This is one of several pages relating to the history of the automatic totalizator, its invention in 1913, the inventor George Julius and the Australian company he founded in 1917 which became a monopoly (later an oligopoly) in this field. This page contains a technical description of the historical Julius tote at the Eagle Farm Racetrack Museum. This is a history only non commercial page. If you wish to start from the beginning then go to the index .
|A description of the Julius totalisator in the Eagle Farm Racecourse Museum|
The Eagle Farm Racing Museum Julius Tote
Page 1 First and Last can be seen in the centre of the tote frame under the sign protruding above the top.
Page 2 An electromechanical shaft adder can be seen on the third adder window from the right hand side.
The description of the public odds indicator drive can be seen in the tenth window from the right hand side.
The RESET READY ON indicators above the frame are described in the "Place Pool Reset and Ready Switches" section below
Charles Barton, the last Chief Engineer of this system would have loved to see this!
This is the machine room or in contemporary parlance the computer room. This is a large scale, multi user, real-time system. These systems existed long before the invention of the world’s first electronic computer. This one had 128 electromechanical Ticket Issuing Machines attached to it distributed in totes around the track.
A Julius tote was installed in Longchamps France in 1928 with 273 terminals and the one in White City London was upgraded to support 320 terminals. One of these systems was demonstrated in Sydney in 1920 capable of supporting 1000 terminals and a sell rate of 250,000 tickets per minute.
There is a school of thought that these early Australian totalisators were the first computers. The director of the London Science Museum wrote in his New Scientist article dated 29 October 1987 titled “A sure bet for understanding computers”; “The Julius tote with its automatic odds machine is the earliest on-line, real-time, data processing and computation system that the curators of the museum have identified so far”.
This frame houses a Win and Place totalisator. The front totalled the Win pool and the rear is duplicated for the Place pool. It supported a field of 24 runners. Inside the windows there are electromechanical shaft adders one for each runner plus one for the grand total. The adders have odds calculating devices attached to them which drive barometer indicators on the outside of the East and West walls of this building for public display. These indicators have two vertical channels for each runner one for the Win and one for the Place marked with imperial odds. Bands rise up from each runner’s odds calculator, win and place, through the roof of the frame and across the ceiling in both directions to the Eastern and Western indicators, then outside to move visible indicating strips in the channels mentioned above. A commission gearbox subtracted the commission from the pool for the displays.
This system was superseded in 1979 by a PDP11 based digital computer totalisator also manufactured by Automatic Totalisators. It was long thought that the last of these Julius electromechanical totalisators to cease operation was at Harringay London in 1987 until an email was received in 2005 from Caracas asking how to make adjustments to their Julius tote to bring it up to modern day standards. At the time of the email, that system had been in operation for 48 years. The lifespan of computing systems has certainly changed over the years!
These once common Australian Totalisators have disappeared mainly under bulldozer tracks. We are fortunate that this example still remains, a reminder of an Australian achievement, courtesy of the Queensland Turf Club.
This shaft adder is of a similar vintage to the ones used in the Eagle Farm system the main difference is that this is a 3 shaft adder and the ones on display are a 2 shaft adder
The heart of the shaft adder is the epicyclic gear train. These adders have two horizontal shafts with epicyclic gears visible near the top of the adder. The adders were powered by a horizontal drive shaft running the length of this frame driven by DC motors at the northern end. The adding shafts in the adders were driven by a spring for each shaft, which was wound up by the drive shaft. When the adding shaft spring was fully tensioned a clutch disengaged the shaft from the main drive. As energy was removed from the spring by escapement wheel movement resulting from bet traffic, the clutch engaged again to wind the adding shaft’s spring back up. The adding shafts have escapement wheels on them, 6 on the front shaft and 2 on the rear. These wheels moved one tooth at a time when the associated solenoids visible underneath the adding shafts were activated by a ticket-issuing machine recording a transaction. The number of teeth on the escapement wheel determines the value of the bet. The more teeth the lower the value. This system supported £5 £1 10s 5s bets which was translated to $10 $2 $1 and 50c when decimal currency was introduced.
Proponents of the electromechanical systems would boast that this system could do something that the next generation computer tote could not. It could record multiple bets simultaneously whilst the new computer tote which existed prior to the days of multiprocessing had to record bets sequentially albeit at a rate that made it all look instantaneous. In other words any combination or all of the solenoids on a shaft adder could be activated at the same time. The epicyclic gear train took the different value bets resulting from the different escapement wheels rotating and activated the display counter to keep a running total of investment. In the event of a drive failure to an adder, all adders were fitted with automatic cut outs, activated by a mercury switch, to prevent loss of bet registrations and raise alarms, illuminating one of the lights above the adder window for attention. Automatic Totalisators even manufactured its own plastic plugs and sockets visible on the right hand side at the bottom rear of each adder.
At the end of each row were the grand total shaft adders with the associated winding gear for raising the vertical lift sliders on every runner’s adder to represent the grand total. The horizontal component for each runner was produced via the adder odds chain sprocket wheels, at the very top of the adder, which let their horizontal sliders out in accordance with the investment on that runner. The hypotenuse arms formed right triangles for each runner as they connect the vertical and horizontal sliders. This produced an angle between the vertical and the hypotenuse, for each runner’s adder, which was always related to perfect odds for that runner. On this angle was a pulley arrangement, which drove some cams operating a small motor, which in turn powered the large barometer style odds displays mentioned on page 1. The hypotenuse arms mentioned above can be easily seen behind the adders or by looking down the length of the frame at either end. They can then be used to identify the vertical and horizontal sliders.
For remote locations such as the centre of the racetrack or a loft, odds transmissions were achieved by the use of a variable resistance, mounted at the important angle mentioned and utilising the Wheatstone bridge principles. The remote receiver sensed any out of balance transmitted by the adder unit and drove itself and the odds display until the bridge centre leg potential was once again null and therefore equaling the transmitted odds.
The distributors are at the top of racks 2, 3 and 4. They are a circular piece of equipment with four concentric rings. Two studded metal contact rings surround two continuous ring contacts with a rotating arm spanning the diameter of the outer set of studs. This arm rotated when the equipment was in operation and electrically connected one of the continuous rings with the inner circle of studs one at a time and the other continuous ring connected with the outer set of studs in sequence. One set of studs serviced the Win pool and the other the Place pool. The function this performed was to allow multiple Ticket Issuing Machines (TIMs) to be attached to a single escapement wheel solenoid in the adders. This was achieved by sequencing the supply voltage in the continuous rings to each stud and consequently the stud’s attached TIM, which provided a circuit to the attached solenoid, in the runner handle selected adder, if the pool selection knob on the TIM was pushed down indicating a sale. In contemporary terminology the scanner provides an enabling pulse, which will allow a transaction cycle if the TIM has a sale pending. The Win/Place knob on the TIM selected the appropriate inner or outer set of studs in the distributor. In the event that multiple TIMs attached to the same adder solenoid, in every runner’s adder, selected the same runner, meaning the multiple TIMs were now attempting to trip the same physical solenoid, the scanner serialised access to the associated solenoid, into the appropriate number of activations, so every bet was registered. There are 16 studs on each of these scanners allowing 16 TIMs to be attached to one shaft adder solenoid. Each of the 8 scanners is associated with one of the 8 solenoids in all the shaft adders, the actual adder being selected by the TIM’s runner handle. This provided a capacity for this system of 16 X 8 giving 128 TIMs. Optimum scanner rotation was between 90 to 120 RPM.
There are overlap relays beneath the scanners connected to the scanner studs. Once a sell transaction was started by the TIM and when the distributor selected that TIM, a set of contacts in the overlap relay connected the supply voltage to the transaction circuit and its own coil, keeping itself activated, which held the supply voltage, after the distributor arm had passed its stud, to the TIM and associated adder solenoid, until the machine cycle completed. Beneath the overlap relays are switch banks, which could isolate a TIM in the event that it had a fault and did not release the transaction circuit.
There are some computer hardware analogies to these distributors, a serial line multiplexer and a sequencer; however the functionality of these devices was replaced in their successor by a polled protocol, which curiously also supported 16 TIMs per line.
It is interesting to note that the Sales Bell relay in the right hand rack at the top, along with two other relays are implemented using Mercury switches. This bell marked the start and stop of betting. Mercury is a metallic element and consequently a good conductor. It is a liquid at room temperature. An arced tube of glass contains mercury and a central electrical contact and a contact at each end of the tube. If the tube is oriented past the horizontal in either direction the mercury pours to the low side and makes a circuit between the low-end contact and the center contact. A restriction in the flow implemented a delay ensuring the bell always rang for the same length of time.
This system had 128 of these Ticket Issuing Machines (TIMs) distributed in totes around this track as follows. Main house 48 Ledger Stand 6 Sub House12 Jackpot tote 8 Members 28 Lady Members 10 Front Public Stand 6 Top Public Stand 10. The Main house mentioned above is the downstairs of this building.
It was at these windows where these TIMs were installed that the punters queued up to place their bets. The word queued, used above, is used rather liberally, photographs of Brisbane racetrack tote outlets in the era that these totes were in operation often had a standing room only crowd swamping the totes. There is half a century of totalisator history based on machines like this one before the inception of the TABs, which diminished the racetrack attendance in conjunction with the introduction of other forms of gambling.
A J8 Ticket Issuing Machine
The handle rotates in an arc and is positioned at the required runner number. The knob on the top of this handle moves backwards and forwards along the longitudinal axis of the handle. The forward position selected the Win pool and the backward position selected the Place pool. When the same knob is pushed down the transaction was recorded and a ticket printed which was ejected from the machine at the slot in the top. A J8 ticket is shown below.
A ticket printed on a J8 ( Animation above )
The value of the bet was fixed in this system, determined by which escapement wheel solenoid the TIM was attached to. Later Julius totes supported selectable value at the TIM. The HAGUT printed on the ticket shown, is a security code, which was different for every race and was automatically selected at race increment. Barrels with the code type were installed in the machines for the meeting. The L down the right hand side of the ticket is an added security feature and was different for each race as was the ticket colour. For this reason the printer paper in every TIM had to be changed every race. The code words and ticket paper security letters were kept secret until the race day manager instructed the tote house managers to open the race day code security envelope.
A trip relay in the TIM terminated the transaction cycle. Its coil and the transaction circuit are in series with a normally closed set of its own contacts so when these contacts open they cut the supply to its coil, resetting itself and the transaction circuit. By the time the current sufficiently built up to activate the trip relay the same current would also have tripped the adder solenoid which is also in series with the trip relay. (Current lags the voltage in an inductor as quantified by the L/R time constant). The overlap relay coil is also in this circuit and consequently it also reset. As the trip relay’s open contacts stopped the current that activated it in the first place it reset and its contacts closed again but it did not reactivate as the overlap relay was no longer providing voltage to this circuit, as it was in the reset state. The TIM was now ready for the next transaction cycle.
Internally, the runner handle moved two contacts over two sets of studs arranged in the same arc with 24 studs in each set. Each outer arc stud is connected to the associated Win adder solenoid for this TIM bank, in the runner’s shaft adder with a number matching the number of the stud. Hence stud one attaches to runner one’s Win shaft adder and stud two attaches to runner two’s Win shaft adder etc. There is the same arrangement for the inner arc studs and the associated Place adders. If runner 15 was selected with the handle, the knob pushed forward for the Win pool and the knob pushed down to indicate a sale, the Win contact attached to the knob, contacted stud 15 on the Win arc completing a circuit enabling the scanner pulse to travel to the solenoid associated with this TIM bank in shaft adder 15. There is a lot more to this transaction circuit however that is well beyond the scope of this introduction.
The switch on the top of the TIM was used to turn it on and off. The last position on the runner handle arc was used to print a test ticket.
When a transaction was registered on the TIM by pushing the Win/Place selector knob down the runner selection handle and this knob were locked in place until the transaction cycle was complete. If there was a fault and the transaction cycle did not terminate correctly the handle release button on the top of the TIM was used to release this lock, after the problem had been investigated, so that the TIM could continue operation.
These machines had to be moved between the Eagle Farm and Doomben gallops, Albion Park trots, Gabba greyhounds and Ipswich gallops as there were insufficient machines to populate all the tracks. Today’s TIMs are still moved for the same reason. Sometimes during the Winter Carnival we feel exhausted moving the large number of PC based modular TIMs; after having to use a hydraulic wheelbarrow to move the J8 recently I will only consider ourselves fortunate in future.
I lament that Charlie Barton, Chief Engineer of this and other Julius totalisator systems in Brisbane is no longer with us to see this system preserved. It was his dream to preserve and possibly restore one to an operational condition for public display. Alas, it was my fate to be Chief Engineer of the first on course digital computer based totalisator systems for the Brisbane metropolitan clubs, which brought an end to the operation of these magnificent machines. I find it ironic that someone who never worked on these electromechanical totalisators nor indeed saw one working is left to write about it. This would have been very different just 10 years ago. It is now 2007, 29 years since this system last operated.
Acknowledgement: Thanks to Ron Findlay for assisting me with questions I had regarding the J8. Ron used to work on the J8s and continues to work on the current generation of TIMs.
The console above is at Harold Park in 1958 however it is the same as the one at Eagle Farm.
When the large knob in the middle of the Tote Control Console control panel was rotated clockwise the race number was incremented. This caused a barrel with the race number type to be rotated to the next number in every TIM. All the TIMs clanked in unison as this selection was made. The scratching switches introduced an open circuit in the escapement coil circuits in the adder corresponding to the scratched runner.
There is an emblem on the large knob mentioned above with Premier on it. This was the product name that George Julius gave to his totalisator. This emblem is also visible on the J8 TIM. Automatic Totalisators is visible engraved below the Knob.
To give some idea what it was like in this room when the system operated the following paragraph is an extract from a company magazine called Tote Topics. This article was written in 1968 and compares electromechanical totes with the then new computer based ones.
In the machine room of an electromechanical totalisator there is motion, constant motion, and noise. With betting in progress, the constant chatter of the escapements blends with the purring of the counters and the low rumble of the drives to give a quite characteristic sound. This sound, both in intensity and pitch, indicates to the experienced totalisator operator, even more clearly than his eyes, the state of the queues outside and the conditions around the selling houses. He scarcely needs a clock, so accurately is he able to predict from the betting pattern the time to the start of the next race. The equipment consists of row upon row of shafts and gears and escapement wheels and mechanical counters. At first sight it seems entirely mechanical as the electrical portions are buried deep inside.
It is interesting to contemplate the human side of this system. The following is a transcription from an audiotape recorded by Neville Mitchell, a long serving Automatic Totalisators Manager and excellent source of information about these systems.
The mystique of the machines was something I experienced, particularly in Melbourne, not so much in Sydney. The men who operated the four major tracks there had been with these machines since 1936 and on the decommissioning day, I saw emotions that were quite unbelievable. They were seeing the last day of operations with this sort of gear. The strictness with which the engineers ran these systems was somewhat akin to a military operation; they really had a lot of power. They had a lot of routines set down and to be an apprentice in those days was a lot of sweeping the floors and making the tea for a long long time before you actually got your hands on any piece of equipment. And I believe in the early days in Melbourne, if an apprentice was seen with his hands out of his pockets in the machine room, he would get a swift slap around the ears. The same thing applied in New Zealand. I read some stories from there and I actually knew a couple of the engineers and they applied the same very very strict mode of operation on their set-ups. They were extremely proud of these machines and some of them spent all of their, what you would call, idle time in routine maintenance and polishing of brass and things like that, that made these machines absolute showrooms.
Go back to the index Go to the bottom of the page
You will notice GAJ part numbers on these drawings. These are George Alfred Julius’ initials. Additionally the draughtsman’s name is Noble. This is Norm Noble, a long serving Automatic Totalisators employee. I remember him well. As the company supported operations, engineers in the field would often require urgent shipment of replacement parts out of the normal working hours of the Head Office and Factory. Long before the days of mobile phones there were times when it was not possible to contact the people responsible for providing support. As these parts requirements were usually urgent it is fortunate that there always seemed to be people willing to go the extra mile and help out in areas for which they were not responsible as they realized that operations was the coal face of the company and was the place our products were judged. Norm was one of these people. I could always rely on him to go to work in Sydney on weekends, public holidays, or otherwise out of hours to send me anything I needed. There is a photo of the ATL drawing office in the "Memories of the Factory" chapter that predates my time which has Norm Noble in it, unfortunately he is facing away from the camera.
These drawings were made at a time long predating computer drawing, CAD and CAM applications. I recall draftsmen in the 1960s in their drawing offices standing, or sitting on bar stools, in rows, at their large drawing stations with their angled drawing surfaces and spring loaded large setsquares that could sweep across the whole drawing area. Earlier photographs indicate that the drawing office draftsmen who made these older drawings sat at normal office desks. It is interesting to note, that documentation for these old systems, seems to always have been released to a particular person. The document from which these drawings are extracted has a title page indicating that it was released to S.Huss in the Assembly section on 12/9/1950. I have seen other manuals in the form of bound books with the name of the employee to whom it was released embossed on the cover.
Following is a transcription of a page from this document to save clarity and download time.
|1 Type Wheel Peg & Frame||23 Other Side of Issuer Wiring Former|
|2 Type Wheel & Win, Place, Show Arm Assembly||24 Selector Quadrant & Wiring|
|3 Locking Rod Assembly & Ribbon Bracket||25 Win, Place, Show Arm, Anchor & Spring. Show Switch Bumper Assembly. Pool Selection & Bracket.|
|4 Intermediate Gear Bracket & Gear||26 Condenser & Clips|
|5 Handle Assembly. Handle Stop & Brush Holder||27 Motor & Brushes, Chain & Split Link.|
|6 Taper Pinning Operation||28 Ribbon Feeding Operation|
|7 Platen Assembly||29 Issuer Box & Hinge &Wiring Clips|
|8 Win Place, Switch Assembly & Slide Rod||30 Plastic Issuer Handle , Pin & Circlip|
|9 Paper Feeding Assembly||31 Electrical Setting Details|
|10 Trip Coil Assembly||32 First Test|
|11 Ribbon Rewind||33 Taper Pinning Operation|
|12 Quadrant Supports, Issuer lifting Handle & Posts & Cover posts||33A Ticket Issuer Chute|
|13 Handle Release, Double Pole Switch & posts, Test Switch, Handle Release Lever & Spring||34 Covers & Horse Number Segments.|
|14 Test Coils, Wiring Brackets & Cover Catches||35 Final Test|
|15 Latch switch & Show switch||36 Spray Finish issuer Box.|
|16 Cam operated Counter switch||37 Attach Nameplate|
|17 Guillotine Lever, Anchor & Spring Printing Lever, Anchor & Spring||38 Clean & Inspect Box|
|18 Value Slide Lever & Spring|
|19 Rotary Switch, Retaining Posts & Value Leaf Switch Assembly|
|20 Value Release & Spring|
|21 Veeder Assembly|
|22 Issuer Wiring Former & Brushes|
Following are image extracts from the document.
One of the Barometer Indicator Drive Units in the Eagle Farm Racing Museum
The scale and weight provide a local indication of the odds. The more the sensing pulley pulls the cable the more it lifts this weight and vice versa. There is a corresponding rotation of the pulley at the top of this device. This light cable mechanism is only capable of rotating this low resistance pulley. It is not capable of moving the heavy indicator band and weight, in the associated runner channel of the indicator on the outside wall, the distance required. This rotary motion of the top pulley on this device turns the cam wheel with the red line across it. When the odds on the outside indicator match the odds displayed by the cable weight here the line is horizontal. As the odds calculating mechanism changes the odds the microswitches attached each side of the cam wheel are activated by cam movement sensing the required direction and amount of movement required causing the motor to drive accordingly. The motor which is capable of driving the external indicator ribbons and weights, drives a second pulley in this device, an arc of which is visible near the base. This pulley has two metal bands attached to it that run through the roof of this frame across the ceiling in both directions to other pulleys on top of the east and west walls. These pulleys in turn move wider metal bands which move a visible weight up and down the channels of the associated runner which are part of the odds indicators on the outside of the east and west walls and the position of the weight indicates the current odds for the associated runner. You may have noticed the circular variable resistances located near the top of the hypotenuse arms. These were used to drive an infield indicator.
A postscript not included in the text displayed in the museum. This is just to provide additional information relating to the photograph provided. The two aluminium vertical bars visible behind the indicator drive unit are the transport mechanism for the vertical slider. The upper pulley visible near these bars is the third pulley attached to the vertical slider as described above. The pulley visible below and behind the third pulley is the angle sensing pulley attached to the extension of the hypotenuse arm as described above. For the particularly astute the cable that should be running around these pulleys has been pulled out of the third pulley and should not be running directly to the ceiling from the sensing pulley on the extension of the hypotenuse arm. This is probably why the wheight on this unit is dangling below the bottom of the scale.
The indicator panels located on the centre of each side of the Win/Place machine frame were to indicate the status of each betting pool. The "Win" or "Place", "Reset", "Ready" and "On" lamps were operated from control switches located on each end of the machine frame and on the remote control console unit. When the engineers had reset the adder counters to zero for each race they would turn on the "Reset" switches. The senior engineer would then check the counters, gearboxes, indicators, etc to ensure that the machine was ready to open and, if satisfied, would turn on the "Ready" switches. The person allocated the responsibility to set the field, scratchings, gearbox settings, etc on the remote control console unit for each race, (sometimes the Tote Manager or the Secretary), after checking all status lamps were indicating the correct settings, would then open the betting by turning on the Win and the Place pool switches. The machine room staff would then know that the betting was "On" from the machine frame centre panels.
The two indicator lamp units, located on each side of and at the end of the machine frame, were to indicate the settings of the automatic display gearbox settings. On the Win gearbox end, "Minimum' or "Maximum" and on the Place end, "Minimum" or "Maximum" and "Two Dividends" or "Three Dividends". Some frames also indicated "Win" or "Place", "Reset", "Ready" and "On" as a second status display on each end of the machine frame. On some systems they had a "Mean" gearbox setting in addition to the normal "Maximum" and “Minimum" ratios.
The Eagle Farm Racing Museum end view of the Julius Tote
The Place Commission Gearbox description following can be seen in the window in the door of the frame.
The paragraphs titled "Page 7 The J8 assembly drawings" above, can be seen in the image above the window.
The J8 assembly drawing images above can be seen below the black indicators.
The 2div 3div and "pla max" and "pla min" status indicators are described in "Place Pool Reset and Ready Switches" section
|Previous page||Go to the index||Top of the page||Next page|