Log of August 2006 events

Back to Home Page

Tuesday, August 15,
Tuesday, August 29, folks measured RAMAC electromechanical parameters :-))

---

Tuesday, August 15
Meeting - at CHM at 2 PM on Tuesday, August 15

• Present were: Al Hoagland, Dick Oswald, Dave Bennet, Joe Feng, Jack Grogan, Ed Thelen, and Herky Hanson.

• This was Herky's first meeting. He isn't in town very much. Dave Bennet drew the proposed block diagram on the white board and we discussed it.
    proposed system block diagram, from July 12.

• We had a discussion on protecting the RAMAC from hazards. Including potential problems, such as
  an air line or wire breaks/ driver fails, we or the sequencer thinks a detent is engauged and it isn't
  a head piston sticks and we think the head is off-disk
  ...
  Probably much of the protection should be built into the servo, as the sequencer might not react intime?
  Jack Grogan to be "Range Safety Officer"? ;-))

  There was a discussion on error reporting, what was the sequence of events leading to the safety/protection halt? FPGA's can have memory, and report the last n commands and conditions.

  IBM and Hitachi are donating help and cash to the 50th anniversary on September 13.

  Dave Bennet reported that about 20 RAMAC heads had been donated to the project. (Prior to that, we had only the heads installed in the machine.) Apparently there are differences between the the upper and lower heads, we now have 4 spares of one kind and many of the other.)

  Dick Oswald's drawings of June 20th were discussed. Dick said that the earlier desired chopping of the clutch current was impractical due to the clutch L/R parameters. He is going to the original DC drive.

  As the meeting broke up, (another meeting, an air bag problem,) I realized that we should have pictures and snapped the remaining folks
  

  Dave Bennet

  Joe Feng

  Herky Hanson

  Dick Oswald

  Jack Grogan

  While showing Herky the RAMAC, a remark was made that the reason the bottom lubricant had not been changed was that the hex fitting (to take an Allen wrench) of the plug had been stripped and was unuseable.

  There must be some method of rotating that/those? plug(s). And maybe restoring the easy rotation -

  Dave Bennet, Joe Feng and Ed Thelen took Herky Hanson on a tour of the museum, including back stage. It is amazing how much folks remember. They look at two identical looking units and say "that is an IBM ..40 and that is an IBM ..50. You can tell because the type with the removable pack has that yellow indicator."

  ---

  
  Tuesday, August 29, 2:00 PM folks came to measure RAMAC electromechanical parameters :-))

  and several to talk about the presentation at CHM September 12, 2006

  Present were Al Hoagland, Dave Bennet, Dick Oswald, "Herky" Hanson, Joe Feng, Ed Thelen, with "visitor" Bill Bernard who is helping plan the September 12th presentation at CHM.

  There is another presentation on September 13 at the IDEMA show.

  A little prior to the gathering Al Hoagland expressed interest in showing a magnetic drum memory. These were very popular from about 1950 until almost entirely replaced by core memory and RAMAC related disk memories about 1958. Al, with a smile, wished to point out that drum memories had fixed heads and were more liable to disk and head damage than the air bearing and flying heads of the RAMAC and subsequent random access disks. (The closeness of the head to the surface was a human adjustment, and varied due to differential expansion of different parts, probably of different temperatures when running.) The following images are of CHM objects.

  An ERA drum, showing signs of a tough life :-((

  This is an English Electric drum, believed to be from a British DUCE machine.

  And further detail

  This is current (vertical) vs. time (horizontal) display of the effects of an inductance in series with a resistance when a step voltage is applied. The time constant of the curve is L/R where L is in Henries, R in Ohms. This is useful to Dick Oswald in designing the RAMAC clutch driving circuit. A schematic of the test circuit is here.

  Same test setup (at 40 volts) operating a RAMAC air solenoid. Note the odd wave form caused by the movement of the iron armature in the magnetic field. With this we can estimate the delay time of opening the air flow. The right image is both applying and removing the voltage from a RAMAC air solenoid. Removing the voltage (and current source) caused an inductive "kick" (the vertical stripe way off scale, and an indication of the time the air solenoid stops supplying air - it now allows air to bleed out of the controlled air circuit.

  Also, using the current source above, and Dave Bennet operating his spring scale, "we" determined the following driving action of the RAMAC clutch on the RAMAC arm going horizontally from out to in: Current through clutch in Milliamps Force necessary to keep the arm moving, friction in pounds, a tough unsteady measurement, +- 20%? 0 1.2 300 1.5 500 3 600 4 700 > 6.5 (We restrained the clutch drive motor from turning, and measured the arm force necessary to slip the clutch) (Dick Oswald figured the above was good enough. The resistance of that clutch was 316 ohms. Assuming that "we" will be driving the clutches at least 700 ma, the instantaneous resistive power dissipation in the clutch ( P=I2*R ) will be at least 150 watts. (This does not include slippage heating and hystersis and eddy current heating.) Good thing there are two clutches, and by the nature of the RAMAC seek operation used less than 50% each :-)) I imagine that after a few minutes of continuous random seeks, the clutches will be noticeably warm.

  The Three Musketeers - Dave Bennet, Dick Oswald and Herky Hanson ;-)) Joe Feng had set up the storage scope, shown us how to retrigger it, and is now talking with Al Hoagland.

  Joe Feng, left, talking with Al Hoagland. Don't they look serious! They are talking about how to present the RAMAC at the September 12th show - serious business! - Joe *REALLY* wants Pat Connolly to operate Pat Connolly's RAMAC moving program and equipment. !! - Pat is starting a new job, and may not be available?

  Bill Bernard (who is currently an IBM Senior Engineer, Intellectual Property, Licensing, Storage) has huddled extensively with Al Hoagland about the show at CHM Sept 12th.

  This track position indication potentiometer has a problem. Two of the seven taps are open, and one end is open. See here for further details. This makes life much more difficult for very tight servo position control. We would *really* like this fixed - but are afraid of doing more damage than good if we try!! Using Dave Dial's flashlight (CHM), Ron Williams' dental mirror (1401 Restoration) and Joe Feng's good eyes, "we" read the name plate as: " Markite 25 K Ohm 10%, Type 2917, Serial# A7356 " Sensor Systems, L.L.C. Petersburg, Fl. - 1-800-688-2181 Pat Grippe - repairs manager - http://www.vsensors.com/rotary_pot.htm claims to now own the Markite Potentiometers, Inc nameplate. http://www.vsensors.com/images/datasheets/model-78-305.htm mentions taps. Potentiomenter: shaft diameter = 0.25", body diameter = 1.325", Gear diameter (measured within teeth) = 2.1" Arm: total travel 6 1/8", travel on tracks = 5.0", 10 detents/inch Other: Center post in stack = 12", actuator cable diameter = 0.060"

---

If you have comments or suggestions, Send e-mail to Ed Thelen

Back to Home Page