Life and Times of a Geek part 3 (HPR Show 2968)

Part 3 of my personal story of experiences with computers

Dave Morriss


Table of Contents

Introduction

In the last part of my story (show 1811 in 2015) I told you about some of my experiences at the University of Manchester as a postgraduate student from around 1973.

Today I want to talk a little more about my time in Manchester and mention some of the things I did that may be of interest to Hackers!

Researching for the episode

As I have been researching for this HPR episode I realise how long ago some of these events were - in Internet years particularly. In many cases I could find no online records of places, equipment or people. This seems to be because any records there might be are on paper and have never made it online. I contacted a company that made some of the laboratory equipment I used that I thought might be of interest, and the person I contacted said that although he remembered what I was referring to the company had kept no records of it and had had to discontinue it due to modern safety concerns.

I find this somewhat dispiriting and it makes me feel very very old!

Being a postgraduate student

A change of location

As I mentioned in the last episode, I started my time at Manchester mainly working in the Animal House in the basement of a building on Coupland Street, in the main University area.

It was not an ideal location; it was beset with cockroaches and mice which lived in the utility tunnels that ran between buildings. Not a good place to house animals or to work.

An Animal House previously owned by UMIST (University of Manchester Institute of Science and Technology) became free, and we relocated there. This was on the top floor of the Roscoe Building (I think), a newer building across Oxford Road. There were multiple rooms there for offices and to house animals and experimental apparatus, so it was an advantageous move. It required a fair amount of relocation of equipment and materials though.

Research

It was expected that we’d be able to build our own apparatus if necessary. There was a quite well-equipped workshop in the department and people with the skills to help, but they were busy and often unavailable. So I had some equipment built for me, but ended up making some of my own.

Arena

I needed an arena in which my Barbary doves would be placed, to observe their behaviour. There was a wooden arena available when I first started but this didn’t prove to be particularly usable since I had to stand beside it and look through a one-way viewing window, and the bird inside could hear me. I ended up recording behavioural data from my doves in a 1-metre square covered arena which I had built myself using Dexion1 and hardboard.

Photo from Wikipedia
Dexion Slotted Angle showing assembly of nuts and bolts with spanners
ConstructorGrp [CC BY-SA 3.0], via Wikimedia Commons

The base of this arena was where the birds were placed but above this I built a 4-sided pyramid for observation. I had access to a monochrome video camera which was mounted at the top of the pyramid. However, it would not work pointing downwards so I had to install it horizontally on a platform and set up a 45-degree mirror to look down. The camera recorded stuff on a reel-to-reel Video Recorder using what I think was ½ inch magnetic tape. I played it back in order to analyse the data using a small black and white monitor.

The arena which the camera was recording was painted white and illuminated with fluorescent lights for the best visibility of the experimental animals. There were feeding stations randomly placed around the floor where the birds would find a metered amount of grain.

I had drawn out a plan of the arena floor with the feeding stations and each bird’s movement was drawn out on a duplicated copy of this plan once the video tape could be analysed.

Skinner Boxes

In another series of experiments I was also using a device called a Skinner Box. This is a chamber (an “Operant Conditioning Chamber”) in which an animal is trained to perform some action in response to a stimulus.

These devices consisted of a small box with metal sides, and a transparent door. The wall opposite the door was fitted with a perspex panel which operated a micro-switch, and underneath was a hopper of bird seed that could be raised and lowered with an electric motor. The bird in the box was trained to peck the switch and would then receive a reward of seed by raising the hopper. The switch could be illuminated from behind so that the bird would learn to peck only when the light was on, or a particular colour was used.

The box itself was built in the department by workshop staff using a product called “Handy Tube”. This consisted of square-section steel tubes that could be fixed together with jointing pieces that were hammered into the ends of the tubes.

Campden Instruments Ltd

The Skinner box electronics was driven by programmable laboratory equipment. This came from a company called Campden Instruments Limited, and consisted of a series of metal boxes containing components which were to be clipped to metal rods through which they were powered. I can find no information on the power requirements, but a leaflet about a later device from this company shows it required 22-30 volts DC through its power rails but drew a maximum of 50 mA.

The departmental workshop had constructed floor-standing racks for these units which could accommodate several rows of the different boxes containing 5 or 6 per row.

The boxes contained a variety of electronics, from logic gates such as simple AND and OR gates and INVERTERS to chart recorders and counters. Each box had a number of press-stud connectors on the front so it was possible to use a piece of wire with suitable connectors to connect the output of one to the input of another.

I remember configuring this system to turn on a coloured key light in a Skinner Box, detect a key press and trigger the operation of the food hopper. The bird’s behaviour was recorded on a paper chart, and the number of sessions counted on a counter.

I recall working out how to make what I think was a flip-flop circuit or possibly an oscillator which I used to make a little speaker buzz when a session finished. This meant I could leave the experiment running and go back to my office where the buzzer was and be alerted when it was finished.

As I mentioned, I contacted Campden Instruments about their 1970’s systems but sadly they had no record of these devices any more. They pointed out that such laboratory equipment could not be sold today for safety reasons with the bare power rails.

Slides

Click a thumbnail to see the full image.

slide_1.png
External view of the arena. There is a viewing window, and control buttons nearby. Note the instructions chalked on the blackout material. I was lucky to have help from one of the technicians to run the experiments.

slide_2.png
The arena floor with its random pattern of feeding stations. The birds had been trained to remove the (high-tech!) lids and were being tested to see if they preferred one colour over the other when it signalled a greater reward.

slide_3.png
A simplified Skinner Box setup. This was used for training the birds. There was constant white noise in the box to prevent the birds being distracted by external sounds. A dim light illuminated the interior and the round key could be illuminated. Later a feeding hopper was added and then used to deliver food rewards.

The picture shows the rack developed in-house holding Campden Instruments equipment used to run the training session.

slide_4.png
Things got a lot more complicated when an experiment was being run in a Skinner Box. There is a mixture of the Campden Instruments and another vendor’s kit here. This rig was set up for two keys illuminated in different colours. There are counters for pecks on these two colours and for numbers of rewards.

It’s a long time ago now but I think the device with what looks like a loop of a plastic material hanging down from it was a way of triggering events at pre-defined intervals. My memory of this is rather hazy though.

slide_5.png
One of the experimental doves receiving a reward in the food hopper. The Skinner Box has been configured with two keys and one hopper. The speaker delivered white noise as explained in an earlier slide.

slide_6.png
This bird looks as if it’s waiting for one of the keys to illuminate so it can peck it and get a reward.

Being a Demonstrator

As a postgraduate student it was expected that we would help out with the teaching of undergraduate students. This took the form of being a demonstrator - assisting in laboratory sessions. We were paid for this, so it was usually something we were happy to do. It could require a fair bit of preparation though, since what the undergraduates were being taught was not necessarily something we had learned ourselves as undergraduates.

I don’t remember all the lab sessions I worked in over the years I was doing this, but I do remember a few:

  • Dissection labs - for first-year students usually. We helped with classes for Medical students, who had the reputation of leaving the worst mess of smashed up animal bits when they’d finished!
  • Microscope labs - looking at slides of stained tissues, identifying and drawing them. Many students were not experienced with microscopes and often couldn’t find their specimens, or crunched through the glass slide with the lens by winding it down without checking.
  • Physiology labs - making physiological preparations with recently killed frogs or cockroaches; monitoring nerve impulses with an oscilloscope.
  • Statistics labs - helping students perform statistical tests on data, and trying to explain what the results meant!
  • Brain labs - my Supervisor ran a lab session dissecting human brains, which were in the Medical School Anatomy laboratory. This was challenging at first since the medical students were often busy working on cadavers at the same time! My fellow PhD students and I had to learn about the structure of the human brain quite rapidly to do this course justice.

I could recount more anecdotes about these sessions. There are probably a number of HPR episodes that could be made on this subject, but I’ll control myself!

Computers

Of course, I made use of the computers that I mentioned in the last episode. To be honest, in some cases the use of these resources was more rewarding than the Biology research.

Computer Graphics Unit

As mentioned before, I had many sheets of paper with plans of my arena onto which animal tracks had been traced - from the video recordings. I needed to turn these into coordinates for computer analysis and to do this I contacted the Computer Graphics Unit, which was part of UMRCC.

They had a PDP11 computer which could be used for data capture and analysis, but particularly they had a digitiser. This was a D-MAC device, made by Dobbie McInnes of Glasgow.

I have not been able to find much information about this machine other than this StackExchange article, so I will have to try and describe it for you. It consisted of a heavy glass-topped surface on substantial legs. It could be tilted for ease of use. The top was perhaps a metre square, and under the transparent top was a space in which an X and Y sensor ran.

The principle of the device was that as a mouse or puck was moved about the table top the X/Y machinery underneath followed it to determine its coordinates. It was possible to configure the D-MAC to output coordinates more or less continuously or when a button on the puck was pressed. Its default mode of output on the models I used was 8-hole paper tape.

I would place my sheets on the table one at a time, holding them in place with masking tape. The device could be zeroed to a corner of the picture, then the track of the bird could be traced, pressing the output button at each point visited.

The Computer Graphics Unit stopped providing this service during the period I needed it, but I was able to visit a local hospital (The Christie Hospital not too far from the University) which had a D-MAC which they kindly let me use.

Data General Nova

A research group studying fish vision within the Zoology Department had bought a mini-computer, a Data General Nova. I don’t recall the model but it may have been a 1200. It was to be used to run experiments in a lab close to where I worked. It was initially set up in the office I was using and I was given free rein to use it for a while.

This machine was 16-bits with ferrite core memory – I have no record of how much – maybe 16K or even 32K. It had a paper tape reader and punch and a teletype. I remember a FORTRAN compiler being bought with it, and this was in the form of paper tape.

To start the machine it was necessary to enter the boot loader by hand using the switches on the front panel. Then the loader program was brought in from paper tape and this could be used to load the compiler or compiled programs.

Using the teletypes on the ICL 1906A

I mentioned the ICL 1906A in my last episode. I found that there was a room of terminals and ASR 33 teletypes that were available to users of the central computing facilities. I learnt how to use the GEORGE operating system on these teletypes, and found out how to write and store programs and data on the ICL machine. The user interface used on the teletypes was called MOP (Multiple Online Programming).

Photo from Wikipedia
Teletype Corporation ASR-33
ArnoldReinhold [CC BY-SA 3.0], via Wikimedia Commons

It was possible to prepare work on the ICL via MOP and then to submit it as a batch job to the CDC 7600. As noted on Wikipedia, the version of GEORGE used at UMRCC had been modified to allow this.

It is now possible to run an emulation of GEORGE 3 on the Raspberry Pi!

Cyber-72

I also used another computer at UMRCC, though just in an exploratory way. This was the CDC Cyber-72. My memory of this machine is hazy now, and it may have been a Cyber-76, or perhaps the 72 was replaced by a 76 at some point.

This machine had terminals with good screens, and offered a variety of programming languages. I used it to experiment with APL, a programming language developed in the 1960s by Kenneth E. Iverson.

APL uses a range of symbols in the language. As I recall, the Cyber-72 version didn’t support these, but it was still quite usable. I was able to write a simple statistics program in it to prove it could be done, but I didn’t carry on using it thereafter since it seemed too limiting in terms of availability.

Funding

As I mentioned, I paid for my first year myself from money saved from working during my year out. I managed to get some grant funding for one or two further years, but this was not enough. I was lucky to get a part-time job within the Zoology department as a Laboratory Technician to help with funding.

This involved helping to set up laboratory sessions, but mostly I was called upon to do driving work. I’d ferry students about from time to time and sometimes buy or collect things for the department at various places around Manchester and its surroundings.

Notable events I now recall:

  • Taking a fellow postgraduate student to collect freshwater mussels in a stream just beside the Jodrell Bank radio telescope.
  • Collecting shellfish at Llandudno Bay in North Wales
  • A regular trip to buy maggots at a local fishing supplies shop
  • Collecting dead gulls around a reservoir (probably Audenshaw Reservoir) for a parasite check
  • Helping to catch fish (perch) with nets in various lakes for experimental purposes
  • Collecting large amounts of cow’s blood from a local abattoir to be examined for parasites

Photo from Wikipedia
Lovell Telescope, Jodrell Bank Observatory
Mike Peel; Jodrell Bank Centre for Astrophysics, University of Manchester. [CC BY-SA 4.0], via Wikimedia Commons

Photo from Wikipedia
Great Orme Panorama from Llandudno promenade
Velela [Public domain], via Wikimedia Commons

Hobbies

Building a calculator

I wanted to own a calculator at this time. I’d been seeing advertisements for the Sinclair Scientific in newspapers and the like for some time. I had learnt to solder at school - in a very basic way using a tinsmith’s soldering iron. This is a shaped chunk of copper on a handle that can be heated in a gas flame like a blowtorch and used to melt solder into a joint.

When younger I had also done some soldering at home, using my father’s electric soldering iron I think.

Around this time I bought myself an electric iron from Antex and a few accessories such as a soldering iron stand and a clip-on heatsink to prevent fragile items from getting too hot as they were soldered.

I bought a Sinclair Scientific kit for the advertised price of £9.95 and managed to build it without any serious mishaps.

Photo from Wikipedia
Sinclair Scientific
Aq42 [Public domain], via Wikimedia Commons

The device was a little odd:

  • it used Reverse Polish Notation to enter calculations
  • it had no decimal point but used exponential notation instead
  • it used repeated addition to multiply numbers and subtraction to divide them
  • other functions such as log and antilog were achieved by iterative methods so the calculator was very slow!

Check the links for the details if you are interested.

Moving on

I had spent over 4 years on the research for my PhD. Some of it I paid for myself, as I have told you, though I received a grant for part of it.

However, during that time I realised the research topic I’d embarked upon was going nowhere, and it gradually dawned on me that I was not cut out to be a researcher.

During the time I’d learnt a lot of stuff including a bit of Electronics, a lot of Biology and a great deal of Computer Science. I felt I’d be better leaving Manchester to see if I could get a job in IT somewhere, and this is what I did.

I’ll tell you more in the next episode.


  1. I still have some lengths of what I assume to be Dexion in my attic, acquired from my last employer who was throwing it away, and recently made some storage shelves with it!