Schedule July 2008
return to main 1401 Restoration Page
go to Team Bios
Wed July 02 - general,
Wed July 9 - general, Sat July 12 - CHM Volunteer Work Day,
Wed July 16 - general,
Wed July 23 - general, Sat July 26 - CHM Volunteer Work Day,
Wed July 30 - general,
Wed July 2 - Generalfrom Stan Paddock, TuesdayDaily Reports, Team Bios and Current Status Report (March 2008 ), Proposal to Acquire & Restore the "Connecticut 1401
Robert Garner says
I just spoke with Dennis Cassar.
- The UPS for the CHM servers is now up and running.
- The electrician looked at our hardware and is going to give us our "60AMP 208 three phase" Thursday of this week. They are going to use the connector we have provided.
- We should be able to apply power to the CT 1401 next week or Saturday if people come in.
Remember: Ron is assuming that us "younger" engineers (i.e, you, Ron Crane, Bob Feretich, and I) will bring up the 1401 this time. We're in for it.... ;-)
- Present were: Ron Williams, Bob Erickson, Bill Flora, Frank King, Allen Palmer, Glen Lea, Joe Preston, Judith Haemmerle, Robert Garner and Ed Thelen.
It was apparent that someone was catagorizing some of the CT spare cards. Judith seemed to deflect questions -
Allen Palmer has just "cleaned" the badly pitted contacts on this 729 column sense (bellows) switch - and is re-adjusting it.
I was out on some kind of raid into the museum. Chris Garcia called out that he had something to show me. I identified it as a 2 gun electrostatic deflection CRT - and figured that Bob Erickson (who serviced Williams Tube memories on the IBM 701) would get a kick out of it. Would you go for a DuMont 5SP7 which is used in a Tektronics 555?
Bob Erickson's grandson "JD" (right) came in with a friend. We gave them a "special" tour, including a comprehensive tour of "Visible Storage" :-)) JD isn't embarrased, he is working on a golf course ;-))
Frank King struggling to connect very stiff cable to 1403 printer and another on the other side - these contain the hammer driver currents from the controller in the 1401 The power cable will be installed in the center after CPU successful power up.
Wed July 09 - generalA BIG day :-)) - Power up on Connecticut system :-)) Lunch at GOOGLE ;-))
- Present were: Ron Williams, Bob Erickson, Bill Flora, Frank King, Glen Lea, Joe Preston, Judith Haemmerle, Stan Paddock, Robert Garner and Ed Thelen.
- As "the rest of us" trickled in Bill Flora and Stan Paddock were finishing connecting the power cable into Connecticut 1402 Reader/Punch and removing the belt from the card reader motor. The reason for that is to verify the sequence of the three phase power will rotate this motor - and by inference all motors - in the correct direction. There is a 50 percent chance that this will be correct, as in general, phase identification is not carried into the building (I don't think.) This is not a big problem, if the three phase motors go in the wrong direction, just reverse two phases and all will be well :-))
While working with the 1402 belts, Bill Flora notice that the belt that enables the card punch to delivers cards into the center pocket was missing, and another broke with only slight stress. So - do we have belts? Lots, lets hope the correct ones :-))
OK - lets start at the wall panel. The bottom three ganged breakers are for the German 1401 - They need to be rated at 90 amps because of the inefficiencies in the 60 Hz to 50 Hz power convert. Just above are the new three ganged (if one trips, all trip) 60 amp circuit breakers for the Connecticut system installed by the electrician ( the normal circuit for an IBM 1401 with tapes. )
The cable from the wall panel terminates in the right angle socket sitting on the underfloor. The cable from the 1401 Card Reader/Punch (the main power connection for the 1401 system) is sitting on the raised floor. Husky looking hardware eh?
Excitement, Ron's lit utility lamp indicates the 1401 now has utility power, like to plug in oscilloscopes - but the rest of the power did not come up. :-| We are scrambling about looking for blown fuzzes, open interlocks, anything ... I hear that Bill Flora leaned on some relay and 48 volt relay power came up and stayed up - dirt in contacts??
How is this for a hopeful look on Robert Garner? 48 volt relay power is up, and the push buttons ready to trigger the power up sequence relays. Well - the power came up briefly, but a -12 volt power supply tripped :-(( What to do? Think, Think - Well, nothing smoked, lets try again, and this time power came up and stayed up :-)))))
Well, at least we have a clock - Now that the machine is alive once more - here is the running time meter :-| Unfortunately, the 1401 is not working correctly yet - a fetch from memory yields all bits on regardless of what you think you put in. :-(( Work to do - maybe much work -
Shortly thereafter, Glenn and Stan started moving the least good looking (decomposed rubber) 729 mod 2 tape units to invisible storage. Life is so much sweeter when you have proper moving equipment :-))
Frank is REALLY eager to get his 1403 printer going - all the shafts rotate, there is sufficient hydraulic fluid and other fluids - Lets get this baby going!! He started out by disabling the 60 volts to the hammer drivers in case there is a problem in the printer driver clock circuit. Too long a pulse to the hammers can blow out soldered-in fuses in groups of 20 - no fun!!
1403 motors going in the correct direction - we assured correct phasing way back in the 1402, above - But a very worrisome racket!! Very noticeable!! Here is Frank using a large screwdriver as a stethoscope, trying to determine the source of the racket. The noise seemed to be coming from the output shaft of the 1/3 horsepower 3 phase motor driving the hydraulic "transmission".
Frank called in Bob Erickson for a second opinion - The racket slowly subsided, and after about 10 minutes, the machine almost sounded normal :-)) Of course, we haven't put a print chain in yet -
About then, John Hollar, (white shirt), the new CEO of Computer History Museum appeared, with Paula Jabloner, Karen Kroslowitz, and Judy Strebel of the museum staff. By then, it was past our appointed time (12:45) for a lunch at GOOGLE, arranged by Judith - but Robert wanted to get a nice group picture of us - It seemed to take for ever to set up the tripod, ..., and all of the usual delays - but the result, at long hungry last, was very nice, allowing for the subject matter and fatigue factor ;-))
And off to GOOGLE
Judith says our host, Marcin Wichary, is a young CHM Docent and avid photographer (took the best images of the Babbage Machine at CHM by waiting for night - no outside light distractions). Wouldn't you just know - I didn't get a good image of Marcin. We had a GREAT lunch - I started out with three desserts and some cookies ;-)) Then a tour of GOOGLE, now in the last headquarters of Silicon Graphics. A classic boom, build fancy headquarters, and bust -
GREAT tour Marcin, thanks much.
After the tour, including many little food places - Marcin says GOOGLE wants free food near all employees - a laundromat, and interesting hallways - here is a mock-up of Space Ship 1 winner of the X-Prize. Marcin returned us to a Slurpy bar, where we topped-off our food reserves. and had a great view of T-Rex ;-))
Sat July 12 - CHM Volunteer Work Day
- Present were: Ron Williams, Tim Coslet, Stan Paddock, Robert Garner.
Report from Stan Paddock.
[Ed Thelen was AWOL, attending a memorial service.]
- I came in early(so the others would not have to breath the dust) and used compressed air to clean every drawer in the CT 1401. Found peanut shells and acorn shells inside.
- I reseated all of the SMS cards in all of the drawers.
- When we powered up the machine, more of it worked!
- We also found out that of the 16,000 locations, only the bottom 2,000 locations don't work.
- Ron entered a 'branch to self' program in upper memory and it worked!
- There were several lights burned out on the control panel. Ron and Robert replaced them.
- There was a hardware error with the transfer between the A register and the B register. Bit 4 would stick high in the B register. Robert, with Ron's help, found and replaced the faulty card. This was Robert's first IBM 1401 hardware bug fix!
- I left at 3:00 and don't know what happened after that.
When we get the lower memory to work right, we can start trying the 1402 and 1403. Exciting.
Wed July 16 - General
After hours, I went to visit Tim Coslet at Sequoia Hospital in Redwood City. Tim had kind of fallen apart, seriously, and Judith had taken him to Kaiser, then to Sequoia on Monday. Tim said that Mike Cheponis had visited him Tuesday. I had to wait half an hour after visiting hours opened Wednesday because "Sandy" and another ex-coworker from WatkinsJohnson got in ahead of me. Tim and I talked of many things until visiting hours closed, and I was asked to leave ;-))
- Present were: Ron Williams, Bob Erickson, Bill Flora, Frank King, Joe Preston, Judith Haemmerle, Stan Paddock, Robert Garner and Ed Thelen. Glen Lea was AWOL - was he talking about Hawaii??
Ron W. Bill F. Bob E. found the 1402 card reader brushes were very frazzled. Resulting in the following error report - (top right). Test card read is below Better brushes right, gave better report - the bottom 3 of the report card. More work yet -
- Frank thought that the 1403 printer fan that cools the fins on the hydraulic transmission didn't turn last Wednesday, and in any case could stand a good cleaning. Frank said it was connected between two of the three phases - about 208 volts. Regarding electric motors as proper "boy's toys" Ed quickly volunteered to have a look :-)) Connecting it up to the 120 volts from the work bench - the motor started gently and ran quietly - at full rated voltage it probably would really whir :-))
Ed disconnected the fan from the whole blower assembly (not shown), removed the fan rotor. Stan and Ed cleaned the assembly and rotor. The dust was really oily, condensed from near by oils and ribbon? We picked off and rubbed off most of the grime then went to the janitor's closet wash sink and blasted off the rest with warm water. Perfection :-)) We noticed the insert broken - but the mounting screw worked OK because its threads pressed into the steel plate :-|
Frank King's son brought in some visitors. Stan, in the red suspenders, is acting as tour guide and operator - About the same time, there was a going away party for Allison Akbay, who is going to a position in an art museum at Stanford University. We will miss her cheery smile and practical good sense -
Frank's words "Driver for paper stacker Drive Mechanism. Runs off of the hydraulic motor. Noise when first started printer (I think)" Joe's pencil points at a grove in the shaft. A seized bearing had been there. The shaft rotated in the seized bearing - bearing and shaft will be replaced.
Most of memory now works: Robert replaced a defective core address driver to fix the inaccessibly of even 2K blocks. - but "odd" gaps remain. The notes on the card indicate that the zone 7,000 to 7,999 and zone 11,000 to 11,999 has the following defects - 7090 through 7099, 7290 through 7299, ... and a similar pattern starting at 11090 through 11099. Ron Williams notes that a similar problem in the German machine was fixed by Bob Erickson doing some high class soldering of core plane memory wiring.
As we were leaving the building, some delivery man asked if this was 1401 Shoreline. We asked "Didn't you see the "1401" on the building? We are working on a machine called 1401." - But the 1401 seems to be missing from the front of "our" building :-((
Wed July 23 - general
Today was apparently Tim Coslet's last full day in the Bay Area. Tim had been very helpful in restoring the CHM IBM 1620 to functionality. Over the past 4 years Tim, in spite of handicaps, has repaired most of the 100+ bad SMS cards removed from the "German" IBM 1401. Tim supplied his own design in-circuit tester and his own scope for these efforts. His efforts are greatly appreciated and he will be missed (like we have no one to replace his capability.)
Tim was hospitalized about a week ago (not the first time) and figures he will never be employed again. Judith Haemmerle has treated Tim like family, helping clean his apartment, caring for his cats, settle accounts, get his (?COBRA?) medical insurance transferred for service in Montana, packing Tim's items for the journey, ...
Thursday, Tim's brother will start driving Tim back to his family in Montana -
Thank you Tim - Fare Well -
- Present were: Ron Williams, Bob Erickson, Bill Flora, Frank King, Allen Palmer, Glen Lea, Joe Preston, Judith Haemmerle, Stan Paddock, Robert Garner and Ed Thelen.
Allen reported that this summer's rocketeering was not totally successful - something about the second stage deciding to go horizontal rather than vertical :-(
- Bill Flora and Glenn Lea were VERY SERIOUS about working on the CT 1402 Card Reader/Punch - no frivolity here !! WORK !!
I suppose if you must, you must - It is a free country - mostly ;-))
- Frank King brought in a specialized stethoscope.
We are all familiar with the medical variety - The medical person puts some cold thing on your bare chest and says "Take a deep breath - hold it - hold it some more " - well maybe not the last phrase ;-))
Well, dig this !! A stethoscope for machinery :-)) No more trying to couple a transmission to your ear with a screwdriver, as a fake stethoscope
Frank wants Ed to listen for the various bearings in the 1403 transmission, and the hydraulic fluid rushing about - this takes some familiarization !!
- Stan Paddock and Bill Worthington had a meeting planning docent presentations of the 1401 and unit record equipment. The goal is to simplify the operation of the demo decks - remove the switch options - which just got added. Back to KISS, Keep I Simple Stupid :-)))
- Allen Palmer spent his day working to bring the second 729 Mod 5 tape drive up to power-on status. It is "almost ready" :-)) (It's power supplies were powered up and checked a year or so ago.
Allen Palmer is pointing to the many relays in a 729 Mod 2 that were replaced by transistor logic on SMS cards in the 729 Mod 5 - giving considerably greater reliability in the Mod 5. The white arrow points to a mechanical timer that helps control the timing of the LOAD and UNLOAD operations. This is replaced by electronic timers in the Mod 5.
- There was considerable discussion about "firing up" the first of the CT 729 Mod 2 tape drives. There are numerous electrolytic capacitors in the power supplies for these drives. But the CT system had been used much more recently than the "DE" system - and all of the DE power supplies came up well - not enough capacitor reforming to be "dramatic" - like heat and explode the electrolytics. And the electrolytics in theh CT 1401 showed no reforming at all -
So "we" decided to save time and the technical risk of removing and replacing equipment and wiring - by just powering up the 1st CT 729 Mod 2 tape drive. We applied power, and no dynamics :-))
The voltages seemed reasonable, but the voltages did not appear on the power supply terminal strips in the same positions as shown on the schematics for that drive :-| Much casting about for why the mis-match.
Here is Ed, probing about near the bus bars, while Joe Preston is paging about in the schematics hoping to find some correlation between physical reality and the schematics :-| Actually all the voltages, 140, 70, 48, +- 12, +- 6 were quite close to nominal. The minus 6 was actually 5.75, while being restored to 6.00, it showed signs of a dirty/noisy potentiometer (very jumpy when moved). One of the two fans in the area did not function although it rotated easily when turned manually.
These are a few of the air filters that were removed from the 1401 and 1406, vacuumed, and today replaced
Sat July 26 - CHM Volunteer Work Day
- Present were: Bob Erickson, Ron Williams, Robert Garner.
Report from Robert Garner
- Ron and Bob chased a couple of 1401 CPU problems. Ron noticed that B execute cycle was staying on longer than it should after Clear Storage instructions. We found the troublesome card and replaced it (using a German spare card this time as there was none in Darien inventory) and problem went away.
A test loop of multiple Clear Storage instructions works fine in single step, but not in Run mode (existing problem). We traced that bug for a while on the storage scope (triggering on the error check), narrowing the problem down to several candidate gates. As it was after 4 pm, and we had places to be, time to stop. We should have it licked next session! ;-)
- During this time, Bob compared the schematics of the German 1406 against the American 1406 and was fun to be around, as always.
- Judith dropped by to keep us company and brought in her home-made, scrumptious, to-die-for chocolate cake. (Her own birthday cake!)
Wed July 30 - general
- Present were: Ron Williams, Bob Erickson, Bill Flora, Frank King, Glen Lea, Joe Preston, Judith Haemmerle, Stan Paddock, Robert Garner and Ed Thelen. Glenn is talking about going to Hawaii next week - he didn't mention bringing "the rest of us" with him. ;-))
- 729 Mod 2 fan problem from last week - to cool circuit cards the rear -
Isn't that a cute little three phase motor - most of us had never seen one that small - We of course wanted to take it apart, but the screws wouldn't budge - sealed with Glyptol? A quick raid to our spare 729s (classified as "educational" as opposed to the museum's "archival") provided a new fan assembly with two working fans :-)) - "Repair" complete :-))
Joe and Stan measuring the resistances between the three wires going to the little three phase motor, two pairs of readings were about 600 ohms, the third was about 2 megohms :-((
Later in the afternoon, this next problem was solved by replacing another card - and we can "sort-of" read a card into memory - about half the columns are correct - Now, is this a problem with the character decode, or with the brushes in the 1402 reader/punch - or both - or neither - stay tuned for next week's episode ;-)) Will the villain drive the hero off the cliff??
There is a certain loss of dignity while sitting on the floor trying to bug-shoot the lower 1401 gates - Here is Robert, caught by the candid camera - checking if a changed circuit card fixed a card read problem. Executing a card read command would clear all of memory, continuously until halted. Fortunately this fixed the continuous clearing, but the card read command now causes another fault - not quite so fatal.
Speaking of cards, here are 4 aging cheat sheets pasted under the front "cover" of the 1401
And speaking of cheat sheets, imagine you have to work with say 150 of these - Ron Williams made a series of "aids" like this
- Of course, everything cannot be serious - gotta have some fun -
And Judith brought in a room air filter - it was getting noisy. She and Bob Erickson disassembled it, but I lost track of those goings on.
Stan Paddock brought in this "boy's toy", with no electrical diagram or instructions.
- Back to business ;-| A couple of weeks ago "we" cleaned a 1403 printer chain. Time to mount it in the print chain assembly. Frank stressed that even a hair between the print slugs could cause a break in the chain plastic binding all the slugs together- "And there ain't no more." And there was the business of the 1403 chain screw that got loose. When it gets loose, its head contacts another screw head and causes trouble - like stripping threads on one or more screws. We talked of "bottom thread tap" which can be a normal thread tap, but with the tapered part cut off so the threads can be cut completely and the screw threaded all the way to the bottom. Apparently such an operation has already taken place in Frank's garage.
Now to put the chain on the two pulleys, with the "9" slug at the correct index point on one of the pulleys.
Now to put the chain on the two pulleys, with the "9" slug at the correct index point on one of the pulleys.
- After considerable discussion "we" (Ron Williams ;-) decided that we should merge:
the spare SMS cards we got from Connecticut, already sorted by Judith and someone with the 6 drawers of German SMS cards
The current overflow from the drawers went into these boxes. We can probably fix one more shelf in the set of drawers to hold cards. There were no 729 cards in the German inventory - but fortunately there are some 729 specific SMS cards from Connecticut
- Ron Williams brought in some of his photos of folks and the installation.
Go to August 2008
go to Team Bios