one end of the setting to the other - but they no longer have stops to indicate when you have reached an end. I had to take the card out and adjust the pots to get them at their halfway point in resistance. Since one of them had fixed resistors across it as well, I unsoldered it first then adjusted it. 

Once I had the circuit behaving well, producing the same low voltage on both the forward and reverse outputs of the card, the arm shouldn't be trying to move either way. These outputs are routed to card M2/M3 which have the four bridge transistors that switch current to the voice coil based on the outputs coming from card F2/F3. 

However, when I tested, card M2/M3 was driving the voice coil strongly to the reverse direction. With the inputs at the same level, it shouldn't be. That was the next candidate for debugging. Meanwhile, I don't have a good setting for the dead band yet because I need to arm to not try to move while I adjust it so that I can see if/when it begins to oscillate and back off the setting. That will come later. 

INVESTIGATING THE DRIVER CARD M2/M3

I pulled the card, which only has five large germanium transistors on it. One of them is used to select which read/write head is active, so I can ignore that. Each side of the circuit (forward or reverse) has an NPN transistor and a PNP transistor that switch +48V and ground to the voice coil, but the pairs are wired to inverse to each other so that a pair either drives the coil one way, 48 to ground, or the other way. 

I hooked the transistors up to my curve tracer to see if one of them was bad. The two PNP transistors performed identically and had good looking traces. However, one of the NPN was quite different from the other. I suspect that one, although the curve shape wasn't definitive. This is an IBM type 257 transistor

FOUND DONOR CARD IN MY STOCKPILE WITH THE TRANSISTOR ON IT

I did find an SLT card in my pile of spare and ebay auction cards that has the 257 transistor on it. I will pull it off, test to make sure it matches the presumed good one on the M2/M3 board, then replace the bad NPN transistor to repair the card. When that is done, during my next shop visit, I should have a drive whose arm is not activated when not attempting a seek. I can make the dead band adjustment and then resume testing of the drive.

Work on the internal disk drive of IBM 1130 - part 1

RESTORED 1130 POWER RAIL LEVEL LEADING TO NEW SYMPTOMS

When working on the disk drive attempting to make the adjustments for seek behavior, I checked the voltage at a pin that the Field Engineeering Maintenance Manual (FEMM) suggested be tested. The expected level was 8V but I found it at 6.2V. I verified the incoming power rails - +3V, -3V and +6V - only to find that the last was only at 5V. 

I adjusted the linear voltage regulator in the 1130 to get the rail back to +6V. Once that was done, I powered up while watching the same pin on the disk drive. I saw the voltage start near 8V and slowly decline to just over 6V. I had made quite a few changes to two potentiometers on the disk drive circuitry trying to follow the adjustment procedure in the FEMM, so that might be the cause.

I spun up the drive, with the cable from the 1130 logic disconnected, so that I could use the manual movement switches on the rear of the drive. However, when the drive went ready, the arm stayed at cylinder 0 and it would not move when I used the manual movement switches. 

I manually released the detent that holds the arm in position. These are solenoid operated chisel shaped teeth that fit into V shaped notches in a rack that moves with the disk arm. There is a detent for even cylinder numbers and another for odd cylinder numbers. The rack has notches for each cylinder - spaced 10 mil (.01") apart. 

Holding the even detent out of the notch for cylinder zero, I tried to slide the arm outward but I felt that the voice coil magnet was strongly energized pulling the arm back to the home or cylinder 0 position. That should not be. The voice coil should only energize to move the arm during a seek operation, taking either one (10 mil) or two steps (20 mil distance). 

One of the adjustments I had previously tried to adjust is called the dead band - it determines how much of a drive signal level is needed to start the arm accelerating or decelerating. Set too low and the circuit could oscillate with small random fluctuations in drive current. I set it much higher just to be safe, but the coil remained pulled at full force to the rear. 

The other adjustment is the dynamic balance of the tachometer amplifier. A tachometer attached to the arm produces pulses as the arm moves - the faster they arrive the higher the output from the amplifier. However, it should be set so that a net zero voltage is produced when the arm is not moving. Nothing I did to the pot for this setting would move the net voltage off of zero, however, that doesn't make sense. It is possible that the amplifier is producing a strong movement signal. 

Still, the movement signal should be gated by the logic circuitry on the disk drive so that it only attempts to move in two cases - either it gets a movement signal from the 1130/manual switches, or it is doing its initial power-on reversal to ensure the arm is always at the home cylinder as it powers up. 

The two sides of the voice coil are driven by a power amplifier which gets its drive signals from the card where I was adjusting potentiometers. It is designed to put 48V across the coil in one of two directions, depending on whether we want to move to higher cylinder numbers or back towards 0. A quick check of the voltages on two of the drive control signals showed them at about 2V and about 5V. That is not a valid combination of voltages so I have some debugging to do. The drive previously worked until I began with the adjustments, but some part may have failed. 

I ran out of the short window I had at the shop, since I had somewhere to be. I will be collecting information and building a debugging strategy to figure out what is failing to cause the arm to be retracting strongly. 

Testing my new seek logic for the Virtual 2315 Cartridge Facility - part 13

ASSEMBLED NEW VERSION OF 2310 INTERFACE BOARD

The Virtual 2315 Cartridge Facility (V2315CF) has a component that is a board that interconnects the cables from the IBM 1130 disk controller logic, the internal disk drive, and the main V2315CF unit, as well as a few other functions. I adjusted the design to include electronic switches that connected the signals between the internal disk and the disk controller logic when the V2315CF is in real mode, the norm. The alternate virtual mode does not use the internal disk thus all signals flow only between the disk controller logic and the V2315CF. 

ADJUSTING THE DISK DRIVE SEEK SETTINGS

I sat down with the maintenance manual, a scope, a voltmeter and tools in order to make all the adjustments to the internal disk drive. Simply unplugging the cable from the V2315CF from the disk drive electronics put it in Customer Engineer (CE) mode where I could operate it without using software on the 1130. There are CE switches to move the arm in and out, in both 10 mil (1 cylinder) and 20 mil (2 cylinder) steps, either once per switch press or continually while the switch is held on. 

The first step is to perform a dynamic balance of the tachometer amplifier output, setting it so that there is zero output voltage while the arm is not moving. The next adjustment is for the voltage at which the mechanism stops accelerating the arm and turns off the go latch. This drops the +Access Ready signal, flips the voice coil polarity to begin slowing the arm and starts to engage the appropriate odd or even detent to lock the arm in at the target cylinder position. 

There are different target voltages for 10 mil and 20 mil, since the arm has to move farther for a 2 cylinder step and that requires a faster speed. The target voltage is compared to the voltage on a capacitor which is (dis)charged by the tachometer output, e.g. by the speed. 

A third adjustment sets the dead band, the level below which the servo does not attempt to move the arm. This is done to avoid oscillation while still providing enough oomph to properly accelerate and decelerate the arm. 

The last of the electrical adjustments is for the timer which keeps the +Access Ready signal low from when the trigger voltage stops arm acceleration until the detent should be fully engaged and the arm not moving. The target is 10 milliseconds. 

STARTED TO CHECK DISK DRIVE SETTINGS AND DISCOVERED VOLTAGE ISSUE

When I began to adjust some of the potentiometers in the disk drive I saw weird behavior. The balance was always zero, no setting of the pot would produce anything else. The oscillations never occurred when changing the dead band. These settings are all on the Amplifier card at slot F2 and F3 of the disk drive card cage. The manual suggested voltage checks at certain pins, one of which should be 8V but came up at only 6.2V. 

A QUICK CHECK OF THE INCOMING POWER RAILS TO THE DISK DRIVE FOUND ISSUE

The 1130 generates the DC voltages for the disk drive, the three main power rails as well as 48VDC. The power rails for Solid Logic Technology (SLT) are +3V, -3V and +6V. The 6V rail was only generating 5V, which is way too low for good operation. I checked back at the main logic gates of the 1130 and found that the voltage was low at that point as well. This is a definite problem for the entire system.

ADJUSTED LINEAR VOLTAGE REGULATOR TO RESTORE RAIL TO +6V

I adjusted the potentiometer on the regulator to restore the voltage to the correct level. I have no idea how this got so far out of spec. It might explain some of the odd results or intermittent issues I observed in the past month or so. 

The 1130 has a circuit that verifies the presence of all three primary voltage rails, shutting down the machine or blocking power-up if they aren't good. However, it isn't highly precise, since 5V was high enough to pass the check while still being far too low for proper operation. It is intended more to detect the lack of one of the rails, not that they are correct. 

Testing my new seek logic for the Virtual 2315 Cartridge Facility - part 12

TRYING OUT A TERMINATOR AT THE 1130 BACKPLANE TO IMPROVE SIGNAL

The +Access Ready signal that is generated by the disk drive in real mode of the Virtual 2315 Cartridge Facility (V2315CF) or by the V2315CF in virtual mode. It is feedback to the disk controller logic in the 1130 as it moves the disk arm to seek to different cylinder locations. There was some ringing and glitches on the signal which did cause issues for the disk controller and for the V2315CF which needs to monitor the disk movement to keep track of the current cylinder. 

I have tried multiple things to improve the signal as received at the pin on the 1130 in gate A, compartment C1 where the disk driver logic resides. I changed the impedance of the terminator installed in the V2315CF, which is on the sending end of the signal to the 1130 and receiving end for the signal from the disk drive. I removed the terminating resistors entirely. I shorted the +Access Ready signals from the 1130 and the disk drive together on the 2310 Interface Board so that the V2315CF is electrically uninvolved other than additional cable lengths. 

All of these changes made some improvement. The last one gave me excellent agreement between the disk drive arm and the intent of the XIO Seek commands I issued. However, the V2315CF failed to completely shadow the movements, missing a couple of cylinders on longer seeks. 

I therefore decided to add termination right at the destination pin on the 1130 backplane. The IBM design did not have terminators there, other than whatever pullup resistor they used on the backplane after the signal arrived. I couldn't find any resistor or source for a pullup. The signal is generated by a single shot timer in the disk drive itself. 

I have no component level circuit schematic for that gate so I can't tell if it has pullup or not. In any case, having the pullup on the source side which is typical for the 1130 logic would not protect the input pin from ringing caused by inductance and capacitance in the cabling over to the SLT backplane. 

I therefore decided to mount termination resistors right at slot L4 of the backplane, where the +Access Ready signal is connected at pin D04. I used a 330 ohm pullup to +3V which is available on pin D02 of the slot and a 470 ohm pulldown to ground which is available on pin D08 of the slot. 

I had to find a good way to install this in that area. I chose to solder the two resistors together with very short leads at the connection. I then installed wirewrap wires to pins D02, D04 and D08, with their other ends loose. Once I find a way to mechanically secure the resistors in place between the rows of pins on the backplane, I will cut the wires and tack solder them to the common point and other ends of the resistors. For the time being, I did a quick and dirty installation.

OBSERVATION OF SIGNALS AND BEHAVIOR WITH TERMINATOR RESISTORS IN PLACE

The signal looked almost exactly like the signal leaving the disk drive at the backplane pin of its electronics, thus I was not getting any distortions based on the run of the signal through various cables, boards and traces on its way to the 1130 backplane pin. 

The drive would still fail to seek properly on some longer seeks, stopping a few cylinders short of its intended destination. It would also lock up and buzz at a point short of cylinder 202 but well along in its travel. Something is still wrong with the drive that I have to resolve. I need to go through the entire sequence of settings per the manual.

Below is the image I took of a seek of 193 cylinders.

Yellow is the +Access Ready signal as it is delivered to the 1130 disk controller backplane pin. Green is the -Access Go command from the 1130 disk controller that triggers each step. Purple is the -10/+20 Mil signal which makes the first step 10 mil and the remainder be 2 cylinders. Blue is the source +Access Ready signal at the disk drive electronics backplane pin. 

The expected waveforms and timing are shown in one of the maintenance manuals, which I will be checking while the drive performs some steps. 

There are four adjustments I will check and change if necessary that control the timings shown above.

1. Tachometer amplifier static balance - no output when carriage is not moving
2. Dead band of servo - adjust to eliminate oscillation but achieve proper movement
3. 10 mil and 20 mil stop voltages - duration of acceleration for each step size
4. 10 millisecond timer duration - time that +Access Ready stays low

I will also verify the 2.6 ms single shot is the proper duration to allow the detents to retract before arm movement starts. Watching voltages and timing of other signals above help identify if any components are out of spec and causing issues.

The FE maintenance manual has multiple methods of adjusting the drive values above, including a very promising method that observes the continuous movement of the arm for erratic detent operation and a means to tweak the adjustments for the most reliable and consistent stepping. 

Testing my new seek logic for the Virtual 2315 Cartridge Facility - part 11

STILL FOCUSING ON SIGNAL INTEGRITY FOR V2315CF SEEK

I was not happen with the glitching and malformed signals going into the 1130 for the +Access Ready feedback signal. I am determined to clean this up and not just hope that the Virtual 2315 Cartridge System (V2315CF) works in spite of the signal distortion. The first thing I did was to remove the terminator resistors for that signal, to see what the signal looked like. It was still distorted.

The new version of the 2310 Interface Board, fabricated and current being shipped to me, adds electronic switches to short the signals between the disk and the 1130 while in real mode, rather than trying to generate the signal as a copy of the incoming one. I installed a short bit of twisted pair to bridge the B07 pin from the disk drive to the B07 pin heading into the 1130 disk controller circuitry. The result was considerably better.

The blue trace on the bottom is the +Access Ready signal generated by the disk drive and the yellow trace is the same signal at the entry to the disk controller SLT cards inside the 1130. The green trace is the -Access Go signal that goes from the 1130 to the disk to start a seek. Finally, the purple trace is the -10 Mil Step signal, which initially requested a single track move and then rose so that the additional steps are two tracks each. 

The temporary bridge connecting the signal straight through

Above is a trace of a seven cylinder seek. The -10 Mil Step signal is low for the first step and high for the next three. It does look good, although with a time scale of 5 milliseconds per division any glitches wouldn't be visible there. 

V2315CF AND 1130 IN PERFECT SYNC, DISK SOMETIMES IS OFF

I pounded away at the disk trying small, medium and long seeks in both directions. I was pleased that the disk arm position is now exactly in agreement. The only issue is that the V2315CF shadowing of the arm position sometimes drops a few cylinders from the count. 

I am hopeful that the new PCB with its electronic switches will further improve the fidelity, as it will manage seven signals that run between the disk and the 1130 controller circuitry. By the end of the week I should have built the new version of the board and installed it. I can troubleshoot further if the issue is still present with the new board version. 

VERIFIED THE POWER FAIL PROTECTION

I wired together the elements that give the V2315CF the ability to detect a power failure in the 1130 and conduct a full writeback of the RAM contents into the mini cartridge leveraging a 12V motorcycle battery. 

The power distribution unit board, in the center of the photo above, routes power between the 1130 12V rails, the motorcycle battery, a smart trickle charter, a timer module, the power supply for the V2315CF and then the V2315CF unit itself in the lower left. At the top is a bench power supply that stood in for the 1130, allowing me to flip it off to verify that the timer powers the unit during a 40 second interval, and that the V2315CF properly detected that the 1130 12V power dropped. It is ready to be wired into the 1130 system. 

WILL MAKE ADJUSTMENTS OF THE INTERNAL DISK DRIVE SETTINGS

The internal disk drive has a set of six adjustments that are important to proper behavior. These include the thresholds for the 10 and 20 mil seeks, the timers for the various actions during a seek, and the sensitivity of the seek movement circuitry. Those, coupled with the mechanical stops at cylinders 0 and 202, will ensure the most reliable operation. 

Testing my new seek logic for the Virtual 2315 Cartridge Facility - part 10

NEW TERMINATOR BOARD IMPROVES BUT DOES NOT FIX V2315CF SEEK TRACKING

The Virtual 2315 Cartridge Facility (V2315CF) watches the control signals between the 1130's internal disk controller and the internal disk drive (named 2310 or 13SD or RAMKIT). It shadows the position of the disk arm as commanded by programs issuing XIO Seek commands, since the data to be returned when the software reads from the disk depends on which of the 203 cylinders the read/write heads are positioned over. The V2315CF generates the data signals instead of the disk heads. 

The current situation is that requests to move the arm are obeyed correctly by the disk drive. The physical arm is in the correct location after every move. However, the location as shadowed by the V2315CF is off, short by 1 or 2 cylinders from the actual final position. The error is cumulative.

The programmer requests a movement of some number of cylinders relative to the current arm position. The disk controller implements that as a series of steps of 1 or 2 cylinders at a time, the first being 1 cylinder only if the requested movement count is an odd number; all others are 2 cylinders at a time. 

Glitches on the signals between the 1130 and the disk drive do not appear to affect the disk drive's operation but confuse the V2315CF logic. I can work on two fronts here - eliminating the glitches entirely and/or ensuring that V2315CF is invulnerable to the glitches. 

OSCILLOSCOPE OBSERVATION OF SIGNAL ISSUES USING NEW TERMINATOR BOARD

I set up the oscilloscope to watch +Access Ready inside the 1130 at gate A, compartment C1, where the signal enters the logic card. A second probe was connected to the V2315CF main box where the +Access Ready signal is emitted. A third probe was attached at the pin on the 2310 Interface Board where the cable is attached running to the 1130 logic where our first probe is placed. The final probe monitors the signal as it is produced by the disk drive. 

I triggered on the falling edge of +Access Ready, using my test code in the 1130 to issue an XIO Seek of 3 cylinders forward which is one that we have seen malfunction consistently. I captured the waveforms and looked to see what was happening to the signal at several points in its routing and hoping to understand what is producing the glitchy behavior. 

I can see a very strong drive from the output chip on the V2315CF and then substantial ringing of the signal where it enters the 1130 logic circuits. Part of the problem is ground bounce between the 2310 Interface Board, the V2315CF, the 1130 and the disk drive. I will add some wire braid for high frequency ground paths, although that is not the issue here. 

NEW LOGIC IN FPGA TESTED TO SEE IF IT MATCHES DISK ARM POSITION

I modified the seek shadowing logic in the V2315CF and installed it to the flash memory that initializes the FPGA. I ran some seeks of various sizes, noting for each whether the arm position matched the intent of the seek and whether the V2315CF reported cylinder matched the other two. 

We were down to one issue with the shadowing. The V2315CF seemed to miss the first seek, that error in positioning remained for all subsequent seeks. I tracked that down to an error in my Verilog, where I look at the state of the Home signal and don't bump the cylinder when it is on in real drive mode. That means we don't count the first step away from cylinder 0.

I updated the logic and tested again. The disk drive and V2315CF are in complete agreement with the seek amounts I executed. I ran a series of forward seeks in a small program, issuing the next seek as soon as the previous one caused an operation complete interrupt. I moved 40, 20, 10, 7, 6, 3 and 1 cylinders with the short program. It completed in a fraction of a second with the disk arm and V2315CF showing the current cylinder location as 87. 

Testing my new seek logic for the Virtual 2315 Cartridge Facility - part 9 - Fixed cylinder 202 issue

ADJUSTING THE FORWARD CRASH STOP ON THE INTERNAL DISK DRIVE

Crash stop when arm is not at 202

Arm against crash stop just past 202

In order to be safe, I changed the stop to a gap of .008", giving the maximum margin for acceleration while still blocking the ability of the arm to move to cylinder 203. With an attempt to move 20 mil, the arm should stop and the detent should still fall back into the notch for cylinder 202. 

Just as importantly, it should not do anything bad if it gets a 10 mil step thus engaging the odd detent instead of the even detent. Looking closely at the positioning rack, I can see there is a notch at the 203 location which the odd detent can enter. 

The way you read the location is to look where the right detent is pointing. The numbers are etched every 40 tracks, with a large vertical bar at the 20 track intermediate point. The fine vertical lines on the positioning rack mark each 4 track step, with a notch in between at the 2 track point. The left detent (odd) will fall in to indicate that the location is the number you read from the right detent plus 1. 

The picture is of the arm when I had a .006" feeler gage between the crash stop and the arm, the arm not quite far enough along to let the even detent fall into 202. However, you can imagine the rack slightly to the right so the right (even) detent enters the 202 notch. This is where the arm stops at the last legitimate location. 

If we attempt a 1 track (10 mil) step forward, the drive flips which detent is used to stop, now engaging the odd (left) detent. Since there is a notch there that the odd detent can enter, it will try to hold the arm there. That means the drive is trying to be at 203 but the mechanical stop is trying to keep it from moving past 202. 

It may be slightly mispositioned to read or write on the cylinder but it is a reasonable state to stop in. As long as the circuitry in the disk drive turns off the acceleration of the arm so it doesn't lock up buzzing, this is a decent outcome. It will back up correctly. A step of 10 mil will bring it back to 202 with the even detent engaged, a 20 mil step will move it to 201 still using the odd detent. 

The IBM maintenance manual blithely instructs the field/customer engineer (CE) to run the drive, use the CE switches to move the arm out to cylinder 202, then adjust the crash stop to a specific gap. The stop is a bolt that is threaded into a big block holding the arm mechanism together, with a nut that locks it in position. 

The issue is that the slot in the bolt that one turns to adjust the crash stop is directly against the 2315 cartridge case that is in the machine. There is no room to fit a tool in there to turn that once the nut is loosened. If I remove the cartridge, the drive will not spin up nor load the heads, thus it wouldn't allow arm movement or perform seeks. 

However, I can manually push the microswitch that detects a cartridge is inserted, to allow the drive to spin up and load the heads. They won't really load for two reasons - no platter to ride on, but more importantly my modifications have stopped the heads from being pushed down so they will remain safely apart 

I did find a way to use pliers to turn the threads of the bolt after loosening the nut, thus could adjust this with the arm actively sitting at 202. The gap was much too small, so I adjusted it out to a suitable amount that would stop the seek action but not move past 202. 

TRYING TO MOVE THE ARM PAST 202 AFTER MY ADJUSTMENT

The disk drive is designed so that if the cable that plugs into the drive from the controller (or from the 2310 Interface Board in our case) is removed, the Customer Engineer (CE) switches and buttons can control arm movement without having to issue XIO commands or involve the 1130 disk controller logic. 

I moved the arm to cylinder 201 from zero by setting the Step switch to 20 Mil and held the switch at Continuous movement until it got close. I then set the Step switch to 10 Mil and advanced one step at a time until I verified that the arm reached that point successfully. This would have the odd detent in the rack.

I then moved the arm 1 track, with Step set to 10 Mil and using a single advance, which should put the arm at 202. That was verified by looking at the arm . The even detent should be engaged.

Now, the moment of truth. I issued a seek of 2 tracks farther, with Step at 20 Mil and a single movement request, which should leave the the arm at 202 and the drive should recover from the attempt to move past. The even detent should remain engaged. Worked great.

I cycled the drive to get the arm at the home cylinder and used 20 Mil steps to advance the arm to 202. I moved 1 more track, with Step set to 10Mil, wanting to see what happens with the detents and the arm. I looked at the detent and the arm position, as well as the V2315CF. If the odd detent is now engaged, a reverse step of 10 mil would actually stop the arm at cylinder 200 instead of 201, which is not desirable. If the even detent remains engaged, I need to understand the mechanism.

After testing and observation, it does engage the odd detent, but has bounced back to 201. I hooked up the system and did plenty of seeks, with the pathological behavior gone. This problem is now resolved. 

Testing my new seek logic for the Virtual 2315 Cartridge Facility - part 8

FINISHED MODIFYING TERMINATOR BOARD WITH NEW RESISTOR VALUES

The terminator board that fits into the Virtual 2315 Cartridge Facility (V2315CF) main unit is the same one used successfully with the unit in its original role as George Wiley's RK-05 Emulator connecting to a DEC computer to emulate a disk drive that is a derivative of the IBM drive. The terminator implemented the same resistance values as the M903 terminator board built by DEC. 

The impedance of the terminator connections is just over 122 ohms, which may match the impedance of the cabling in the DEC computers. The RK-05 emulator kit offers an option A1 RK-05 to RK8-E that has two IDC 40 pin ribbon cables to make the connections to the controller card, since the RK-05 drive needs 36 signal lines for the interface between disk and controller, deploying ground lines between each signal line on the ribbon cables. 

Ribbon cable characteristic impedance for single ended signals (not differential pairs) is 80 to 100 ohms with interspersed ground wires. Using the DEC terminator board yields a big mismatch in impedance, which will lead to reflections of about 20% of the signal level. If the rise time of signals is faster than 6 nanoseconds the signal will ring and appear to have spurious transitions. 

I determined resistor values to drop the terminator impedance to about 100 ohms, which would minimize the issues even with the current sharp rise times. The Lattice FPGA does not support the means to slow the signal transitions to 7ns or slower instead typically rising at 1 to 1.5 ns.

I removed all the original resistors - a mix of 1206 size and 0603 size for the pullup to 5V and the pulldown to ground respectively - and cleaned up the board.

The original terminator board has a few unpopulated resistor locations - the crude circles show you were to concentrate to see them. 

I installed my 179 ohm pullup and 249 ohm pulldown resistors onto all the positions where I have signals connected to the ribbon cables. The new terminator board was tested and then plugged into the V2315CF to do more testing. 

DETAILED VIEW OF EFFECT OF SIGNAL GLITCHES ON SEEK BEHAVIOR

A glitch on the +Access Ready signal causes the 1130 disk controller logic to malfunction, thus not forming correct step requests to the drive. The disk drive responds to brief glitches in -Access Go that are induced by the +Access Ready glitch, so it does move to the requested cylinder. The V2315CF fails to recognize the malformed sequences thus it does not reach the same cylinder number as either the drive or the intent of the XIO Seek commands. 

The dance of signals we should see is the controller drop -Access Go with -10 Mil Step low, holding it until the drive drops +Access Ready about 5 milliseconds later. When +Access Ready drops, the signal -10 Mil Step will rise from low to high since only the first seek should be a 10 mil step. 

Also at the time when +Access Ready drops, if the count register (Bits 7 to 14 plus the -10 Mil Step which is bit 15) has all bits on, the -Full Word Count signal is dropped which terminates the seek in the 1130 disk controller logic.

The disk drive, meanwhile, moves the arm for another 9-10 milliseconds then raises +Access Ready to indicate the seek step is complete. If the count register is not all ones yet, another cycle of -Access Go dropping with its +Access Ready responses will occur, iterating until -Full Word Count drops. 

Above, we will examine a successful seek of 1 cylinder. We can see the signals conforming exactly to the protocol. After -Access Go is dropped with -10 Mil Step low, we see +Access Ready drop 5 milliseconds later. At that time, the controller changes -10 Mil Step to high, which turns on -Full Word Count, stopping the controller from trying any more steps. At 15 ms from the start of the step, +Access Ready returns to high indicating the move is complete. 

When I did a move of 3 cylinders, however, the glitches arrived and the process breaks down. The arm moved three cylinders but the V2315CF only saw a single cylinder step. Lets look in detail below, zooming in to the time when the glitch first arrives:

At the 5 ms point, when +Access Ready should drop for 10 ms, we see a very short blip downward, circled in red. This turns off the -Access Go and makes -10 Mil Step go high as pointed by the green arrow.

We see another -Access Go, but very brief, indicated by the blue arrow. Since -10 Mil Step is now high, this asks for a 20 mil step. The short drop of +Access Ready just to the left of the blue arrow causes the disk controller to add 2 to the count register, making it reach the all-ones state and turning on -Full Word Count pointed to by the pink arrow. The seek is over so no further step requests. 

The drive did respond to the additional -Access Go, in spite of the prior seek not being finished. It is only a 20 nanosecond pulse, borderline for the Solid Logic Technology (SLT) edge detectors to see but obviously long enough. Somehow the drive managed to do the second step correctly, as the arm did end at the proper spot. 

Since the V2315CF logic is generating the correct protocol dance - keeping +Access Ready low for 10 ms starting 5ms after the falling edge of -Access Go. It doesn't watch for the next -Access Go falling edge until after the 5ms and it misses the defective short mistimed pulse that the disk drive responded to. 

The drive even responded to a 5 cylinder seek with a correct movement. The signal pathology is even uglier, as seen below:

We see that a single step is observed as far as the V2315CF, based on the 15 ms dance expected. Zooming in on the glitch area, we can see how it managed to move 10 mil, 20 mil and 20 mil to achieve the 5 cylinder goal. 

The glitchy +Access Ready is see to the left of the green arrow, resulting in the -10 Mil Step signal going high where the green arrow points. The first step we began is 1 track, but the subsequent moves will be 2 tracks each. A glitch from the controller shown with the first red arrow drives the arm another 2 tracks. Later we see the -Access Go drop for a longer interval, driving the last 2 track move. This adds 2 to the count register, completing the count of 5 and turning on -Full Word Count to complete the XIO Seek. 

Testing my new seek logic for the Virtual 2315 Cartridge Facility - part 7

TESTING NEW SHADOWING MECHANISM

I loaded the FPGA and PICO of the Virtual 2315 Cartridge Facility (V2315CF) with the updated functionality and repeated the tests from before - using the disk in real mode, issuing seeks of various distances. I wanted to check that the V2315CF reported the ending cylinder number that is expected from the distance of the seek, that the disk arm moved to the correct ending cylinder number, and that the program seeks a successful completion. 

The arm appears to be reaching exactly the cylinder it should based on the XIO Seek commands issued. The V2315CF is still missing some counts which is still symptomatic of the glitches from +Access Ready as it appears at the 1130 backplane. This is a combination of the disk controller logic being tripped up by the glitches and the V2315CF misbehaving. 

Interestingly, the disk drive still sees the correct number of steps, even if it is a short glitch on -Access Go that was triggered by the glitch on +Access Ready. While I could probably develop logic in the V2315CF that would be insensitive to the glitches and mistakes of the disk controller, the right thing to do is to fix the glitches themselves. 

The top signal is -Access Go as emitted from the 1130 backplane. The second signal is +Access Ready as it comes back from the V2315CF. This is produced from a timer based logic chain thus the signal going out of the FPGA pin is unglitched but it sure isn't when it arrives at the 1130. The third signal is the -10 Mil Step signal, being advanced to a 20 mil step because the glitch appears to be the drive dropping +Access Ready as it is in the midst of the 10 mil first step. 

SIGNAL INTEGRITY WORK PLANNED TO ELIMINATE THE GLITCHES

I had already ordered new terminator resistors to improve the impedance matching. The parts arrived today and I carefully removed the 113 resistors on one of the terminator boards and clean up all the pads. It is ready for me to solder on my 1206 size 179 ohm and 0603 size 249 ohm resistors, when I get back to the shop.

The new version of the 2310 Interface Board is still in fabrication in China, but I have the parts to put it together if I believe that the signals will be improved by connecting source to destination directly for all the important seek related signals, eliminating the FPGA in the signal path other than to monitor some of them. 

I am going to work on the twisted wire shielding for the cable I made to run from the 2310 Interface Board to the disk drive. I may be picking up induced signals that are leading to the glitches. 

I need to put the oscilloscope on +Access Ready both at the 1130 backplane at as close to the V2315CF circuitry driving it as I can. I want to see exactly what these signals look like that are causing the issue. 

ORDERED THINNER CARTRIDGE BOTTOMS FOR MINI CARTRIDGES

The mini 2315 cartridges used with the V2315CF are built with a circular white PCB marked to look like the top of a 2315 cartridge, mounting the micro SD card socket and other parts underneath, and placed on a 3D printed bottom cover to make it look more like a tiny version of the physical disk cartridges. 

I thinned out the bottom cover a bit. I can't make the depth proportional to a full sized 2315 cartridge because that doesn't allow enough room for the sd card socket and the header pins that connect the mini cartridge to the V2315CF. However, I had excess material on the bottom which I cut back to improve the esthetics a bit. 

Actual color is white

I ordered a batch of 25 from CraftCloud3D.com in white PLA. These will be glued to the bottom of the PCB to make up the mini cartridge. Proportionally, the walls would need to be about 1/4 of the height in order to look like a tiny 2315. 

I could get closer by notching out the space where the right angle 2x4 header is installed (the opening in the front of the image above) so that it hangs down below the bottom of the mini cartridge. The plastic and copper of the header would be visible below the mini cartridge yet the rear portions look like the 2315 in miniature. 

INVESTIGATING THE MECHANICAL STOP AND THE CYLINDER 202 ISSUE

I had observed an issue when the disk controller attempted to move the arm past the physical limit of the disk drive. The disk implements 203 cylinders, thus can move from cylinder 0 (home) to cylinder 202. Seeks are relative operations, requesting a movement of a number of cylinders from the current position. Thus, it is possible to request a movement that would naively move the arm past cylinder 202 (or backwards past the home cylinder). 

The disk does not directly support a relative movement of an arbitrary number of cylinders. Instead, it can move exactly 1 or 2 cylinders, a step size of 10 mil or 20 mil. The disk controller converts the count of relative movement given by the 1130 XIO Seek instruction into a series of 1 or 2 track steps to move the arm to the final destination. 

The drive has two mechanical stops that block the arm from moving past the two limits. In addition, there is a microswitch that turns on when the arm is at the home cylinder, which the disk controller uses to block any further reverse movements. There is, however, no microswitch to indicate that we are at cylinder 202, so the disk controller will try to move forward to complete its count and the disk drive will try to move 1 or 2 cylinders per step until the controller stops requesting steps. 

Cyl 202 crash stop in red, home crash stop nut in green

Moving a step in the drive involves releasing a detent, a tooth that fits into a notch in a rack to hold the arm at its current position. It then accelerates the disk arm using a voice coil magnet up to a set velocity, then decelerates to bring the arm to a stop. The detent is also released to fall into the appropriate notch for the new cylinder location. 

If the mechanical stop blocks the arm from moving, it can't accelerate to the trigger velocity. The circuitry to stop the voice coil acceleration depends upon that trigger being reached. Thus, one of three conditions can occur. First, the velocity reaches the trigger level and the movement stops. Second, some backup method stops the movement of the arm even though it doesn't reach the trigger speed. Third, it doesn't reach the trigger speed and the backup method fails to work, resulting in the circuit perpetually driving the voice coil to accelerate the arm. 

In the third case, the drive makes a continual buzzing sound and even a system reset of the 1130 will not stop it. Only by turning off the run switch can we break the drive of that pathological state. This is exactly what we are experiencing when we try to move past 202. 

There is a vague mention of the backup method - stating that "a resistor" causes the trigger to be turned off in about 40 milliseconds because it charges the integrator that is otherwise only charged by the tachometer sensing the speed of the arm. Normally the acceleration reaches the trigger level in less than 4 ms, so this would definitely be a safety net for the drive when it is blocked from actually moving. 

However, nothing in the manuals defines where this path is or how it works. I suspect I understand the mechanism and the way it might fail - only a very few components could still allow normal arm movement but fail to trigger when the arm is blocked from moving.

My first check will be the setting of the mechanical stop that blocks the arm from passing 202. It must have a gap between the arm and the stop, when the arm is at 202, of .003 to .009 inches according to the manual. This must be wide enough that the tachometer can charge up to the velocity trigger before the arm stops moving, otherwise we reach condition 3 that is the pathology observed. 

A single track is 10 mil wide, .01 inches, thus the outer range of the setting has to be less than this to ensure we don't actually move to cylinder 203. Thus, the max gap is .009" for the stop setting. However, we don't know how close the minimum gap can be and still block a 20 mil step. It depends on how long it takes to accelerate the voice coil to its target velocity for that step size - the velocity needed for 10 mils being lower than that for 20 mils. 

Ideally the design of the drive provides some safety margin so that even with the minimum gap of .003" the target velocity can be attained, but I have no data to see the speed we could reach in that case nor the trigger level needed to shut down the acceleration. 

The gap did appear to be too tight- with a .006" feeler gage in place, the arm did not get the even detent tooth fully to the notch for 202. 

Something occurred to me as I thought about the condition when we are already at cylinder 202 and attempt to move beyond it. If we try to move 10 mil, the disk drive circuits alternate between odd and even detents to achieve a 1 track movement. If we step past 202 with a 10 mil step, the detent will flip from even to odd. 

The tooth can't fall into the notch for 202 - that is a notch on the rack under the even detent. It must fall into a notch under the odd detent, which might put the arm back at 201 or let it set at 203. There is no circuitry to stop the alternation of detents, thus no way to block this if we are at 202 and try a 10 mil step. This does not sound like a useful behavior for the drive, but I can't see how it will detent properly back at 202 in this scenario. 

TESTING THE SEEK PAST 202

I did some testing using the small test program I put into memory. My first seek from cylinder 0 was to reach 201, which it did successfully. I verified the arm position and that the odd detent was the one engaged in the rack. 

I then issued a move of 1 cylinder forward, which caused the drive to successfully stop at 202 with the even detent engaged. Now I was ready to attempt to move past 202 to see what would occur. 

I did a seek of 1 cylinder again, which switched to the odd detent and tried to move past 202. It entered the pathological condition, buzzing with the voice coil steadily energized trying to move the arm forward. This did confirm however that the design of the disk drive does not recover to a good state if the advance past 202 is done with a 10 mil step. 

I believe I tried again, getting the arm to 202 but this time I moved 2 cylinders so that the even detent will remain engaged. This time it attempted the move, ended the seek normally and was sitting at 202 afterwards. My conclusion is that the drive should be shut down and restarted if the program ever tries to move beyond 202, since in the best case it is on 202 but it could slip back to 201, try to hold position without a detent in a notch, or enter the pathological case.

Improved Virtual 2315 Cartridge Facility behavior when drive powered down then back up

SITUATION BEING ADDRESSED

When we have loaded a valid mini cartridge into the Virtual 2315 Cartridge Facility (V2315CF) it should remain active regardless of whether we power the disk drive down and back up. In the real world, a 2315 cartridge that had been inserted in the drive will still be there when the switch if flipped to run. So too, the virtual cartridge should remain available if we spin up the disk drive.

When we have powered down the disk drive, the V2315CF state machine sits in its run level 9, waiting for the disk drive to signal File Ready after the drive believes it has loaded the read/write heads on to the disk platter inside the 2315 cartridge. Once the drive goes ready, V2315CF advances to run level 10 where the 1130 can access the disk however it wants.

Powering down from run level 10 will simply step back to run level 9, since we do have a valid virtual 2315 image loaded into the V2315CF. The next time the disk asserts File Ready after we spin it up, we will go to run level 10 and be able to use the cartridge again from the 1130. 

What was not properly modeled before was that the cylinder number that V2315CF records to match the position of the real disk arm does not go to zero when the drive is powered down (File Ready goes off). That is what the real drive hardware does, but V2315CF naively keeps its previous cylinder number instead.

SIMPLE FIX APPLIED

The same logic in the PICO code that turns the File Ready lamp on or off on the V2315CF main box will now reset or set a flag in the FPGA that forces the cylinder number to 0. Anytime the RDY lamp on the V2315CF is turned off (and the File Ready lamp on the main 1130 console is off), the cylinder will be locked to 0 (home cylinder). Turning on the lamp also unlocks the arm so that seeks from the 1130 can be recorded to set the cylinder to the appropriate number.