return to main page

1402 Maintenance
Rust on the IBM 1402 Reader/Punch

Table of Contents

Ron Williams' - Seven Steps to a Happy Card Reader via Stan Paddock
Ron Williams said" the best way to fix the reader is to do a 7 step process:"
  1. Check the dynamic timer to make sure the clutch engagement point is 315 degrees.
  2. Check to picker knives to insure the card trailing edge is under the feed at 72 degrees.
  3. Check and adjust the RL cams as required
  4. Check and adjust the RC cams as requires.
  5. Check and adjust the optical timer as required
  6. Check and adjust the check brush block as required.
  7. Check and adjust the read brush block as required.
Ron, Frank and Iggy spent most of the day following the above list.
Several of the items above needed adjustment.

At the end of the day, the infamous "cards of dubious value" were fed into the machine.
A few cards were kicked out for reader stops
- - - - (cards had previous damage so they would not feed) and several validity checks.
Some of these turned to be fed in upside down (12 edge first) and some were valid validity checks.
When these cards were pulled out, we read cards all day long.

As of 5:00 PM last night, the Connecticut machine is the machine to use for demonstrations.
I do not know the status of the tape drives on the machine.

Rust on the IBM 1402 Reader/Punch - Report # 1 by Grant Saviers
The large photos below (accessed by clicking on the thumb-nail sized photo) are 2/3 the edge length of the originals supplied by Grant and considerably reduced in .jpg resolution. For originials (about 350 k bytes each) contact Grant Saviers or me ( giving picture name.

All text and photos below by Grant Saviers.


1. 1402 CPPA orig cond - as removed from 1402 card punch assembly. Note: the shaft end bearings are backwards


2. 1402 CPPA shaft - note rust, which is cosmetic, not functional. Also note red marker "witness" marks on the arms and shaft - somebody has had this assembly apart before us!


3. 1402 CPPA arms - same red "witness" marks, probably not accurate enough and can be easily removed. A fine scratch would be better


4. 1402 CPPA pick shoe - lots of cosmetic rust, the other side is the business end with carbide inserts and looks ok. These will be measured in phase 2.


5. 1402 CPPA cam follower - lots of card dust amalgamated with oil & grease. The cam follower bearings are hard to turn.


6. 1402 CPPA cam follower brg - dry brushed to remove gunk, surface corrosion and grime is visible


7. 1402 CPPA cam bearing clip - fortunately, the shaft is not pressed in - is cleaning this area an infrequent PM?


8. 1402 CPPA cam arm dirty - note the packed gunk which jams the bearing


9. 1402 CPPA cam arm spacer - some scraping out of gunk reveals a shaft spacer (small washer)


10. 1402 CPPA arm bearing - this bearing is a cam follower; note the thickness of the outer race, designed to prevent its distortion from the cam pressure. The bearing is MRC double sealed type R-B/8-F/P-F/P-P/R-8/B (each digit is most likely character/alternate) or most likely RBFFP8. The id is stamped into the rubber seals on both sides of the bearing and difficult to read, even under moderate magnification. The dimensions are .187" id x .625" od x .195" width. Since this is a sealed bearing I used a little steel wool to clean up the outer race surface. It appears to be ok functionally and turns very easily.


11. 1402 CPPA arm shaft - the cam follower shaft shows an unusual wear pattern of repeating diamonds. No dimensional differences could be detected with a dial caliper.


12. 1402 CPPA end bearing - backwards on the shaft again. Note the screwdriver "oops" (ding) probably inflicted in the disassembly. A better tool would have prevented this. Should be easily fixed with a little filing.


13. 1402 CPPA end bearing detail 1 - dirty, but mostly gunk and a little cosmetic rust. The bearing type is SKF double shielded, marked 393214, dimensions .375" id x .875" od x .250"? width (I could not measure it directly without removal). The shields are the pressed in type, not removable without destruction. The bearings (both ends) turn freely.


14. 1402 CPPA end bearing detail 2 - other side of detail 1. Not sure of the engineering function of the rubber damper/seal/??. Anyway, it is coated with a different kind of gunk, semi-transparent dry grease, assembly adhesive, or ?


15. 1402 CPPA end bearing parts - careful prying and massaging by hand separated the rubber inner part from the two sleeves. The rubber part is a soft durometer and in good shape, no cracks, tears etc. It went back together easily. Again, what is the function?? I don't want to clean it until some light on this issue and an idea of the composition for solvent resistance, etc.


16. 1402 CPPA cam arm cleaned - a soft brush and some paint thinner made it look new. Note one bearing spacer for the cam follower bearing. This shaft had one, the other shaft two spacers, one each side of bearing. Again, SOMEBODY HAS BEEN HERE BEFORE ME! There should be two spacers, since the bearing inner race is the same width as the outer race. The spacer keeps the outer race from rubbing on the side of the arm. A slight wear pattern is visible on the arm side without a spacer. The spacer is 0.020" thick x .187" id x .375" od, probably steel as there is some surface corrosion present. I need to make a replacement for the missing spacer and install it.


17. 1402 CPPA detail cleaned - I didn't loosen any of the cam follower or pick shoe arms as I saw no need. With a new spacer I think these parts will work ok and the bearings seem to be in very good shape.


18. 1402 CPPA clean - Further action on this part is a. cosmetic (except needed spacer) and optional and b. measure the pick shoe steps for wear


19. 1402 CPPA tools - nothing fancy hand tools, a magnifying visor and dial caliper. The follower bearing shaft did not need a press to remove.


Overall conclusions and lessons learned:

  1. Interestingly, IBM chose inch standard bearings for this unit. Metric standards are much more common. Further research is needed to verify replacements are available. The bearings seem to be in good shape, the main functional problem was gunk jamming them.

  2. More care/right tools need to be used for disassembly to prevent damage (end bearing holder screwdriver ding)

  3. Even though the team indicated this level of disassembly was not done in the field, this unit has been apart before and a spacer was lost. Note the witness marks, certainly not the accuracy needed by the factory! Lesson - make no assumptions about what we have - look carefully, think about the mechanical design, and proceed with caution and copious notes. (and patience!)

  4. Further investigation is needed to understand the end bearing assembly functions.

  5. A standard and index for identifying photos might be a good idea.

  6. Time involved in this project 1.5 hours work, incl photos. 2.0 hours for report. Camera: Nikon Coolpix 880, 3.3 megapixel, used XGA normal resolution.

Next phase:

  • Remove the pick shoes and measure the wear on the carbide (not easy).

  • Clean end bearing assembly when understood better.

  • Decide if a light cosmetic cleanup should be done.

  • As usual, comments, feedback, and all suggestions/help welcome.

Philosophy - Why relay logic, not SMS cards, in the IBM 1402 Reader/Punch? July 2015
On Jul 17, 2015, Claunch,Carl < > wrote:
Hi Guy [ Fedorkow ]

I will indeed be there on Wednesday and happy to spend some time with you poking around at interesting parts of the system.

I will show you what I mean about the relay logic and how cumbersome the documentation is.

Almost all the control unit functionality for the 1402 is implemented in relay logic locally in the reader, with the interface to the 1401 being little more than the brush wires and a very sparse set of signals. When we put a scope on an interface line, or more usually just on the logic line inside the 1401 upstream of the interface, it tells us little more than that the 1401 side is okay. When there are problems in the reader, the diagnostic work is almost totally internal to the unit and unrelated to the interface signals.

When IBM 'refreshed' the 1402 to sell it with their 360 mainframe series as the 2540 reader/punch, they replaced the relay based logic for the control unit with some SLT card logic that sat inside the peripheral. While the kept all the SMS generation logic for the 1403 printer when that was attached to 360 systems, they decided it was worth the cost and delay to replace the relay logic. Would have been much cheaper to keep it like it is and stick some SMS inside a 2821 for interfacing, but someone wisely decided to ditch the failure prone and tough to diagnose guts of the 1402.

I bet that if IBM had known they would sell as many 1401 systems as they did, the team that converted the 088 collator into the 1402 would have put the extra effort into a transistorized state machine, but with the low forecast, the engineering cost of replacing 088 relay based logic didn't seem warranted


On 7/20/2015, Robert Garner wrote:
I bet that if IBM had known they would sell as many 1401 systems as they did, the team that converted the 088 collator into the 1402 would have put the extra effort into a transistorized state machine, but with the low forecast, the engineering cost of replacing 088 relay based logic didn't seem warranted.

The engineering team from Chuck Branscomb, Fran Underwood, Jim Ingram, et al, on down believed early on that the 1401 was going to be wildly successful. Engineering development wasn't constrained by central planning's low market projections -- that disconnect occurred later on...)

I've cc'd 1401 "founders" to ask for their thoughts on why the 088/1402's relay logic wasn't changed to SMS control for the 1401 system.

- Robert

From: Charles Branscomb < >
Subject: Re: Why doesn't the 1402 have SMS control circuits?
Date: July 26, 2015 7:13:27 PM PDT
To: Robert Garner < >
I frankly am surprised that the 1402 had relay logic controls rather than SMS. But I must admit that I do not recall any discussion on this subject. My surprise is based upon the major push we had on reliability which should have led to SMS logic. I do recall discussion and concern for punched card hole sensing reliability in the 1402 but again I do not recall any significant technology changes compared to the 088.


Here's the response from 1401 designer Jud McCarthy:
Robert: The amount of logic in the 1402 did not warrant the additional SMS power supplies and an additional special input transformer required to support SMS. The alternative of providing these voltages from the 1401 was prohibitive, as the voltage drop would be to great due to the distance between the two units. Anyway, the real complicated control logic for the 1402 is handled with SMS logic within the 1402 attachment logic located in the 1401. ---- Regards ---- Jud