Wednesday, December 22, 2010

Attendees: Bill Newman,  Bob Erickson, Douglas Martin,  George Ahearn,  Joe Preston, Ron Williams, Stan Paddock, Robert Gardner and Allen Palmer

Visitors: none today

Allen Palmer re-joined us after a long term absence. 

Joe Preston continued his work on the third tape drive of the German Machine. He misses Glenn Lea's help.

The rest of us continued to work on the card reading problem with the Connecticut machine. After checking the 20 volt, the 30 volt and the differential power supplies to no avail, we looked at the voltages at the read brushes.

The roller is suppose to ground the brushes when a brush falls through a hole in a punched card.

If more that 60 columns have the same punch in the same row, it is not able to do this. We have seen this on a scope with the advancing width cards.

Bill Newman is looking for replacement transistors for the ones in the 1402 that may be failing.

We ran the same test on the German machine and did not seethe failure pattern.

While I cannot say this is the problem, we will know in two weeks when we get back and try the fix.

Stan

Friday, December 17, 2010

Attendees: George Ahearn, Stan Paddock

With open questions, George Ahearn and Stan paddock decided to go into the CHM and try to answer the pending questions.

Questions we had were:

1.     What happens if the temperature sensing card reports the correct temperature?

2.     What happens if we run the cards starting with the 80 punches first instead of last?

3.     What happens when the punches go from right instead of left to right?

When we entered the room it was very cold. Way below 70 degrees F.
We ran the PTEST program and go the same results as before.
We turned the cards around for case 2 and 3 and got errors until the number of columns went bellow the thresholds as before.

We pulled the temperature card and soldered all sensors closed.  This told the machine that the temperature was less than 70 degrees F.
We ran PTEST again and got just about the same results as before.
We pulled the card (indicating that the temperature was over 95 degrees F  and got significantly more errors. 


Following the ALDs for the temperature sensors, they control a bias to the memory current controller circuit.
While the temperature sensors are broken, they are not responsible for the errors we are seeing. However, the change in the errors with the cold and hot said we are on the right track.
Next week we will inspect the current control circuit and see if we can find the problem there.


We have found some small devices online that will open and close as the ones in the IBM 1401 are suppose to do. We will being buying a batch of these and update the cards we have in both machines so they work properly. 


While we had not used up the time Ann Ahearn had allocated to us, we decided to see if see could start trying to find out what was wrong with the German machine. Much to our surprise, it seemed to work fine.


We ran the PTEST program on the German machine with no reported errors.


Bill Flora said that he could not get the 'ripple read' program to run on the Connecticut machine.


We tried it on the German machine and got the same results.
We ran the cards through the PTEST program and 5 cards were reported with errors.


Next week, we will use the IBM 513 to produce a new card deck for the test.

Wednesday, December 15, 2010

Attendees: Bill Newman, Bill Worthington, Bill Flora, Bob Erickson, Douglas Martin, Ed Thelen, Frank King, George Ahearn, Glenn Lea, Joe Preston, Ron Williams, Stan Paddock

Visitor(s): Robert Garner had two more visitors and Don Luke had one.

Joe Preston continues to work on the last failing 729 of the German system. He is making good progress. Now the tape drive waits until a tape is installed before it tries to load the tape.

During a demonstration on the German machine, the 1403 printer dropped a column. Frank King came to the rescue and replaced one of the 66 hammer driver cards. He commented how dirty that card was. 

Stan pulled all of the hammer driver cards and went outside to brush the dirt off all of them. 

When Stan installed the hammer driver cards back in, the German machine decided to stop running programs. As time was running late, we had to leave the machine as it was and we would attack it next week.

 

Work on the card reading problem with the Connecticut machine, we took a suggestion from Jud McCarthy in Boca Raton Florida. We created a card deck of 80 cards. The first card had a ‘G’ punched in column one. The second card had a ‘G’ punched in column one and two. The 80^th card had a ‘G’ punched in all 80 columns. A program (PTEST) was written that read cards with the I/O check switch off. It would print the data and add an ‘E’ in print position 82 if a card read error was sensed.

If the card had 63 or less ‘G’s, there was no errors. Starting at card 64, more errors and columns started to be missing data.

Another deck of cards were punched but using an “*”. The ‘*’ has three punches per column instead of two like the ‘G’.

When this deck was run, the errors started after the 25^th card. The errors were sensitive to the number of absolute punches in a card.

 George Ahearn suggested the problem might be temperature sensitive. We pulled the temperature sensing card over the main memory stack that contained the three bi-metallic thermostatic sensors. The first is specified to be closed until the temperature reaches 70 degree F. The second opens at 82 degrees F and the last one opens at 90 degree F.

Under testing, the first bi-metallic thermostatic sensor worked as designed. The second (hotter) and third (hottest) failed open. At this time, we had run out of time for that day.

Questions we had were:

1.     What happens if the temperature sensing card reports the correct temperature?

2.     What happens if we run the cards starting with the 80 punches first instead of last?

3.     What happens when the punches go from right instead of left to right?

Wednesday, December 8, 2010

Attendees:  Bill Flora, Bill Newman, Bob Erickson, , Ed Thelen, Frank King, Joe Preston, Judith Haemmerle, Robert Garner, Ron Crane, Ron Williams, Stan Paddock AND George Ahearn

Visitor(s): Robert Garner invited Martin Haeberli, a colleague of his from Xerox days (early 80's).

Joe Preston continues to work on the last failing 729 of the German system.

Frank took a set of brake and two clutches from a spare 729 and re-filled the brake with the required amount of power.  Replacing the brake on the right hand tape shaft looks like a major undertaking.

Ed Thelen was checking out the new display area when he found a interesting gentleman who also was checking things out. Ed struck up a conversation with him and he turned out to be Dennis Van Sickle of Van Sickle & Rolleri. They have the contract to do the new display area. He is an interesting guy, said he flew in from Pennsylvania to review the progress to date.

December 1, 2010

Attendees: Ron Williams, Bob Erickson, Robert Garner, George Ahearn, Stan Paddock, Glenn Lea, Frank King, Don Luke, Douglas Martin, Bill Newman, Joe Preston, Ed Thelen, Bill Worthington, Ron Crane and Bill Worthington

Glenn Lea and Joe Preston continue to work on the last failing 729 of the German system.
With some technical help from other members, they found that the neon bulbs that have stopped working can be given a second life by reversing the polarity.

Ed Thelen had decided to work with Doug Martin and clean the dirt off the boards in the German 1401.
Ron Williams mentioned that the timing signals of the German Machine have not been checked for some time.
Ed thought that was important and jumped right on it.
Now the German machine is having a problem operating.

George Ahearn brought in his external garage door keypad that would kill a new battery in 2 hours. That is just the sort of problem Ron Crane, Doug Martin and Bill Newman like to sink their teeth into.
After some checkout, Ron Crane said "I think I know what it is" and disappeared down the hall with the board in hand. He had gone into the men's room and washed the circuit board. When he returned with the board dried, it worked fine. He had noticed some corrosion around one of the transistors!
Good Catch Ron.


The Connecticut 1401 is having a problem reading cards again. Last week Bill Newman, Ron Crane and Stan Paddock read up on how the reading circuits work. Without Ron Williams present, we came up with a number of possibilities that were all wrong. But it was fun.
Today, Ron Williams, Stan Paddock, Doug Martin and Bill Newman worked on the problem some more.
Ron Williams showed us how to switch two connections so the first set of brushes is switched with the second.
Doug Martin entered his first program into memory that prints out cards that fail.
This showed that the problem is not associated with just one set of brushes.
We will be working on it again next week.