Schedule October 2007
<!FONT SIZE=-1> return to main 1401 Restoration Page
go to Team Bios
Wed October 03 - general, Wed October 10 - general, Thurs October 11 - 729 tape drives, Sat October 13 - 2nd Sat. Wed October 17 - general, Wed October 24 - general, Robert is visiting N.Y., Sat October 27 - 4th Sat. Wed October 31 - general,
Wed October 03 - general
- Present were: Ron Williams, Bob Erickson, Frank King, Allen Palmer, Joe Preston, Stan Paddock, Ed Thelen.
Joe and Stan have been working with the type 026 Key Punches. Stan is currently involved with the read stations of the key punches (used in copying a card).
This was not Bob Erickson's finest day !! He pushed too hard on something and a number of punch dies popped out :-(( Again !!!
It ain't all glory ! Here is Stan cleaning off 40 years worth of grease and grime.
- In the afternoon, Frank, Stan and Ed met with Alex Bochannek about the eventual displaying of the 1401 System (IBM 1401 computer, 1402 card reader/punch, 1403 printer, 1406 extended memory, 729 magnetic tape units) and the Unit Record Equipment ( type 026 keypunches, type 083 sorters, type 077 collators, ...).
- Report here.
Wednesday October 10 - general
- Present were: Ron Williams, Bob Erickson, Allen Palmer, Joe Preston, Stan Paddock, Robert Garner, Ed Thelen.
Allen Palmer brought in a framed organization chart from "The Tabulating Machine Co." with Thoma J. Watson as President and Director. Included are the board of C-T-R-Co. All of the above companies were renamed International Business Machines" in 1924. Apparently Joerg & Gisele Ebel of Saratoga, CA gave the picture to Allen.
Bob Erickson has been trying to re-assemble the punch die and magnet assembly. It has been "tedious" yet "interesting", as in the ancient Chinese curse "may you live in 'interesting' times.
- All but the last row of electro magnet assemblies has been inserted and connected mechanically. - but this last row is stubborn :-((
We were happily doing our things, when all of a sudden - THAT SMELL - overheated lacquer, - something is way too hot.
- Allen was powering up and testing one of the recently re-assembled 729 tape drives. The Rewind Motor (left center) overheated. What was causing that?? A missing phase on the three phase motor?? Miss plugging?? - It turned out that one of the clutches (right center) was locked up (energized) while the Rewind Motor was trying to spin the tape reel. Now what??
- Alex Bochannek dropped by with a count of Unit Record (IBM card equipment) that we had been talking about at the previous "1400 Demo" meeting. Looks as though there are sufficient 083 sorters such that one can be run as part of a "unit record" demo. The name "unit record" comes from the fact that a punched card can (usually ;-)) contain all of the data for a transaction. This makes easy correction, sorting, merging, ... and was a major factor in promoting punched cards as the medium of choice rather than the much cheaper paper tape and other equipment. Also there are sufficient 077 collators :-))
The three card controller for a 729 tape exercizer has been returned.
Towards the end of the day, Tim Robinson dropped by, and we leaped at him to give us a demo of his Babbage type machines. My goodness - It all seems to grow daily !! Tim's web site
- It is fascinating to hear Tim tell about his "models" to an interested audience :-)) There is so much to know, and he has studied it so thouroughly. See this web site for an introduction on what values to use to initialize difference engines for different functions.
- This year for Vintage Computer Festival, November 3,4 at CHM, Tim is showing his Mecano based Differential Analyser - first time in several years - I think that alone is worth the price of admission. Torque amplifiers, mechanical integrators, plotters, ... :-))
Thurs October 11 - 729 tape drives,
1401 TAU Debug Status - (Jeff S., Ron C., & Bob F.)
Note: The jogger in the 1402 read output hopper #1 was not working. We were limited in the size of decks we could read.
We started the day running tape diagnostics successfully several times. The diagnostics started to fail as the room warmed (about 30 minutes without the Liebert on). The most common symptom was the records being read were missing the first few characters. (The diagnostic program writes a record, backspaces, and then reads the record.) We suspected that the 729 tape drive start/stop times needed to be adjusted.
We wrote the tape with one very long record of all bits set. Then read the tape while we probed the “GO” signal on scope CH1 (triggered on this transition) and amplified Read Data from the head in Ch2. We watched the amplitude of the Read Data signal change as the tape drive started and stopped. Both start and stop times were longer than spec.
The pictures we took of the tape start/stop data envelope are
First "Prolay_right...jpg" is a closeup of the right side prolays. The gaps I mentioned are: (measured with the prolays in the neutral position and tape threaded through)
- Start gap (spec= .003"-.005") - from the center plastic roller to the capstan roller on its right.
- Stop gap (spec= .004"-.005") - from the center plastic roller to the brake just above it.
More pictures and thumbnails are available at http://www.irhino.com/ibm1401/1401_101107.zip (Jeff Thanks for processing the pictures.)
RiseOfGo...jpg is the "start" spec showing the envelope being at full amplitude at 2.7 +-.5 msec. StartBefore.jpg is the "start" waveform after prolays current adjustment, but before gap adjustment. StartAt2mils.jpg is the "start" waveform after the gap was adjusted to .002". fallOfGo...jpg is the "stop" spec showing the envelope being at full amplitude for about 1.5 msec (coasting) and then dropping to zero in about .6 msec. Stop...jpg is the "stop" waveform after gap adjustment to .003".
Upon starting, the data amplitude took about 1 ms longer than spec. to reach 7Vp-p (specified full amplitude is 8Vp-p). Upon stopping, the data amplitude took about 6 ms longer than spec to die out. Stopping seems to have two components, coasting and braking. Both were too long by >2ms.
We checked the prolays gaps, they were all within spec. We checked the prolays current; it was very close to spec.
We boosted prolays current 10% above spec. This helped a little.
We decreased the gaps to meet the desired growth/decay waveforms. We were able to achieve the correct waveform at about 50% the specified gap sizes. However, this resulted in tape stretching and breakage. We backed off the gaps to the spec minimums, but the data amplitude growth/decay is still significantly longer than spec.
Our best guess is that the start/stop problems are due to hardening of the rubber on the drive capstans and brake wheel. After 40 years, the rubber is probably not as sticky as it once was.
If you have ideas on how to further isolate the root cause of bug, of have repair ideas please reply. It’s clear the start/stop times are a problem, but I have no idea why the errors seem to be sensitive to a warm room.
Another note: At the room temperature where the tape drives started failing, the 1402 reader also became very unreliable.
Tape Analyzer Software
Jeff has made progress with the GUI imagery. He left it running in a browser window on Ed's notebook. Take a look.
I am seeing some signs of life from the Windows kernel driver, but the "Blue Screen of Death" is occurring regularly. I have enough coded to send and receive messages. A hook exists for the tape drive emulation functionality, but the code is not written yet. I suspect that it will take at least a month to work through the current set of bugs/issues, then a few more to make it emulate a drive..
Saturday October 13 - 2nd Saturday
- Present: Ron Williams, Bob Erickson, Judith Haemmerle, Dan McInnis. Jeff Stutzman, in afternoon.
- This report courtesy of Ron Williams.
- Bob says he and Judith good progress on 513 summary punch, maybe apply power next Wednesday
- Ron and Dan worked on debugging Dan's switch based IO exercizer, Demo Assurance Program. Pretty much got it working.
- At the lunch for the "2nd Saturday Volunteers", Judith presented a slide show of progress at the new Milpitas warehouse.
- Jeff Stutzman worked on tape drives.
- This last Thursday, Ron Crane, Jeff Stutzman, Bob Feretich worked on the 729 tape drives. Their log says the on-line tape drive works well when cold. It appears to have a start/stop problem when warm. That problem recurred today.
- About 20/30 seconds after the 1401 was turned "ON", the -12 volt power supply tripped. This occasional (once a month?) problem is getting higher priority now that the system is getting more stable. A more full discussion of this problem is on August 8,2007. Basically that power supply is marginally overloaded. Circuit breakers made now have a different (smaller) form factor :-(( What is reasonable to do?
- On the 1402 card punch, column 54 picks up extra punches if another hole already punched in that column swapping driver card with col 74 made problem went away :-((
- On the 1402 card reader, the joggler in normal read stacker is not moving -
Wed October 17 - general
- Present were: Ron Williams, Bill Flora, Frank King, Bob Erickson,Glenn Lea, Joe Preston, Stan Paddock, Ed Thelen
The 513 summary punch, which had a badly damaged punch mechanism from some mis-adventure before arriving at CHM, was powered-up today for the first time. No one dared? punch a card, but the card path was sufficiently well timed for correct card passage. :-))
Unfortunately, this cylindrical slotted rubber thing which is supposed to help push the cards is hard and not at all flexible - how to get a replacement?
Bill Flora trying to re-assemble a tape exerciser controller. It seems that some vital information was scrambled during dis-assembly. Bill is not his normal smiling self.
Frank King and Joe Preston are re-furbishing a spare 026 print mechanism for installation on the 026 from Illinois, which came with out a print mechanism.
Granddaughter Rose is not pulling her fair share during this walk Would you believe that the handle of her lunch box, with new toy inside, is still firmly clasped in her right hand?
Wed October 24 - generalSome 729 tape capstan rubber info:
From Grant Saviers
From Allen Palmer
Allen loaned me a "like new" stop capstan and I measured the durometer of the the rubber.
On a Shore A scale it measured 90 and on a Shore D scale it measured 50. Although the A to D scale comparison is not reliable, these are within the +/- 5 durometer typical of rubber specifications.
The other question is what type of rubber and my bet after a little research that it is neoprene (Dupont trademark) alias for chloroprene. It is unlikely from the little history search I did that EPDM or polyurethanes were common in the late 1950's although the dates of introduction of these materials is hard to determine. Perhaps some one of our mail list has more data.
In any case, neoprene is available from McMaster from Shore A 30 to 80 step 10. I suspect the capstan has hardened some over time so we might try the A80 material. Machining this stuff is difficult and not fun, but can be done. Grinding is the way to finish the capstan OD's. Very messy.
I also advise starting with known property rubber as using hammer heads etc. is a complete random walk.
From Lynn Liebschutz
Lynn dropped 4 of the rebuilt stop- capstans. I will bring them in Thursday & if you are in the area please stop in to look at them. Allen j Palmer
Grant's comments are well received.
Since I have provided 5 pieces to Allen ( and I still have one completed with me ) from the mallet material and have suffered through the odorous fab process, let's give these a try. The material is about the correct Durometer or maybe sightly softer and it is reasonably robust. I still have some spools and I can retire these with Buna N, Neoprene or an equivalent depending on the results of live testing.
408 - 717 - 8137
- Present were: Ron Williams, Bill Flora, Frank King, Bob Erickson, Glenn Lea, Joe Preston, Stan Paddock, Robert Garner, Ed Thelen. Later Dr. Mike Williams, author, very special ex-head curator of CHM, now president of the IEEE Computer Society, came for a visit. He was in town briefly and (among other things) wanted to see how we were doing :-))
We started the day in snappy fashion - "new" old IBM 083 sorter arrived, and "old" (rusty German) 083 got moved into the ramp area to make room. This machine looks excellent - we went through a quick capacitor reforming exercize and started sorting cards - Frank reports it was at least 98% perfect, and didn't even jam once. (Frank says the sealed bearings didn't need checking for grease.) (We never did get the other rusty one to quit jamming.)
Bob Erickson and Glenn Lea continued bringing up the completely overhauled IBM 513 summary punch that had been brought in badly damaged internally. Now to assure everything correctly timed -
Mike Williams, Robert Garner, Ron Williams and Dag Spicer watching 729 Tape Drive in action. Ron, Dag, Mike looking at BigPrint printout of Mike Williams visited CHM October 24,2007/ Joe Preston and Frank King in background.
Mike Williams examining a pre-IBM organization chart of "Computing- Tabulating- Recording Company (C-T-R)," with T. J. Watson as its head :-))
- This is on loan to Allen Palmer from one of his neighbors.
- The 1401 system successfully reproduced a 200 card deck BigPrint for backup. This was the first time the 1401 system successfully reproduced a non-trivial deck of cards :-))
There were two "Smart Cars" on exhibit in the CHM lobby. A few years ago, Frank King negotiated for one - but couldn't agree on a price.
- Looks as though Frank is still lusting for one ;-)) He wants these pictures.
Robert Garner is in the state of New York to visit this 1401 - Oct. 26/27, 2007, images by owner's family
Sat October 27 - 4th Sat
- Bob Erickson, Ron Williams, Judith Haemmerle & Tim Coslet - This report by Ron Williams
- no tape people - Ron tried to run DFT tests and own test on 729 tape drive - everything failed.
- Bob & Judith worked on 513 and ?077? plug board programming.
Wed October 31 - generalNotes from meeting - October 24, 2007 Wednesday afternoon
Preparation for VCF demo series, starting 3:00 Saturday Nov 3, Present Bill Worthington, Stan Paddock, Jim Somers, Ed Thelen (After the meeting, Bill Worthington and Stan Paddock discovered they were unavailable for the VCF 1401 demos) Discussion of rights to IBM images - 029 key punch, ... Vintage Computer Festival ------------------------- open house ? VCF is the one day we really try to control access to Visible Storage - because these guys really like to touch things !! Have the public stay behind the 1401 room rail - Have a red band with stanchions at the top of the ramp - How many people can see from the ramp? 15? Jim will make a sign saying when the presentations start likely 3:00, 3:30, 4:00, 4:30 how to terminate a session, we can't turn out the lights as in the PDP-1 demos last year Ideas - sign name as come in, we can key punch names off list feed to cards into the BigPrint program, return cards and printouts as souvenirs (Ron Williams knows how to run the program) - Stan has put the CHM logo and web site on the backs of IBM cards Stan has say 300 already in a drawer near the good Liebert - relief for presenters after 1/2 hour - 15 minutes for talk - 15 minutes Q&A then time for a mini break, and/or other presenter - Jim Somers will work on sign on the outside of the 1401 room. - Robert has camcorder / Video camera? record one of the sessions, to present on monitor on outside later? Jim S can get volunteer to operate. Between tours? - Jim will contact Sellam about above schedule
Go to November 2007
go to Team Bios