Repair sessions
From May to July 2015, I repaired those six PC05 (and cannibalized an incomplete 7th one). I stopped any other RetroComputing activity, since working in an assembly-line manner is highly efficient. When repairing, most time is spent in understanding the machine under your finger tips and becoming familiar with the documentation, this part has only to be done once for all PC05's together.
On the backside, it was very boring at the end. And often I felt like in a private little hell, where a handfull of problems reappeared in a seemingly endless loop.
Standard procedures
For each PC05, I performed these steps:
- outside cleaning
- removal and reseting of all FASTON terminal cable plugs
- opening and refastening all electrical conducting screw joints (transistors, rectifiers, capacitor terminals and ground connections)
- washing the sticky grease out of the Roytron puncher with highly fluid penetrating oil.
- calibration procedure according to document "PC04-PC05 paper-tape reader-punch (feed-hole strobed) 5th printing (May 1973,DEC-00-PC0A-D)"
A special Flip-Chip extender.
For debugging and repairing, I needed to operate some electronic modules outside the card cage. I build a special flexible flip chip extender from an flip chip backplane socket.
DEC Field Service had no need for such a tool: they did not repair, but just located the bad modules and replaced them.
After having used that extender the first time, I could not imagine how I ever could work without it!
The Big Failure List
This is the list of all failures I found and fixed. Interestingly, every device had several severe errors. Since a PC05 surely was taken out of service on the first unrepaired failure, follow-on damages must be attributed to the conditions in the warehouse where they were stored for years.
PC05 #1
- on one M040 stepper motor driver the screw connection to a power transistor's case was bad: it had 30 Ohms instead of 0 Ohm, so the motor was not working correct.
- on a M7050C the Quad-FlipFlop IC Motorola MC4015 was dead. I replaced it by a rewrapped TTL 74175.
- in the puncher two control rods were sticking together by remains of ancient grease. So sometimes two holes were punched were only one was intended.
- cable to PC11 defective
PC05 #2
- the potentiometer on M715 was teared off. M715 is the most outside module in the card cage and exposed to physical damage.
- the replacement potentiometer caused a strange oscillation involving a feed back loop over M715 and M7050. It was built in with extreme setting, which caused the motor control circuit to produce a strange stepping signal to the motor: it was constantly running at half-speed after manual FEED had been pressed once.
- the tape depressor assembly of the reader head was broken into pieces. It was replaced from spare device, and needed precise adjustment afterwards. Uneven pressure onto the tape caused high punctual tape wear. Paper fibers gathered in data holes of the test tape loop resulting in read errors.
- the puncher motor axis was blocked by sticky grease.
- on G918 photo amplifier one transistor 2N2904 in the threshold regulator was ripped off the board, invisible to the eye.
- M7050: TTL inverter E17 dead.
- G918: five more transistors were replaced, as they strangely quit working after a few minutes of operating time.
- on the photo transistor array the cell for hole 1 was dead. It was replaced against a photo array, were the unused "end of tape" hole 9 was dead.
- the TRIAC on the puncher motor control TRIAC-ASSY was dead.
- in the puncher a return spring for the control rod of hole 2 was broken
- M710 "Punch Control": TTL latch E1 replaced.
This PC05 had almost all possible failures. It could have been used in a Field Service school for training purposes!
PC05 #3
- the whole light bulb assembly was missing, it was taken from the spare PC05.
- M710 Punch Control: TTL monoflop E5 replaced.
- Roytron Puncher: main drive shaft broken. This means total loss of the puncher, device is a pure reader now.
- Reader: rotation angle of transport wheel adjusted.
PC05 #4
- chassis front was broken and bent.
- Puncher: the clutch solenoid for the dented tape wheel was blocked by ancient grease.
- reader cable to PC11 seemed OK first, but had an open "Error" signal wire.
- burned resistors on punch solenoid drivers M044 replaced by 2W type.
- the reader light bulb had an bad contact inside the bulb, brightness was unstable.
- one stepper motor driver M040 was dead: TTL chip 7401 replaced.
- on the photo transistor array the "end-of-tape" hole 9 was dead. Replaced with spare device, and re-used the defective array on PC05 #2.
PC05 #5
- the puncher motor axis was blocked by sticky grease.
- a capacitor on M715 was ripped off the board.
- the power regulator board had a solder break around one FASTON tab.
- reader lamp had faint brightness fluctuations, causing read errors. Reason was a very resistent bad contact to chassis ground.
- Threshold potentiometer on G918 was dirty and gave unreliable values as it was moved.
- TTL data latch 7474 on M7050 "Reader Control" had an instable output.
Here reader data bit #2 had intermittent failures, caused by three different sources: amplifier threshold, lamp brightness and a defective data latch! This took a while to recognize ...
PC05 #6
- puncher motor made running.
- I tried to operate this PC05 with disassembled backplane and made a short cut of the 5V supply voltage. The big rectifier in the power supply died and was replaced with the one of a spare PC05.
- burned resistors on punch solenoid drivers M044 replaced by 2W type.
- Reader: data bits not read in correctly in all ZPCA tests. Rotation angle of stepper motor to chassis and angle of transport wheel to motor adjusted.
AND FINALLY: The PC05 of my computer club
- one reader stepper motor driver M040 dead.
- one punch driver M044 dead. This module could be moved to another backplane slot were the dead output was not used.
- reader light bulb replaced.
- rotation angle of the reader tape wheel corrected.
All these PC05's pass XXDP diagnostic "ZPCA" now!
