Small WANG museum

	Small					Museum
	*Jan van de Veen, Nijkerk, The Netherlands*

last update July 2011

You can click on any machine in this overview to see the full image

Links

Credits

Manuals and spare-boards

Steam-page
(just another hobby)

Great News ! (April 2007) It took some years, but the Small WANG Museum now has the complete line of the 2200 series That is the 2200B - VP - LVP - SVP - MVP - CS - CS turbo, and the PCSII and the E model ranging from 1973 to 1995. Yes, it still has to be arranged and that will take some time as I am now extremely busy with my steam hobby: www.stoomwerktuigen.nl July 2011 Somehow I have not yet been able to create a new site to show you all the 2200 products. For eight and a half year now we are restoring this unique German Traction Engine from the First World War and it is taking up all my spare time. The good news is the we finished the restoration in April 2011. For most of the summer we will be on tour with the traction engine, but later this year, there should be more time for me to spend working on a new site. I will do my best to have the new site for the WANG museum ready by the end of 2011 ! (Jan van de Veen, The Netherlands)
1972 - 2011 Form 1972 to the 30th of April 2011, I worked on a Institute in the Netherlands that is involved in research in the field of Animal Nutrition. For all these years I have been responsible for the data collection. First mainly manually, but as computers became more sophisticated, this whole complicated process was automated. The Institute used WANG computers from 1972 to 1995. As it is, I did start to work on the Institute at the same moment a WANG 720 configuration was bought, in September 1972. So this site tells a little bit of the WANG history as well as the IT process that took place on the Institute for research on Animal Nutrition www.schothorst.nl/
First some mechanical calculaters that are in the museum too

	MonroeMatic This is an electric driven mechanical calculator made by Monroe. It still works. It was mainly used to carry out simple statistical analyses. You could quite easily make sum of squares and total sum of squares that are needed when doing Analyses of Variance. For more complex analyses the Institute used time-sharing of Honeywell-Bull on a 300 baud tele-type !. This machine was made in Amsterdam, as Monroe had an office and factory there for the European market. I did never work with this type of calculators, but I know how to make it move and have it make a lot of noise.

	Friden SBT This is a calculator from another famous company, called Friden. To know about the history of Friden click here. You could perform correlation and regression analyses on this machine. In order to do so you need sum of square X, sum of square Y and sum X * Y. To get this result, you would enter X and Y as one number, with sufficient zero's in between. If X = 35 and Y = 48, you would enter 3500048 and have this number squared. The result will be 12250336002304. The left four digits are 3535, the last four digits 4848 and in between is 3360, which is 2(3548). Add this result to the total sum, and progress with the next entry. Extreme care had to be taken to avoid overflow, and that is why the carriage has so many digits. By 1977 these machines where replaced by electronic calculators, (Model 1930) also made by Monroe

	Friden 1151 This electronic calculator was bought by the Institute in 1969. It had a paper writer and 4 numbers could be stored in memory. It was even possible to make a small program of 15 steps at a maximum. Today we would call it a macro as the machine ' learned' what to do, because you performed the calculation with the 1151 set to learn mode. You would than enter the starting values and had the program running. I cannot recall what we did with this wonderful machine, but I will never forget how to find the square root of a number by using the Newton algorithm.

	This image shows the simple keyboard of the Friden 1151. The learn button is at the right, just next to the wheel where you could set the number of decimals to be printed. How do you find the square root of a number ? Suppose you want to know the square root of 285. You know that it is more than 10, but less than 20, so make a guess in between, say 15. Now, you divide 285 by 15 with 19 as result. Take the mean of (15+19) = 17 and start again, dividing 285 by 17 with 16.75 as result. Take the mean of (17+16.75) = 16.87. Again divide 285 by 16.87 with 16.89 as a result. It is now clear to see that 16.88 is the square of 285 and that we found this result in an iterative way, using only 3 cycles. When the Friden had learned this trick, you would enter the number you want the square root off, than enter your guess, and hit the auto button. The result would be printed and the program had to be stopped manually as soon as there was no more change in decimals in the printout.

	This is a Friden 130 from 1963. It was the first all electronic calculator with a 4 line CRT display. Although never used on the Institute, I did add this calculator because it has delay lines as memory. Please visit the site Vintage Calculator Web to learn more about this calculator. (Courtesy Mr Horsman)

The WANG 720C (1972-1977)
	In 1972 the Institute had a need for more calculation power so they could perform more complex analyses without using time-sharing. They bought a WANG 720C in 1972, together with a program to do Multiple Regression Analyses. This picture, taken on the 21st of April 1973 at 2.30 in the morning, shows the 720, together with the dual cassette drive and the IBM-selectric printer. What can not be seen, but is there is a 708 extended memory. This black-box contained another 4K of core-memory. Added to the 2K of the 720, there was 6K all together. In a 720, you can store 256 numbers and no program, or 2048 program steps and no data and anything in between. Care had to be taken to avoid a collision of data and program. The extra memory was used for data-storage, as a Multiple Regression Analyses needs a big matrix of numbers. The program was made by WANG Belgium in Ottergem
Is soon learned to program the 720 and still today I know almost all the codes by head. A program would look like this: (it stores the value of zero in the first 100 data registers) Step Code Shorthand Explanation 0001 7000 0 In words it means that a zero is placed in the X-register 0002 0604 UP and moved to the Y-register 0003 0408 MARK A flag is set 0004 0001 0001 with label 0001 0005 0700 0 zero is put in the X-register 0006 0504 ST.IND and stored in the data-register who's number is in the Y-register (indirect storage) 0007 0701 1 one is put in the X-register 0008 0600 + and added to the Y-register 0009 0701 1 the number 0010 0700 0 100 0011 0700 0 is put in the X-register 0012 0509 SKIP X=Y X en Y are compared. If not equal 0013 0407 SEARCH the program jumps back 0014 0001 0001 to the flag with label 0001 0015 0515 STOP if equal the program stops A 720 has a X and Y register so you can step through the program and see it happen right in front of you.
	The machine that gave the most trouble was the dual cassette drive. The tapes were not formatted and you could store blocks of 1 byte up to 256 bytes in sequential order. In practice you would always use 256 byte blocks, but still there where too many problems. The tapes where stored in albums, 12 data-tapes in an album. I still have several of these albums. For those of you who want to know more of the technical details of the 720, please visit the site of Rick Bensene and have a look at his wonderful page about the 720 In 1974 we bought a Memorex dual 8 inch floppy-drive and a Centronix high-speed impact printer
	This is the same 720C, still running perfectly, with the Memorex dual floppy drive (740-2) and the 709 tape drive. You can clearly see the X and Y register with the nixie-tubes. The floppy drive was a wonderful extension. It was fast and reliable and gave us the possibility to do much more with the 720. Soon we had a herd-management program running for our breeding sows, hens, dairy cattle and so on. Certainly the management program for the sows must have been one of the first in the world. I later on developed a commercial sow-management program for the 2200 and this is still widely used by thousand of farmers in the Netherlands. But, the basic idea was born on this 720 !. In my opinion the 720C was more a desk-top computer then just a calculator.
	IBM selectric. The 720 series needed an output device. The IBM selectric was converted to work with the 720. Amazingly there is no on/off button, it can only be switched on by sending an special code from the 720 !. It is a wonderful piece of engineering with an incredible speed and print quality. The ball shaped typing head spins around like mad. I have made two small clips with the selectric in action. In close up, the image could not keep up with the speed of the typing head but the sound is real !. Without the cover and from a further distance you can see the typing head moving along, but in less detail. It is typing the famous sentence "Now the time has come to see the quick brown fox jump over the lazy dog" (Courtesy Mdm Föllmi, Heemstede)
	This is a recently acquired C50 scientific calculator. It has a so-called left and right calculator that do work independently. As they share a common memory, complex calculations can be carried out without to much trouble. Of course, this machine also has special function keys that can perform pre-programmed tasks like calculating standard deviation etc. It was even possible to link an IBM punch card reader to make small programs.
	The manual is very well made. It is simple and clear and within 20 minutes you know how to work on the C50. ( Courtesy Mr Palache)

2200(1976-1995
	This is one of the very first models form the 2200 line of products It is a model 2200 E2 Built around 1976. It may look like a terminal, but it is a full scale computer. All the logic boards are inside. Just plug it into the mains and it will show READY BASIC: The CRT is 16 x 64. The cassette drive on the right could store programs and data. As the cassette drive was operating, the screen would immediately loose its stability and start to shake each time a block of 256 bytes was read from the tape. It even holds a D25 socket so you could attach a modem or perform serial communications. There is a Centronics parallel interface as well for connecting a printer. Many thanks go to Georg Schaefer from Germany who did an excellent job in repairing the CPU board, so the 2200 is in working order again. (April 2005) Courtesy Mr Ed Schijf (2005)
	This is an 2200 PCS II built and approved in May 1978. Much like the 2200 E2 model showed above, but with a wonderful extension of two 5.25 floppy drives mounted on top of the case. Again, all the logic is inside, making it quite heavy indeed. PCS stood for Personal Computer System, and indeed, with this machine on your desk you could perform almost any task, not being dependent of anyone else To see more about the PCS II click here Courtesy Mr Ed Schijf (2005)
	I had seen the 2200 A en T models of 1974 at the WANG office at Utrecht. The CRT was a great advantage, but I did not like the tape, nor the screen (16 x 64) and these were slow computers. But..... than in 1976 I was allowed to do a little work on the 2200 VP. This was different ! There was a Diablo 10Mb hard disk attached and this machine went like a rocket. It could perform matrix operations, even calculating an inverse matrix, by just a single statement. A 20 x 20 matrix was inverted in just a few seconds. So, at the 20th of April 1977 our 2200VP arrived at the Institute. Dr Wang was a true engineer and that was shown in the program language that came with the 2200. Basic already had a bad name, certainly at the Universities as DEC had a very poor Basic, but this was different. The 2200 works as an interpreter, solving and executing each program line at the time. You can run single lines, what makes debugging easy, provided that you understand what you are doing. On this picture you can also see the Memorex dual floppy that was connected together with the 10Mb Diablo hard disk. A 2221W impact printer completed the configuration. We worked very successfully with the VP and at one time had two systems multiplexed.
	Eventually we needed more terminals and that made the 2200MVP the obvious choice. A MVP is a VP, but with the memory divided into partitions. Each user was assigned to a particular partition. The processor would work in a round robin fashion, giving each partition a time slice of a few milliseconds. At first 8 terminals could be connected, later on expanded to 16. The Diablo soon was too small and a Phoenix drive was connected with 80Mb. (4 fixed spindles and one removable). A wonderful machine that hardly ever gave a problem. Shown here is a Shugart floppy drive, that was IBM compatible WANG had a very good team of field engineers that would do everything for the customer. Sales representatives where a different kind. I never had high thoughts about these people. It was a miracle as you had the same salesman for more than half a year, so they had no understanding what we where doing anyway. By 1984 WANG - Netherlands office had moved to a third location and even that location was to small. WANG was doing very well indeed ! But, the first PC's had entered the market and they would soon bring WANG in great problems.
	WANG had entered the world of finance and banking in 1978 by introducing the VS computer. From 1982 to 1987, not much attention was given to the 2200 line. Their installed base must have been several 100,000th, all over the world. As no new products for the 2200 came on the market, people where looking for alternatives. Some former WANG employees had made a product called NIAKWA, that would emulate a 2200 on a PC. Much cheaper and faster, but for single user only. In 1988 WANG must have realized that they where loosing customers and new products where introduced. The 2200 was now called the CS. Soon there came Disk Storage cabinets with Winchester drives, cheap terminals and printers. We soon had the Phoenix drive replaced by two DS cabinets with tapes as backup medium..
	What you see here is a very special machine. In 1990 WANG decided to have the 2200CS running on an Intel 386 chip ! The code was developed in India as I believe. This machine could have partitions much larger than the 56K that was the limit on a MVP. On the screen (image above) you can see that partition 2 has been assigned 400K ! For old 2200 users this must be a miracle. The CPU is in the right corner of the picture, still a gray, not very good looking box. You see two different terminals, an old and expensive one on the left and a cheap one on the right. The two Disk Storage cabinets are under the table. This machine still works perfectly, although it has not done any real work since July 1995. We used this CS from 1992 to 1995 very successfully
In England a company by the name of Kerridge had decided to go on with the beautiful and powerful language but not the 2200. They developed an emulator for UNIX, written in the C language and with a lot of C functionality added to it. In 1995 we bought an HP9000 and the Kerridge software, called KCML. On the 15th of July 1995 we switched off the WANG 2200CS for the last time. As we had decided not to use the WANG terminals, but use PC's as workstations connected to the HP9000, this was really the end of WANG at the Institute. It is amazing that from 1977 up till now we have used the same operating system. But, all our data are still easy accessible and the Institute has had great benefit from this. We can compile data from many experiments in one data-base to do all sorts of complicated analyses, that sometimes make live experiments superfluous. By 1995 WANG was in great troubles and had even entered a Chapter 11 status. This gave them time to restructure the company while keeping creditors at bay. Dr Wang had died some years before and I believe that his vision was badly missed. Eventually WANG became part of the Dutch company Getronics. If you would like to know more about the story of WANG, you should see to get a copy of "Riding the runaway horse". I was told that it is out of print, but perhaps some bookstores might still have it. This excellent book is written by Charles C. Kenny and covers the whole story of Wang Laboratories Publ. Little and Brown, ISBN 0-316-48919-0
The PC world
	WANG felt that they could make a better PC than IBM and so they did. The WANG PC had its own operating system, not compatible with IBM or any other make using MsDOS. I believe that we bought our first WANG PC in 1984 to do word-processing. WANG was the leader in word- processing at that time. A daisy-wheel printer was connected for output and this all went very well. Not being compatible was a big disadvantage and after a couple of years WordPerfect was the thing to have, but not on this PC ! These PC's are now called the WANG Classic's

	This is a PC 240 made by WANG, probably in their factory at Ireland. It still works well. If I am correct, this PC is equipped with an Intel 286 processor. We had several of these PC that worked very well. For some odd reason WANG had the 5.25 floppy of 1.2Mb as the standard while the 3.5 floppy was the thing to have. If you look insight, you see that it is designed very well with a very pleasant keyboard and fully IBM compatible. WANG never used that word, but said that it did comply with the industry standard. It came with a lot of manuals that contained useful information

	This is also a WANG PC, with the new WANG logo attached. But, this PC was manufactured in the Netherlands by the company called Tulip as a private label PC. It is a 386 machine, still working well, as it was used at the Institute up to November 2000 ! Note that it has a 3.5 floppy drive

	This picture shows some of the manuals that are in the museum, all in all some 50, mostly technical reference manuals dealing with the 2200 models and their peripherals. I have made up a list of all the manuals. In case that you would be interested in a copy I could make one, provided that you will pay for the cost of copying and shipping

	This is a micro-film viewer that was used by WANG field engineers. It is contained in a simple suit-case. The micro-fiches replaced probably hundreds of manuals. I now have almost all the micro-fiches that where used in the field. (Courtesy Rob Hulleman, Mr Stenvers)

	All in all I have some 250 spare-boards in the museum, mainly for the series 600, 700 and 2200. Most of the boards are in good condition, but there is no guarantee. We never used a 600 system on the Institute, so I would not mind to help anyone out there with a spare-board if needed. For 700 and 2200 boards it would depend on the number of boards present. I try to keep all my systems going and I will not ship boards that I might need for my own machines

	Dr An Wang published a book called "Lessons" in 1986. It was translated into Dutch and given as a present to all people employed with WANG in the Netherlands. The title refers to Dr WANG as the godfather of word-processing ! Probably it was also distributed in Belgium as well as they use almost the same language (Courtesy Rob Hulleman)

	The museum is now completed with a genuine WANG neon sign which will illuminate as soon as the lights are turned on. (Courtesy Mr Stenvers)

If you like to see some more images, please click here		All rights by : Jan van de Veen @2011 Nijkerk, The Netherlands