(none)
  

The University of Iowa's DEC PDP-8

Problem List

Part of the UI-8 pages
by Douglas W. Jones
THE UNIVERSITY OF IOWA Department of Computer Science

Contents, By System Component

Contents, By Bug Number


Introduction

This is a list of all of the problems encountered with the University of Iowa's PDP-8 computer. Problems are numbered sequentially in the order of discovery. Click on any thumbnail image to see the full-sized image.


Bug 1: Broken handles on switch register

  thumbnail image     thumbnail image
Front panelHandle repair
The plastic switch handles at the left end of the row of switches are chipped, as if they collided with something. They feel loose, and it is probable that the plastic ears that hinge some of these switch handles to their switches have broken. Some of the switches may also need rebuilding, as we did with Bug 44.


Bug 2: Loose cards on side of core memory

thumbnail image  
Loose cards

There are 8 card-edge connectors facing to the front on the side of the black box holding the core memory. All but one of the circuit cards that should be plugged into these connectors have fallen out.


Bug 3: Decayed foam strips in cabinet sides

  thumbnail image  
Right side
The front posts of the racks have aluminum channels holding plastic foam weatherstrip against the cabinet skins. Whatever its purpose, the foam in these channels has decayed. It is brittle, and it crumbles when disturbed. We will have to replace it with something less problematic. The problem is similar to Bug 11, involving a foam strip in an aluminum channel, but the dimensions are different.


Bug 4: Decayed ribbon cables between half-backplanes

  thumbnail image  
Cable Assembly
The ribbon cable assemblies that connect the left and right half-backplanes of the machine are severely decayed. The insulating layer on the two flex-print cable sthat make up each cable assembly have turned to goo. We should find an appropriate replacement material.


Bug 5: Stability of the computer rack

  thumbnail image     thumbnail image  
Stability
When the front panel with its lights and switches and the entire computer is slid forward on its rack slides, the power supply and rack are just barely balanced. The risk of the whole thing falling over is significant.

Bug 6: Sticky latch for computer half backplanes

  thumbnail image  
The latch
The latch that holds the half-backplanes of the computer in their closed positions is very difficult to turn. The key is in the latch in the photo.

The solution is to grasp the frames for the two half backplanes near the lock with one hand and squeeze them together while turning the key with the other hand. This releases the friction on the latch making it easy to turn.


Bug 7: Broken paper-tape reader on Teletype

  thumbnail image     thumbnail image  
Broken reader
  thumbnail image     thumbnail image  
Repair in progress
The paper-tape reader on the Teletype was broken. The cast ears that attached the frame of the paper tape reader to the frame of the Teletype broke off, probably when someone used paper-tape reader as a handle to move the Teletype. Fortunately, these could be replaced with home-made brackets. Also, the control lever on the paper-tape reader was broken, probably in the same accident.


Bug 8: Teletype power cord and data cables

  thumbnail image     thumbnail image  
Before and after
The electrical wires at the back of the Teletype were pulled out of their strain reliefs and needed to be redone before anyone even thinks about plugging it in. Also, there seems to be an extra data cable.


Bug 9: Teletype print hammer

  thumbnail image     thumbnail image  
The old and replaced hammer

The rubber head on the Teletype print hammer was shot. For many years, the standard fix among Teletype collectors has been to replace the rubber with a piece of vinyl tubing, but recently, a serious collector has manufactured new replacement hammer heads.


Bug 10: Reform electrolytic capacitors

  thumbnail image     thumbnail image  
Power supplies

The electrolytic capacitors in the power supplies need reforming.


Bug 11: Decayed foam gasket in cabinet doors

  thumbnail image  
Door gasket

The foam gaskets that seal the vertical edges of the rear plenum door of each cabinet need renewal. The problem is similar to Bug 3, involving a foam strip in an aluminum channel, but the dimensions are different. Each door has two gaskets, one on the latch side that is easy to see, and one on the hinge side that is most apparent only when the door is open and the hinge-side skin of the rack is removed.


Bug 12: Rebuild the Tally reader cable

  thumbnail image  
Tally reader cable
Where is the Tally reader interface that the card-edge connector on the reader cable connects to? And, what slot in the reader interface? As of the identification of this bug, the cable has been plugged into the back of the Tally reader. Mapping of the 34D Scope display interface revealed that this backplane segment at the bottom of the ADC rack includes the high speed reader interface. Unfortunately, on further investigation, it turned out that the cable, while it may have originally served this reader, had been extensively modified and must be rebuilt.


Bug 13: Make 110V power connection for Tally Reader

  thumbnail image  
Dangling Jones plug
The 2-pin Jones plug on the back of the connector to the Tally paper tape reader appears to be intended to bring power to the reader's drive motor. Verify this, and then make the appropriate secondary cable to bring 110V power from the power distribution panel on the ADC rack to this plug.


Bug 14: Repair or replace Teletype data cable

  thumbnail image  
Cable repair
The Teletype data cable was cut between the W070 connector to the computer and the 12-pin Jones plug at the back of the Teletype. The bad parts near the ends of the cable were cut out and the salvaged remainder was reinstalled with a 6-pin Jones plug at the point where the cable had been cut. The resulting cable is shorter than the original, but with the jones plug at the point of the cut, an extension cable can easily be added.


Bug 15: Inappropriate Teletype data connector

  thumbnail image  
Inappropriate plug
The 12-pin Jones plug in the Teletype cable was not really appropriate. By 1970, DEC was using an 8-pin in-line Mate-n-Lok connector at this point, but (as evidenced by the 12-pin connector here), earlier DEC users who wanted a plug in their cables frequently used Jones plugs. 6-pin Jones plugs are still available, so we installed one.


Bug 16: Check and repair 110V rack wiring

  thumbnail image  
Cut power wires
The power distribution wiring harness between the ADC Rack and the PDP-8 rack was cut, close to the terminal strip at the right rear corner of the rack. The power cabling for the both racks will need careful checking.


Bug 17: Teletype cover broken

  thumbnail image     thumbnail image  
Broken cover, start repair
  thumbnail image     thumbnail image  
Repair completed
The transparent plastic insert in the Teletype cover was cracked and entirely broken free from its mounting. It appears that it was thermally riveted to the inside of the cover. The whole cover was filthy, and had a broken mounting ear. There are also cracks and broken mounting fingers in other parts of the cover.


Bug 18: Missing ADC plenum panel and screws

  thumbnail image   thumbnail image
ADC Plenum door
The rear plenum door of the ADC rack is missing one panel (the one below the ADC interface panel) and about 20 screws. This admits dust and degrades the effectiveness of the rack's cooling fan. A somewhat battered replacement panel has been located -- original DEC manufacture, formerly part of PDP-9 serial number 120.


Bug 19: Inventory and clean the computer backplanes

  thumbnail image     thumbnail image  
Preliminary backplane maps
We need to inventory the boards in both backplanes, then remove them and clean both the boards and the backplanes, and then re-insert the boards where they came from. Wherever there are discrepancies between the boards and the maintenance manuals, we will need to do more to figure out how to put things right, suspecting boards might have been pulled and replugged incorrectly over the years when this machine was not operational.


Bug 20: Repair and re-install PDP-8 power supply

  thumbnail image     thumbnail image  
The supply out of the rack
  thumbnail image  
Opened up for capacitor access
To allow reforming of the capacitors in the PDP-8 power supply (see Bug 10), the supply needed to be removed from the rack and partially disassembled.


Bug 21: Bad foam rubber pad under Teletype typing unit

  thumbnail image     thumbnail image  
Decayed foamNew foam
The foam rubber pad under the typing unit was decayed and crumbled when disturbed. This is Teletype part number 181138, pad, pages 25 (drawing) and 32 (parts list) in Parts Model 32 and 33 Page Printer Set, Teletype Corp. Bulletin 1184B, Nov. 1964, change 5.


Bug 22: Bad vibration isolators in Teletype

  thumbnail image  
Decayed rubber
The rubber grommets that serve as vibration isolators were decayed and crumbled when disturbed. This is part number 181109, mount, vibration, pages 25 (drawing) and 32 (parts list) in Parts Model 32 and 33 Page Printer Set, Teletype Corp. Bulletin 1184B, Nov. 1964, change 5.


Bug 23: Missing Teletype sub-base mounting bolt

  thumbnail image  
Base re-mounted
One teletype sub-base mounting bolt was missing. and two more appear to have been inappropriate substitutes.


Bug 24: Missing Teletype call control unit screws

Two call control unit mounting screws were missing.


Bug 25: Clean, and de-gum the Teletype

  thumbnail image  
The print head
The Teletype mechanism was dirty and the old oil and grease has congealed to an adhesive gum. For example, the print head carriage bearings were so thorougly gummed that they would not rotate at all before several soakings in WD-40. Cleaning these parts requirede nearly complete disassembly of the machine.


Bug 26: Re-assemble the Teletype

  thumbnail image     thumbnail image  
Major subassemblies
Cleaning and de-gumming the Teletype required large-scale disassembly, so we need to put it back together.


Bug 27: Teletype Keys are severely crazed

  thumbnail image     thumbnail image  
The keyboard, before and after
The Teletype keyboard keys were in very bad shape. The outer grey layer of many keys has almost shattered into multiple fragments that are not strongly attached to the white core of each key. Zoom in on the photo for details. Can we buy or make new keys? Consider the keys here that were made with laser-engraved acrylic filled with contrasting paint, then laser cut and laminated to build up the required depth. We eventually found a source of new-old-stock keys and replaced the worst of the bad ones.


Bug 28: Clean and reassemble the Tally reader

  thumbnail image     thumbnail image  
Tally capstan and take-up reel drive
The Tally 424 paper-tape reader mechanism needs to be properly lubricated. This will require partial disassembly and reassembly of the reader.


Bug 29: Replace the Tally reader motor run capacitor

  thumbnail image     thumbnail image  
The capacitor and replacement
The Tally paper-tape reader motor uses a 3µF paper and oil capacitor. Given the time period when this was made, it is almost certainly filled with PCB oil and reports from others suggest that failures are not uncommon and occasionally explosive. Therefore, it must be replaced.


Bug 30: Replace the Tally reader motor leads

  thumbnail image  
The motor stator
The power leads to the Tally paper-tape reader motor were in bad conditon and needed replacement.


Bug 31: Repair and Reconfigure ADC Rack

  thumbnail image  
Installing a rivnut
When the Psychology department mounted the Tally reader and magnetic tape interface in spare spots in the ADC rack, they drilled out some of the rivnuts DEC had installed and drilled (carelessly) some new holes. Before we remount the Tally reader and any other peripherals in this rack, we should replace the missing rivnuts. We will need to replace other missing parts of this rack, and if we change the configuration of equipment in this rack, we may need to add new rivnuts.


Bug 32: Reassemble Type 779 power supply

  thumbnail image  
The supply
In order to reform the capacitors in the Type 779 power supply, it had to be extracted from the ADC rack.


Bug 33: Test Type 708 power supply

  thumbnail image     thumbnail image  
Test load 10V at 10A 300mV
With the capacitors reformed and the power supply reassembled, it is time to test it. This involves constructing a dummy load and measuring all of the outputs under this load.


Bug 34: Reverse engineer the Tally reader interface

The Tally paper tape reader interface is a subsystem of the Type 34D Oscilloscope Display backplane. DEC designed this interface as the Type 750C High Speed Reader, but that interface appears to be designed for a photoelectric reader. There is evidence of tinkering in the backplane here, so we need to thoroughly understand what was done to make the DEC interface work with the Tally reader.


Bug 35: Fix Marginal Check supply panel meter

  thumbnail image  
Trimmer added
The Marginal Check power supply panel meter reads about 10% high compared to the VOM used in power supply testing.


Bug 36: Fix the Inhibit and Read/Write supplies

The Inhibit and Read/Write supplies do not work. These are regulated supplies with outputs that only connect when both voltages are within spec. A relay driver performs the logical and function.


Bug 37: Get or build Flip-Chip Module Extenders

  thumbnail image  
W982 Board
The Inhibit and Read/Write supplies do not work. These are regulated supplies with outputs that only connect when both voltages are within spec. A relay driver performs the logical and function.


Bug 38: Repair and mount Teletype tape-reader supply

  thumbnail image  
The Supply
The power supply for the Teletype mounts either on the Call Control Unit or on in the Teletype pedestal. Our supply was missing, but we obtained a broken one from eBay for a pittance. Unfortunately, it is broken, and it came without a mounting bracket.


Bug 39: Repair or replace broken Teletype screw

  thumbnail image
Old and new screws
Teletype part-number 182891 is a very unusual screw. Unfortunately, the fat screw head invites overtightening. Fortunately, the huge screw head means that this screw can be repaired.


Bug 40: Repair G007 sense amplifier

  thumbnail image   thumbnail image
The broken board The repair
The sense amplifier in backplane slot MB30 was damaged during the backplane inventory and cleaning (see Bug 19). The repair was essentially straightforward.


Bug 41: Repair Teletype keyboard cover

  thumbnail image  
Lugs replaced with screws
The cover over the Teletype keyboard was supported and aligned on 3 mounting lugs, all of which have broken off.


Bug 42: Rebuild Teletype keyboard contact block

  thumbnail image     thumbnail image  
OriginalRebuilt


Bug 43: Test PDP-8

With the power supply tested and the rack wiring fixed, we can turn on the computer. Now we must see if it works!


Bug 44: Repair PDP-8 marginal check switches

  thumbnail image     thumbnail image  
Switch parts during cleaning
After turning on the computer, we found that power was not reaching all parts of the PDP-8 backplane. The problem turned out to lie in the switches that allow switching each row of the backplane from the primary power supply to the variable voltage marginal-test power supply. Repair or replacement of these switches is in order.


Bug 45: Clean PDP-8 backplane fans

  thumbnail image  
Cleaning the fans
There was considerable dust on the bottoms of the 3 fans at the bottom of each half backplane of the PDP-8, so we removed the fan blades from their motors, cleaned everything, and then opened the motor hubs and oiled them.


Bug 46: Reassemble PDP-8 front panel

Front panel
  thumbnail image  
To deal with Bug 1 and Bug 47, we had to disassemble the PDP-8 front panel. It will need reassembly.


Bug 47: Fix burnt-out front panel bulbs

  thumbnail image   thumbnail image  
The lamps.All lamps lit!
When we began testing the PDP-8 (Bug 43) we found that many of the bulbs on the front panel did not light, although other evidence suggested that the bulbs ought to light.


Bug 48: Test Teletype

  thumbnail image   thumbnail image
First test11 weeks later
Having reassembled the Teletype, it is time to see how it works.


Bug 49: Repair R211 board

  thumbnail image  
Testing the R211
The R211 boards are double-height boards. Each hold one bit of the memory address, program counter and memory data registers, plus the multiplexer logic for loading those registers.


Bug 50: Reverse engineer the R211 board

  thumbnail image     thumbnail image  
The R211 H board.
  thumbnail image  
The R211 H reverse engineered.
The R211 boards are extraordinarily complex double-height boards, where relating components on the board to the schematic diagram is not easy.

The image showing the result of the reverse engineering shows both sides of the board, superimposed (with traces on the solder-side shaded red). The text gives part numbers; those prefixed with equals signs (=) reference the 1965 schematic of the R211 H; those without reference the 1966 schematic of the R211 J. Markings on vias relate them to pins on the card edge connector or to the voltages distributed on that circuit trace.


Bug 51: Build Flip-Chip board test rig

  thumbnail image     thumbnail image  
Test fixtureLogic Probes
Testing boards from the PDP-8 is easier if they can be tested outside the PDP-8 backplane. This required building a test rig and a logic test probe.


Bug 52: Adjust the Teletype

  thumbnail image  
Reader stop latch
  thumbnail image  
Adjusting platen height
There are numerous adjustments that must be made after disassembling and reassembling the Teletype.


Bug 53: Missing felt washers in Teletype punch

  thumbnail image  
Missing felt washers
The outboard end of one of the pivot shafts in the Teletype is missing several felt washers.


Bug 54: Loose code bar bracket in Teletype

  thumbnail image  
Code bar bracket
The bracket that serves as an alignment comb at the left end of the code bars that serve as the main mechanical data bus in the teletype was not attached. Floating loose on the code bars, it did not seem to be doing any harm, but it should be attached.


Bug 55: Broken Teletype carriage-return lobe plate

  thumbnail image   thumbnail image
BrokenReplaced
The carriage return lobe plate that mounts on the left side of the carriage was broken. This prevents the spacing mechanism from working after a carriage return because without it, the return lever stays latched, disengaging the spacing pawls.


Bug 56: Teletype control-character linkage engagement

Most of the control-character actuating levers was not engaged properly with the corresponding linkages, and the corresponding function levers were not properly engaged with their function pawls.


Bug 57: Teletype bell rings with every character

During reassembly of the Teletype, the automatic codebar was inserted in the wrong slot in the codebar assembly, causing the function bell to ring with every character typed.


Bug 58: Teletype bit 4 problem

After enough of the Teletype was repaired that the typing mechanism began working reasonably reliably, there were misprints that focused on incorrect values of data bit 4.


Bug 59: Teletype platen needs reconditioning

The rubber of the teletype platen is very hard, limiting the print quality. Is there some way to recondition it?


Bug 60: Teletype stand is missing its back

The back panel of the teletype stand is missing. Can we manufacture a replica of the original?


Bug 61: Debug PDP-8 PC register

When you increment the PC, carry does not propagate from PC bit 2 to bit 1. This is most obvious when you hit RUN with the memory disabled so all fetches get opcode zero (AND). The PC lights up as 00X XXX XXX XXX, where bulbs marked X are on at half intensity because they are flickering between 0 and 1.


Bug 62: Fix Teletype linefeed mechanism

  thumbnail image     thumbnail image  
The screwdriver Adjusting LF drive link
We broke the linefeed mechanism while adjusting the platen height.


Bug 63: Fix Teletype reader-run circuit

  thumbnail image  
The problem spring
The paper-tape reader only runs if the clutch-trip magnet's armature at the rear of the Teletype is held down manually. This is evidence that there's something wrong in the wiring of the reader-run circuit.