FIRST STEP - CLEANING DIRT AND DEBRIS FROM THE DRIVE
I spent a good hour using compressed air cans to blow dust and other loose material out of the drive. Each area was cleaned using 409 and paper towels until the towels were no longer black with dirt. There was a thick coating on some of the plastic parts that was particularly troubling.
REASSEMBLING MECHANISMS TO RECIEVE DISK CARTRIDGES
I put together the drive receiver mechanism - this is a shroud that tilts up when the loading handle is open, into which the disk cartridge is inserted, then it tilts back to position the cartridge perfectly for the heads to load onto the surface. There are springs, pivot bolts and other parts that fit together. The mechanism is quite clever, as it has to open the cartridge as it slides into place.
A disk cartridge is a housing that covers the 14" disk platter with its brown magnetic oxide coating. There is an opening on the front that is pushed open by the receiver while the cartridge is being slid into place. There is a plate on the underside of the cartridge which keeps dirt from entering the cartridge, but that has to be pushed up inside the drive to allow clean air to be blown inside. It is impressive how it accomplishes all this with just one smooth insertion of the cartridge by the operator.
 |
 |
| How valve lets in clean airflow |
SCRAPING SOUND LEADS TO SCARY DISCOVERY
I put a cartridge in the drive to assess how the heads are positioned relative to the disk platter surface. The lower head was very very close to the disk surface, much closer than it should be. I hand rotated the spindle and heard a scraping sound, which I at first feared was the lower head scaping the disk platter surface.
However, after I removed the cartridge, I turned the platter in the cartridge and still heard the sound. When I looked underneath, I could see that the metal plate which should be spring loaded to close the cartridge, was instead collapsed inside sitting on the platter surface. That caused the scraping sound.
I quickly opened the cartridge and checked the cause. The spring loaded plate is held by plastic rivets but the degradation of old plastics (through evaporation of the plasticizers over time) makes them brittle. The rivets just gave way and let the plate fall loose. Fortunately the disk surface itself was undamaged. I can relocate the good platter into another cartridge whose platter is more scraped up.
The plate should be pressed down over the opening to keep dirt out. The drive has a plastic point that raises the plate to let air inside.
LOWER HEAD HAS BEEN BENT IN THE PAST
My inspection of the head positioning uncovered the lower head failing to rest firmly against the solid holder. The head is on springy metal that has been pre-bent so that when it is screwed onto the holder, the head is pressing down on the holder. This positions the head with a gap from the disk surface.
 |
| How the springy metal should fit on solid holder |
The mechanism that loads heads onto the spinning disk platter presses the backside of the head, forcing that spring loaded head to move off the solid holder and up against the surface. However, when the head is not loaded, it should be pulled back against the holder firmly. The lower head was not.
Looking very closely, I can see a bend in the springy metal, not the result of manufacturing. I suspect some ham-handed prior repair. I was planning on removing the head and getting the springy part returned to its intended condition.
LACK OF CONTINUITY IN LOWER HEAD WINDINGS, A FAILED CONNECTOR
One of the tests I did was to verify continuity through the erase and read/write coils of each head. The upper head was perfect but I didn't get good readings from the lower one. The worst case would be an internal break inside the heads; it is essentially impossible to find replacement heads.
I separated the connectors to isolate the failure to either the side with the cable down to the head or in the side with cable back to the disk electronics cage where I tested. Immediately I could see the problem - two of the three pins of the male connector from the disk head side were broken off in the female drive side connector. I carefully extracted them, so that I can figure out how to repair the connector and restore continuity.
REMOVING HEADS IN ORDER TO TEST OUT REST OF DRIVE
I am not yet ready to risk the heads or a cartridge by lowering the heads down onto the surface. I could block the mechanism that loads the heads, so they stay above the disk surface. The arm could move in and out, letting me check a lot of the drive out before loading heads.
However, because of the bent lower head, which is already too close to the surface, it wouldn't be prudent to move the arm in and out as there is too much risk of an impact with the lower head. I therefore chose to remove the heads from the drive, putting back only the solid holders. That way, the arm could still move back and forth but there will be no risk of damage to heads or platter surface.
VERIFYING TWO KEY MICROSWITCHES
The function of the drive depends on the correct operation of two microswitches. One is turned on when the disk arm is fully back at track zero location - called Home position. The other is turned on when the head loading solenoid pulls in; it indicates that the heads are loaded at the same time that the mechanical bits are pushing the backs of the heads onto the platter.
I used the VOM to verify a good connection and correct operation. The Home switch only activated when the carriage was pulled back to the home position. The head load switch worked properly when I moved the pivot that is pulled by the load solenoid.
WIRING UP POWER SUPPLIES TO BENCH TEST THE DRIVE
To test the drive outside of an 1130, I can use the CE switches to command movements when the drive believes that the heads are loaded on a spinning platter. This requires that I provide all the DC and AC voltages with my own power supplies.
I set up four supplies, which will provide the +3V, -3V, +6V and +48V that the drive requires. I crimped ring terminals on wires from the supplies and screwed them onto the terminal block where DC power enters the drive. I have an AC cord attached to the AC box which powers the blower motor as well as the spindle motor that spins the disk cartridge. Finally, I wired a switch with ring terminals to attach where the start switch inside the 1130 would be connected.
SPINNING UP DRIVE BUT BEGAN TO SMELL SMOKE FROM CARD CAGE
When the start switch is turned on, the drive will spin up to speed (1500 RPM) and wait for 90 seconds. This allows dust to be blown out of the cartridge and for the platter temperature to stabilize. At the end of the 90 seconds, the drive will activate the head loading solenoid, which pushes the backs of the heads down onto the spinning surface of the platter.
The drive should then switch into Ready status and accept commands issued to it. My intent is to let the head loading appear to succeed, but of course the heads were removed so they won't actually be flying. The CE switches will allow me to move the arm back and forth in single and double track steps.
The drive did spin up but after only about fifteen seconds I began to smell smoke so I shut everything down.
VOLTAGE REGULATOR CARD SEVERELY OVERHEATING
I found that the voltage regulator card in the electronics cage was the source of the smell and a resistor on it was quite toasty. This card is said to reduce the +48V down to +35V for use in read/write and access circuits, plus a highly regulated +6V for disk head bias.
I can see that the load resistor is damaged and cracked, the source of the smoke. Whether it is just a bad component or their is a short elsewhere is hard to tell without the schematics.
NEED TO SEARCH FOR SCHEMATICS - MY ALD DOES NOT MATCH THE DRIVE
However, I ran into a snag trying to move forward. On the ALDs that came with my machine the internal disk drive has a single card in slot L2 that does the voltage regulation. On the actual drive I am working with there is a double card 01352 in slots L2 and L3 instead.
I will be looking at any other drive ALDs I can find to see if any of those have the double card alternative. I wasted hours debugging the 1132 printer controller because of a difference between the VCF machine's design and the ALDs from my system, until I realized that the logic was moved to a different card slot and type of card in the VCF system. This disk drive difference, occurring at the same spot as a failure, could be equally confusing.
This is why it is so essential for those buying or keeping vintage computers to keep the ALDs and other service documentation with the machine. While I can get close using the few sets I have found online or my own, there are always situations like this.
MAINTENANCE MANUAL SUGGESTS TESTS ON POWER TRANSISTORS
The maintenance manual for the disk drive suggests that the power transistors often fail and should be checked quickly with a VOM. Because some of these are Germanium transistors, seemingly bad readings can be acceptable transistor condition. for example, a few hundred ohms resistance across the transistor can be acceptable for the Germanium type 042 transistors. However, this makes it harder to check for shorted junctions.
I even removed them from the circuit to be certain that I was measuring them without interference from other components. I used the diode testing function which showed good junction voltage drops for the silicon power transistors but only .1 volt for the Germanium 042 components. This is indeed the drop for low currents as used in a VOM. Based on that, these are likely okay.
SUGGESTED QUICK TEST WITHOUT A CARTRIDGE
The design of the drive has it attempt to retract the heads all the way to the home position (track 0) if it is turned on but not yet ready. To test this, I would remove the cartridge, manually position the arm out about halfway, then apply the DC power. It should move back to home.
No comments:
Post a Comment