IBM Electronics Circuit Analysis Program - ECAP

[Richard Stofer]

Does anybody have the source for ECAP in machine readable form?  I have the manual with the listing but I would like to avoid typing a couple of hundred pages of code.

 

Thanks

Richard

 

[Dave G4UGM]

When I did CSMP I got about 80% of it via OCR from scans, but I guess ECAP is larger…..

 

Dave Wade

G4UGM

--
You received this message because you are subscribed to the Google Groups "IBM1130" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Carl Claunch]

Pretty sure I have it - machine readable but as a long sequence of cardboard rectangles with holes punched in Hollerith encoding, nothing more easily transferable at the moment. Once I can read and transfer card files you are welcome to a copy. 

Carl

[Michael Short]

I have a copy of ECAP which was obtained from the DECUS library

and modified by Western Michigan University. I have it running under

MUSIC so it might work on the 1130. Let me know if you would want to 

try this copy. It is 2640 lines of FORTRAN.

--

[James Field]

Lazy bones! I have learned a lot about another coder's style and insight by typing his/hers code.
Comments are the most revealing.   I remember a comment in APL XM6. "loop until brown on top"!


You could scan the code pages optically and learn a bit about ocr.


Jim

--

[Dave G4UGM]

Which Fortran as you using on Music. I would like to try it on VM/370 so would like a copy, but I guess I will also need a manual..

 

Dave Wade

[Michael Short]

I am going to have to retract my offer for source code. I had worked on the code

sometime ago and I ran into syntax problems and didn't get it to compile. DEC had 

some interesting syntax in their FORTRAN.

[John McKee]

How bad could it be?  Could you show an example?

John McKee

[John R Pierce]

On 12/17/2014 12:48 PM, John McKee wrote:
> How bad could it be? Could you show an example?

DEC Fortran had a LOT of extensions, I remember from my VAX 11 daze...



--
john r pierce 37N 122W
somewhere on the middle of the left coast

[Michael Short]

Here are some statements that got flagged by the base MUSIC compiler:

      COMMON /LANG/MODEL(0/15,125),IPLACE(6),ISTORE(6),ERRCHK,NEXT     

     1,ITAB(20),ITRACK(20),NEXADS                                      

      DIMENSION ICTL(0/20),IHELP(11)                                   

                                                                       

-->   DATA (ICTL(I),I=0,20)/15,'B','L','H','E','C','D','A','M','N','S  

     1','T',' ','      ','*','R','X','X','X','X','X'/                  

IF(IDVI.NE.'TTY') GOTO 99

 DO 59 J=75,I,-1

Some of the errors are literals and a negative trending DO loop index

The COMMON and DATA are the ones that may require more a rework.

[John McKee]

Interesting.  Been a long time since I worked with 1130 FORTRAN.  I sqaw the IF error - since logical IF was only available if EMU FORTRAN was used.  I thought the DO was allowed.  I am really fuzzy on COMMON and DATA.  At least what you have shown is not as truly odd as some of the stuff I have seen in DEC FORTRAN - but other adventures might also exist.  Almost comes across as a challenge.

John McKee

[Dave G4UGM]

I would still be interested, I own a couple of Vaxen….

[David Nagel]

Been a long time also but the
DIMENSION ICTL(0/20),IHELP(11)        statement is miss typed. The ICTL(0/20) should be ICTL(0,20). However I don't remember the use of (0) as a dimension value. I could be wrong there.

Dave Nagel

IBM 1130 1968-1972

[James Field]

ALL of this [below] could have been obviated by the sparse or even clear use of "comments"! Those of us who hide their cleverness under the mask of secrecy are fools to themselves.
COBOL forced clarity, Admiral Grace M. Hopper knew well the benefit of clarity.

Cleverness is best demonstrated openly where it can be examined by peers.


Jim Field

[Tom Watson]

The '/' in DIMENSION (REAL/INTEGER) statements was to specify the lower
bound. This was an optional feature that wasn't implemented in many
Fortran compilers. If you had an 029 character set, the ':' was used
for this purpose.

The character sets differences is similar to using '$' as the prefix for
statement numbers on 026 keypunch (48) character sets. With the 029
(64) character set the '&' was used for this purpose.

Of course the standard 1130 Fortran didn't have either of these features
(at least while I used it 1967-1970).

And so it goes.

> > +unsub...@googlegroups.com.

> > For more options, visit
> > https://groups.google.com/d/optout.
> > --
> > You received this message because
> > you are subscribed to the Google
> > Groups "IBM1130" group.
> > To unsubscribe from this group and
> > stop receiving emails from it, send
> > an email to ibm1130

> > +unsub...@googlegroups.com.

> > For more options, visit
> > https://groups.google.com/d/optout.
> > --
> > You received this message because
> > you are subscribed to the Google
> > Groups "IBM1130" group.
> > To unsubscribe from this group and
> > stop receiving emails from it, send
> > an email to ibm1130

> > +unsub...@googlegroups.com.

> > For more options, visit
> > https://groups.google.com/d/optout.
> >
> > --
> > You received this message because you are
> > subscribed to the Google Groups "IBM1130"
> > group.
> > To unsubscribe from this group and stop
> > receiving emails from it, send an email to
> > ibm1130+u...@googlegroups.com.
> > For more options, visit
> > https://groups.google.com/d/optout.
> > --
> > You received this message because you are subscribed
> > to the Google Groups "IBM1130" group.
> > To unsubscribe from this group and stop receiving
> > emails from it, send an email to ibm1130

> > +unsub...@googlegroups.com.

> > For more options, visit
> > https://groups.google.com/d/optout.
> >
> > --
> > You received this message because you are subscribed to the
> > Google Groups "IBM1130" group.
> > To unsubscribe from this group and stop receiving emails
> > from it, send an email to ibm1130

> > +unsub...@googlegroups.com.

> > For more options, visit https://groups.google.com/d/optout.
> > --
> > You received this message because you are subscribed to the Google
> > Groups "IBM1130" group.
> > To unsubscribe from this group and stop receiving emails from it,
> > send an email to ibm1130+u...@googlegroups.com.
> > For more options, visit https://groups.google.com/d/optout.
> >
> >
> >
> >
> > --
> > You received this message because you are subscribed to the Google
> > Groups "IBM1130" group.
> > To unsubscribe from this group and stop receiving emails from it,
> > send an email to ibm1130+u...@googlegroups.com.
> > For more options, visit https://groups.google.com/d/optout.
>
>
>
> --
> You received this message because you are subscribed to the Google
> Groups "IBM1130" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ibm1130+u...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
--

Tom Watson
t...@johana.com I'm at home now

[Dave Wade]

Did any more information on this emerge....

[Richard Stofer]

Carl Claunch thought he might have the card deck but he is not currently able to convert the cards to a disk file.  I’m hoping at some point that this will all work out.

 

I do have the source listings but at least one page had a coffee spill before it was reproduced.  Even if that weren’t a problem, typing that much code is a fools exercise, especially with failing eyesight and dubious typing skills.  It doesn’t help that, in typical IBM fashion, the text has been deblanked wherever possible.  It makes it hard to follow.

 

OTOH, I bought a text book “IBM Electronic Circuit Analysis Program” by Jensen and Lieberman and it’s a very complete explanation of the program and its application.  Toward the end of the book, they get into modeling mechanical systems (heat and hydraulic systems) as well as modeling analog computers.  These topics were way above my pay grade when I was using the program as an undergrad.  Today, I find analog computers to be fascinating.

 

The IBM Continuous System Modeling Program is another nice way to play with analog computers.

 

Richard

--

[Pete]

If for some reason Carl doesn't have it, why not "crowdsource" it, or whatever name is applicable? Distribute scans of pages to volunteers to be keyed in. If enough people work on it it shouldn't be too bad. It would still need cleaning up, but would still probably be cleaner than OCR.

I'm still looking for the source for SPL (I believe, I may have forgotten the acronym over the past several years) - a PL/i interpreter written in FORTRAN that ran on the 1130 among other systems.

[Carl Claunch]

Hi Pete

The IBM catalog of programs listed an SL/1 which is probably what you remember. I too would love to find the source (or even binaries) for that. 

Carl

[Bob Flanders]

Crowd sourcing is a great idea. You can probably offer a few cents a page. We have used Amazon Turk machine with success.

On Monday, January 19, 2015, Pete <Peter...@yahoo.com> wrote:

If for some reason Carl doesn't have it, why not "crowdsource" it, or whatever name is applicable?  Distribute scans of pages to volunteers to be keyed in.  If enough people work on it it shouldn't be too bad.  It would still need cleaning up, but would still probably be cleaner than OCR.

I'm still looking for the source for SPL (I believe, I may have forgotten the acronym over the past several years) - a PL/i interpreter written in FORTRAN that ran on the 1130 among other systems.

[Lucio Fassio]

Have you an autofeed scanner? I wrote a C# program able to read punched holes from an image of the card and produce a text file with the Ascii content, good for the 1130 emulator.

--

[Carl Claunch]

I just finished reading in the approximately 9200 cards and sent the files to Richard for checking. Once it is cleaned up and working we can see if Brian and Norm want to host it on ibm1130.org.

I am amazed by how well the Documation reader handles old, swollen and bent cards. Feeding four and a half boxes of cards with as few stops as I got is impressive. Glad the card handling skills don't go away, riffling, loading, packing away and keeping everything straight. Because of the condition of the cards, the reading averaged about 200 cards per minute, inclusive all all the stops, card handling and cards that didn't pick on the first try. That means a box of 2000 cards takes ten minutes to scan, not too bad at all. I have about 50 boxes worth of software on cards and this is a tolerable processing rate. 

Carl

[Richard Stofer]

I have looked at the files briefly and I will have to run them through some kind of script to replace & or < with ( or ).  Not a big deal at all.  In the real world, this probably doesn’t matter but my system assumes ASCII.  We had a similar problem with the parenthesis and the plotter output.  I think it is related to 026 vs 029 keypunch machines but I am not certain.

 

I will probably break up each file and keep the results as one file per subroutine.

 

Richard

 

 

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Carl Claunch
Sent: Monday, February 02, 2015 5:47 PM
To: ibm...@googlegroups.com
Subject: [IBM1130] Re: IBM Electronics Circuit Analysis Program - ECAP

 

I just finished reading in the approximately 9200 cards and sent the files to Richard for checking. Once it is cleaned up and working we can see if Brian and Norm want to host it on ibm1130.org.

--

[Tom Watson]

Richard (and others)...

The Fortran compiler would accept both the 026 and the 029 coding of
Fortran characters. If I remember correctly, they are (form looking at
an EBCDIC table):
& -> +
# -> =
< -> (
% -> )
@ -> '

The 407 (aka 1132 printer) used a 48 character set, which is what
Fortran uses. The 029 keypunch uses a 64 character set (upper case).
The keys on the keypunch generate codes based on their location
regardless of their key caps. The shifts on the alphabetic keys (other
than those used for numbers) were the added to generate the x-2-8,
x-5-8, x-6-8 and x-7-8 hole combinations.

Yes, keypunches are weird and some more weird than others. The keypunch
keyboard was carried over to key-tape machines. I don't have experience
on lower case devices (129 keypunch as I recall). I left the world of
punch cards in 1978.

One of these days I'll get the Documation reader I have functional to
read in some of the OLD decks I have from my 1620 days.

...Tom Watson

[Dave G4UGM]

An I say thanks as well Carl. Looking forward to trying this.

 

Dave

 

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Carl Claunch
Sent: 03 February 2015 01:47
To: ibm...@googlegroups.com
Subject: [IBM1130] Re: IBM Electronics Circuit Analysis Program - ECAP

 

I just finished reading in the approximately 9200 cards and sent the files to Richard for checking. Once it is cleaned up and working we can see if Brian and Norm want to host it on ibm1130.org.

--

[Dave G4UGM]

1978.
>
> One of these days I'll get the Documation reader I have functional to read in
> some of the OLD decks I have from my 1620 days.
>

You don't have a copy of PACTOLUS in there do you ?

> ...Tom Watson
>

[Tom Watson]

No I don't have PACTOLUS (I don't remember this name, but somehow it
seems familiar. I do have the sources to 1620 Witran (and a listing
that came out of a 1443 printer). So on to the Documation reader one of
these days.

...Tom Watson

[Peter Flass]

Great job. I don't know how many cpm the reader is rated at, but I think the slower readers are a little more tolerant to out of spec cards than the faster ones. Looking forward to seeing what you have!

[Richard Stofer]

Well, the easy way would be to just read the files as is and see if the compiler chokes. After all, they are valid ASCII characters. I have an ASCII to Hollerith translation table as part of the FPGA implementation. I would have to see what I did with the odd characters.

It might work. If not, even a global search and replace isn't all that difficult.

Many thanks to Carl for all his efforts to make this possible.

Richard

[Miles_gmail]

Just a side note about ASCII and EBCDIC... We put a Logicon async adapter on our 1130. We wanted to communicate to Texas Instruments 742 programmable terminals. The 742 communicated using ASCII and that was totally foreign to me at the time.  The 1130 needed EBCDIC.  There was a routine that IBM provided that would accept a conversion table and translate whatever it was given. 

 

Great fun making that all come together.  If I remember correctly the 742 used an Intel 4004 chip.  It had a neat language “TICOL”.  As a “service bureau” we placed the 742’s in offices of accountants and building material dealers. Our night operators would “poll” the 742’s one at a time and then process the data and we’d ship invoices and reports back the next day via courier and UPS.  All done at the amazing speed of 1200 baud!

 

In our building material systems we used a Mod 11 check digit scheme for customer numbers etc.  The IBM 129 keypunches had that feature (or maybe it was an option, can’t remember).  We used it on the 742 too.  Thus if the operator dropped a digit or transposed some, the check would fail and they’d have to correct it in order to proceed.  As I write this, it’s interesting what comes to mind… “93017” was the number we used for the cash sales account. We printed out lists with allowable check digit numbers and as customers were added, numbers were crossed off the list.

 

Another “trick” was encoding certain fields in reports that could be seen by folks that needed some semblance of secrecy (for things like item cost, pay rate etc.)  Take each digit and either add or subtract 5 to/from it but don’t do any carry/borrow.  Thus a cost 15.76 would look like 60.21.  Those “in the know” could readily understand the data, others didn’t have a clue.

 

Did a quick look on the Internet and here is an ad for a TI 742.  It had two cassettes,  One contained the TICOL program, the other contained the data that the operated keyed in.  We made our data look like the 80 column card images that our 1130 used. 

 

 

 

Miles

 

 

-----Original Message-----
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Richard Stofer
Sent: Tuesday, February 3, 2015 9:22 AM
To: ibm...@googlegroups.com
Subject: RE: [IBM1130] Re: IBM Electronics Circuit Analysis Program - ECAP

 

Well, the easy way would be to just read the files as is and see if the compiler chokes.  After all, they are valid ASCII characters.  I have an ASCII to Hollerith translation table as part of the FPGA implementation.  I would have to see what I did with the odd characters.

 

It might work.  If not, even a global search and replace isn't all that difficult.

 

Many thanks to Carl for all his efforts to make this possible.

 

Richard

 

 

-----Original Message-----

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Tom Watson

Sent: Monday, February 02, 2015 9:14 PM

To: ibm...@googlegroups.com

Subject: RE: [IBM1130] Re: IBM Electronics Circuit Analysis Program - ECAP

 

Richard (and others)...

 

The Fortran compiler would accept both the 026 and the 029 coding of Fortran characters.  If I remember correctly, they are (form looking at an EBCDIC table):

& -> +

# -> =

< -> (

% -> )

@ -> '

 

--

[Carl Claunch]

It is the Documation 600 so a 600 CPM machine. 

Its advantage is how it feeds cards - not dependent on a carefully adjusted throat slit like IBM machines, instead uses an air stream to blow the cards apart and then a rubber foot with vacuum to grab onto just one card and move it into the rollers. Very tolerant of variations in card thickness and condition. 

Carl

[Carl Claunch]

I did just that and the simulator and Fortran compiler had no problem at all with interpreting the characters as parentheses. 

The only problem with the current four files I scanned is that they break in the middle of some Fortran source modules. If they are put in the virtual CR one at a time, the program that spans each file break won't be successfully compiled. A bit of reorganization should suffice, or read these together using a simulator control file.

Carl

[Jeff Jonas]

> Just a side note about ASCII and EBCDIC...
> We put a Logicon async adapter on our 1130.

Neeters! Did it use the IBM async card positions
and emulate the IBM SCA (sync comm adapter)?
Or did it tap into the bus elsewhere?
I remember comments how the SCA was used like the
Unix/Linux poll()/select() for a timeout without any I/O.

> We wanted to communicate to Texas Instruments 742 programmable terminals.
> The 742 communicated using ASCII and that was totally foreign to me
> at the time. The 1130 needed EBCDIC.
> There was a routine that IBM provided
> that would accept a conversion table and translate whatever it was given.

I vaguely remember writing 1130 assembler for card to paper tape.
I probably made a ZIPCO table for converting card code to ASCII.
Or perhaps an EBCDIC to ASCII table in anticipation of other inputs.

-- jeff jonas

[Miles_gmail]

The Logicon LI/ON (Logicon Input Output Network) was a SAC device.  At that time we had a 1403 printer attached which used SAC I.  So we ordered a SAC II attachment.  Our CE's installed the SAC II and we couldn't get the LI/ON to work.  Fiddled around and the CE's had no interest in assisting us (justifiably so, IMHO).  Finally we disconnected the cable from the SAC I and plugged the LI/ON into it. Worked perfect.  The CE's came back and worked several days trying to find the problem.  Turned out to be an incorrect wire-wrap from the factory.

 

Logicon provided some rudimentary software that we could "call" from FORTRAN to make it work.

 

We had adopted the EMU (Eastern Michigan University) version of FORTRAN.

 

Sometime later we added more Logicon stuff and sent the 1403 back to IBM as it had been rented.

 

Here are some pictures in those later years... 1977

 

 

 

 

 

 

We also had a continuous form “burster”.

 

As I remember things, that red plastic thing would light up when a check digit failed.

 

 

 

We had fun…

 

Miles

 

 

 

-----Original Message-----
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Jeff Jonas
Sent: Tuesday, February 3, 2015 3:06 PM
To: ibm...@googlegroups.com
Subject: RE: [IBM1130] Re: IBM Electronics Circuit Analysis Program - ECAP

 

> Just a side note about ASCII and EBCDIC...

[Miles Sandin]

The Logicon LI/ON (Logicon Input Output Network) was a SAC device.  At that time we had a 1403 printer attached which used SAC I.  So we ordered a SAC II attachment.  Our CE's installed the SAC II and we couldn't get the LI/ON to work.  Fiddled around and the CE's had no interest in assisting us (justifiably so, IMHO).  Finally we disconnected the cable from the SAC I and plugged the LI/ON into it. Worked perfect.  The CE's came back and worked several days trying to find the problem.  Turned out to be an incorrect wire-wrap from the factory.

 

Logicon provided some rudimentary software that we could "call" from FORTRAN to make it work.

 

We had adopted the EMU (Eastern Michigan University) version of FORTRAN.

 

Sometime later we added more Logicon stuff and sent the 1403 back to IBM as it had been rented.

 

Here are some pictures in those later years... 1977

 

Sometime after these pictures we added a Logicon paper tape reader.  It was a long time arriving and we used a teletype for a while which we connected to the LI/ON async device.  When the reader finally arrived, it had serial # 0001.  Probably had something to do with its delay in delivery.

 

 

 

 

 

 

 

We also had a continuous form “burster”.

 

As I remember things, that red plastic thing would light up when a check digit failed.

 

 

 

We had fun…

 

Miles

 

 

 

-----Original Message-----
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Jeff Jonas
Sent: Tuesday, February 3, 2015 3:06 PM
To: ibm...@googlegroups.com
Subject: RE: [IBM1130] Re: IBM Electronics Circuit Analysis Program - ECAP

 

> Just a side note about ASCII and EBCDIC...

[Richard Stofer]

Carl did all the heavy lifting and he has a good deck of cards.  ‘Good’ means they compile and at least one example runs.  I’m not sure where the example came from (yet).

 

I took his decks, changed the *IOCS statements, corrected the oddball characters and fixed up the LUNs for the 1403 and 2501.  My files also compile and execute producing identical results with an IBM example.  My test was only DC Analysis so there is a lot of testing yet to go.

 

FWIW, I don’t seem to need the 8 *LOCAL statements.  My 32K machine seems to be able to hold all the code and data.  More to follow as I get into testing AC Analysis and Transient Analysis.

 

At the moment, I am totally impressed.  I remember this program from 44 years ago and I specifically remember what a help it was with tuned circuits for EE school.

 

I have a book, “IBM Electronic Circuit Analysis Program – Techniques and Applications” that has a bunch of examples of VERY complex circuits including a chapter or so on analog computing.  More to follow as I work through the examples.

 

Richard

 

[Bob Flanders]

Richard/Carl,

Please post examples of input and output.

Thanks,
Bob

--

[Richard Stofer]

A DC Analysis

 

*CONTROL,TYPE

       DC ANALYSIS

B1    N(0,1),R=10

B2    N(0,2),R=25

B3    N(1,2),R=5

B4    N(2,3),R=75

B5    N(3,4),R=40

B6    N(1,3),R=12

B7    N(1,4),R=60

B8    N(2,4),R=150

B9    N(4,0),R=70,E=0.25

B10   N(4,5),R=1000

B11   N(3,5),R=125

B12   N(3,0),R=200,I=0.03

B13   N(5,0),R=500

       PRINT,VOLTAGES,CURRENTS

       EXECUTE

 

 

EXECUTION

 

 

 

 

NODE VOLTAGES

 

 

NODES               VOLTAGES

 

  1- 4      0.15887159E 00    0.14576512E 00    0.38168126E 00    0.15532991E 00

  5- 5      0.29170739E 00

 

 

ELEMENT CURRENTS

 

 

BRANCHES            CURRENTS

 

  1- 4     -0.15887156E-01   -0.58306045E-02    0.26212930E-02   -0.31455480E-02

  5- 8      0.56587830E-02   -0.18567472E-01    0.59027967E-04   -0.63765241E-04

  9-12      0.57904264E-02   -0.13637746E-03    0.71979093E-03    0.19084068E-02

13-13      0.58341468E-03

 

An AC Analysis

 

*CONTROL,TYPE

C     NAGNETICALLY COUPLED NETWORK

       AC ANALYSIS

B1    N(0,1),R=3.4,E=10

B2    N(1,2),L=55E-6

B3    N(2,3),C=.0076E-6

B4    N(2,4),C=.0421E-6

B5    N(4,5),R=120

B6    N(3,6),L=725E-6

B7    N(6,0),R=5.1

B8    N(8,3),L=106E-6

B9    N(5,7),L=450E-6

B10   N((7,8),R=.5

M1    B(6,8),L=268E-6

       FREQUENCY=50E3

       PRINT,VOLTAGES,CURRENTS

       EXECUTE

 

 

EXECUTION

 

 

 

 

FREQ =  0.50000007E 05

 

 

 

       NODES           NODE VOLTAGES

 

 

MAG    1- 4  0.99733295E 01  0.97114544E 01  0.59348611E 01  0.11601186E 02

PHA          0.29630512E 00 -0.49926817E 00  0.19363559E 02  0.69566178E 00

 

 

MAG    5- 8  0.11672561E 02  0.87175265E-01  0.81749420E 01  0.81777000E 01

PHA          0.15601831E 02 -0.62555122E 02  0.19417514E 02  0.19503547E 02

 

 

 

 

 

    BRANCHES           ELEMENT CURRENTS

 

 

MAG    1- 4  0.17095863E-01  0.17095334E-01  0.10972710E-01  0.25164715E-01

PHA         -0.62539550E 02 -0.62542404E 02  0.63474807E 02 -0.83194091E 02

 

 

MAG    5- 8  0.25164749E-01  0.17093349E-01  0.17093189E-01  0.25164611E-01

PHA         -0.83194137E 02 -0.62555404E 02 -0.62555107E 02 -0.83203155E 02

 

 

MAG    9-10  0.25164749E-01  0.25166101E-01

PHA         -0.83194198E 02 -0.83201111E 02

 

These examples are from the IBM ECAP Manual and the results match exactly.

 

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Bob Flanders

Sent: Wednesday, February 04, 2015 12:49 PM
To: ibm...@googlegroups.com

Subject: Re: [IBM1130] ECAP - Revisited

[Bob Flanders]

Thank you!

--

[Peter Flass]

Cute staff.

[Clare Owens]

Well, that brings back memories for me.  I had purchased a KLH model 16 transistorized stereo preamp/amp in about 1963 - I was a Hi Fi nut but with a limited budget since I was working for the Univ. of Buffalo Computing Center running an IBM 1620 over in the Medical building and the pay was miserable - but it was a lot of fun.  The KLH 16 was my first stereo amp and it worked fine for probably a year and then got noisy but they sent me a replacement board and then things were fine for another year or so until one channel went out completely.  A little testing in my workroom found that apparently one pair of the four PNP Germanium output transistors (2N2147) in that channel were shorted.  They were quite expensive but I bought two and installed them.  They probably lasted a few milliseconds since I had failed to notice that the other two in that channel also were shorted.  Buying another set of four was out of the question since I had recently married and money was really tight even though by then I was an IBM systems engineer in the Buffalo branch.  I had access to a 274x terminal to an on line internal IBM service that had ECAP and I used that to model the output circuit, and then did it again using the much cheaper 2N3055 NPN silicon transistors.  It worked OK with the balance control cocked a little to the left.  Although the KLH 16 is long gone I still have its file folder and am looking at the printout of the runs I made and the schematics I drew for reference.  Some old files just seem to be too good to discard.

I did have the 1620 ECAP manual and it is gone now.  It had page after page of source listings.

Sorry for the long reminiscence...

Clare

--

[John McKee]

Clare, your comments were most interesting reading.  Don't apologize!

John McKee

[Richard Stofer]

Here is the final example in the manual, a Transient Analysis

 

*CONTROL,TYPE

       TRANSIENT ANALYSIS

B1    N(0,1),R=(0.1,1E6),E=15

B2    N(0,1),R=(1E6,0.5)

B3    N(1,2),R=0.1

B4    N(2,3),L=0.2E-3

B5    N(3,4),C=2E-6

B6    N(0,4),R=1

B7    N(0,3),R=6

B8    N(3,0),R=1E6,E=(-12,-11.7)

S1    B=8,(1,2,8),OFF

       TIME STEP =1E-6

       OUTPUT INTERVAL = 25

       FINISH TIME =75E-6

       PRINT,VOLTAGES

       EXECUTE

 

 

EXECUTION

 

 

 

 

 

    T =  0.00000000E 00

 

NODES               VOLTAGES

 

  1- 4      0.15000000E 02    0.15000000E 02    0.11670554E-04    0.11555006E-04

 

 

 

    T =  0.24999957E-04

 

NODES               VOLTAGES

 

  1- 4      0.14846353E 02    0.14692708E 02    0.58473768E 01    0.56191086E 00

 

 

 

    T =  0.49999806E-04

 

NODES               VOLTAGES

 

  1- 4      0.14775024E 02    0.14550048E 02    0.11438943E 02    0.34327679E 00

 

 

 

    T =  0.53781441E-04

 

NODES               VOLTAGES

 

  1- 4      0.14769691E 02    0.14539382E 02    0.12000015E 02    0.30306816E 00

 

 

SWITCH  1 IS 0N

 

 

 

    T =  0.53781441E-04

 

NODES               VOLTAGES

 

  1- 4     -0.11515297E 01   -0.13818376E 01    0.12002500E 02    0.30253034E 00

 

 

 

    T =  0.61164406E-04

 

NODES               VOLTAGES

 

  1- 4     -0.90801000E 00   -0.10896139E 01    0.11699995E 02   -0.13400653E 00

 

 

SWITCH  1 IS 0FF

 

 

 

    T =  0.61164406E-04

 

NODES               VOLTAGES

 

  1- 4      0.14818395E 02    0.14636791E 02    0.11698652E 02   -0.13400653E 00

 

 

 

    T =  0.74999639E-04

 

NODES               VOLTAGES

 

  1- 4      0.14797887E 02    0.14595775E 02    0.11781913E 02    0.57484544E-01

 

 

 

   T =  0.75999633E-04

 

NODES               VOLTAGES

 

  1- 4      0.14796489E 02    0.14592981E 02    0.11821661E 02    0.64823344E-01

 

 

As near as I can tell, the program is working properly.  Things may turn up later when I try to get the output to the printer instead of the console or when I start messing around with the console switches.

 

Richard

[Yvette S. Hirth, CCP, CDP]

On 02/04/2015 12:43 PM, Richard Stofer wrote:

> ‘Good’ means they compile and at least one example runs.

ok, now put it into Production! 8^)

yvette

[Richard Stofer]

I have been carrying the IBM 1130 ECAP manual and other documentation around for more than 40 years.  I KNEW it would come in handy one day.

 

Richard

 

 

From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Clare Owens
Sent: Wednesday, February 04, 2015 1:25 PM
To: ibm...@googlegroups.com
Subject: Re: [IBM1130] ECAP - Revisited

 

<snip>  Although the KLH 16 is long gone I still have its file folder and am looking at the printout of the runs I made and the schematics I drew for reference.  Some old files just seem to be too good to discard.

I did have the 1620 ECAP manual and it is gone now.  It had page after page of source listings.

Sorry for the long reminiscence...

Clare

[Bob Flanders]

I agree. Thanks for that.

I wonder... how many people in this group first used a 1620 before an 1130? I had a year to use a 1620 in high school, but never understood what I was doing. Just kept playing around and getting things to print. The school replaced it the next year with an 1130...

Maybe this should start a new thread....

--

[Bob Flanders]

Very cool. Those packs look like 2314 packs .. or 3330. Do you remember how many platters?

On Wed, Feb 4, 2015 at 4:22 PM, 'Peter Flass' via IBM1130 <ibm...@googlegroups.com> wrote:

Cute staff.

[Dave G4UGM]

The IBM1620 was the first commercial computer I programmed. We had a small demo machine in what I think the US would call “high school”, so 16-18 that was built with germanium transistors, that had to be programmed with buttons and had lights for readout.

 

Dave.

[gah4]

The 2314 has 20 tracks/cylinder on 11 platters, not using the top of the top,

or the bottom of the bottom.

The 3330 is similar, but one whole platter is timing tracks, so only 19 tracks/cylinder.

The 2314 has 7294 bytes/full track.  The 3330 has 13030 bytes/full track.

More common, for the 3330 is 1/4 track blocking at 3156 bytes.

[gah4]

I had your files running on gfortran a year or two ago.

The biggest, and strangest, problem is that all the COMMON blocks are wrong on one,

I think the transient analysis module.  It seems it is a different version.  But putting the

matching COMMON in, and renaming some variables, fixed that one.

gfortran won't initialize non-CHARACTER variables with apostrophe constants,

only actual Hollerith constants.  So, changed them all to Hollerith, which should

also work with old compilers.

And gfortran has no idea about chaining.  So, at the end of each one, I write out

all the COMMON data to disk, and the next one reads them back in.

And I tried many of the examples in the ECAP book.

About 50 years ago, I had the ECAP book from the library, but didn't know anyone

who had it.  So, finally, got it running!

On Tuesday, December 16, 2014 at 11:35:03 PM UTC-8 Carl Claunch wrote:

Pretty sure I have it - machine readable but as a long sequence of cardboard rectangles with holes punched in Hollerith encoding, nothing more easily transferable at the moment. Once I can read and transfer card files you are welcome to a copy. 

Carl