Lister Start-O-Matic
Background to the DC Start-O-Matic system
This 'lighting plant' consists of a 2.5-kilowatt diesel-engined generating set, a 110-volt storage battery and an automatic control panel. It was built in 1947 and installed in a remote Dartmoor location where for decades it was probably the only source of electrical power, sufficient to run the lighting and general appliance load of an average house. The principle of the "floating battery" system used by the plant is quite straightforward; When the electrical load was light, in this case up to about five lamps or 500 watts consumption, the battery would meet the demand for a long period and there would be no need for the generating set to run. Heavier loads would be also be supported by the battery for long enough to get the set started. Once running, any reserve generator capacity would be used to recharge the battery. Provided the consumption followed a regular domestic pattern of varying heavy and light loads, the generator would spend enough time running to maintain the average state of charge of the battery near to maximum. What really makes this system practical is that the generator starts automatically when the load exceeds the threshold, regulates its output to keep the battery charging steadily and finally shuts down when not required. Apart from a monthly "equalising charge" to top the battery right up to full capacity, and routine maintenance as required on any generating plant, the system would look after itself. Start-O-Matic plant was available with ratings up to 20 kilowatts for large premises, and also with interlocked control of multiple generating sets. If two sets of unequal size were installed, the control gear would start whichever set (or both) was appropriate for the load.
It is worth noting that this system applies specifically to DC supply, which was obsolescent in 1947 when the plant was commissioned. At that time, an increasing number of electrical appliances were made for AC only, so there was a definite disadvantage to the DC floating battery scheme used on its own. AC cannot be stored in a battery, so two alternatives existed where AC was necessary; either generate AC directly, or generate and store DC but convert it to AC on demand. AC automatic start systems were made by Lister under the Start-O-Matic name, in which the generator runs all the time load is detected. This is a less efficient technique if small loads are run regularly, as generator efficiency falls when lightly loaded, and slightly inconvenient when the inevitable delay occurred as the set started up. It was necessary when using an appliance such as an electric drill either to leave a light burning or to operate the override switch, so that the set would not stop every time the switch was released. The alternative solution was to run a converter from the DC supply to provide AC for those loads that required it. This approach also reduces efficiency due to the power loss in the converter, but allows moderate AC loads to be used without the generator, and this method was apparently adopted with our plant. A DC-AC motor-generator of 800 watts rating was with the plant when it arrived, and there was an additional contact fitted to the control panel which may have been used to extend the remote starting function to the AC supply.
Operational sequence
The "automation" is controlled by six relays, three of which are fitted with oil dashpot time-delays enabling them to control the duration of the several steps in the starting sequence.
Limitations and eccentricities of the floating battery Start-O-Matic system
Under automatic operation the load is always connected to the battery. The battery voltage when charging must be higher than when discharging, thus the voltage delivered to the load must vary. In practice, it is not possible to charge the battery to 100% capacity without raising the voltage so high that the load might be damaged, typically 35-40% in excess of normal system voltage. In order to prevent the battery losing capacity through sulphation, which affects lead-acid batteries that are not kept fully charged, and to avoid the cells of the battery getting "out of step" in their charge status, it is necessary to start the set manually from time to time, and adjust the dynamo output to the full charging voltage of the battery. During the equalising charge it is important to disconnect any load that may be on the system from the charging voltage. A switch is provided to allow it to be connected to a tapping in the battery that delivers the proper system voltage during charging, enabling the supply to remain alive even when the full dynamo voltage would be dangerously high. Large floating battery systems, such as those in telephone exchanges, were equipped with a complex system of counter-emf-cell switching to limit the load voltage, but such extra expense is not warranted in a plant of this size.
The control gear on this plant does not start the engine in response to the battery charge state being low. If just one light were left on continuously, the battery would eventually run down because there would be no peak consumption periods to signal the need to start the engine. In this case, the automatic gear would not be able to turn the engine over and the plant would lock out. The remedy would be to start the engine by hand-cranking, and then let it run through a full equalising charge to restore the battery to normal.
Dynamos used to charge batteries are normally of shunt-wound construction, rather than compound. A cumulative compound generator can be made to give an exactly constant output voltage with varying load, but the series-field winding required to achieve this can cause mischief if the machine "motors" from the battery through it. When this occurs, the current through the series field reverses, and with it possibly the net magnetism in the machine field system. This can cause the machine to build up with reversed polarity at the next start, which would prevent it being able to charge the battery and would cause destructive arcing when the cut-out attempted to close. Note that the series field provided on the Start-O-Matic is connected with opposite polarity to that of a cumulative compound generator, and is cumulative when motoring; it cannot reverse the machine remanent field. The plain shunt machine avoids this risk but delivers a "drooping" output characteristic that would render the charge rate of a floating battery unstable as the load current varies. To counteract this the magnetic governor increases the engine speed (and hence the output voltage for a given setting of the field rheostat) with load, achieving the constancy of voltage without boosting the field strength. Accuracy of the voltage is critically affected by any adjustments to the system. When the plant was tested at the factory, a line was scored on the field rheostat bezel showing the correct setting for normal float operation. Lister strongly forbade any user adjustment to the governor whatsoever.
CS 5/1 Start-O-Matic No. 6413
Acquisition and preparation
We obtained the plant in remarkably complete condition. The pipework and cables had been cut in-situ prior to removing everything intact ready for collection. The accessories were all present and original, including the fuel tank and pipe, cooling tank and even the three-way water cock bearing the Lister name. We found the switchboards reasonably sound although there is some impact damage and corrosion to various parts of the automatic starting gear. The first job was to clean the crankcase, which was full of sludge and emulsion. It may not have received proper maintenance attention during the last phase of its working life, during which the lubricating oil had become diluted with fuel and contaminated with water. The sump was then filled with non-dispersant monograde oil, as required in an engine with oil circulation but no filtration. At this time the amount of wear on the main bearings, crank and con-rod was assessed; about twice the official Lister recommended maximum but acceptable for test purposes without overhaul. The exposed moving parts and valvegear were lubricated, as was the cylinder bore, prior to turning the engine over for the first time. After exercising the compression changeover valve to clean its seats the engine was found to have excellent compression. The cooling water passages in the block and head were found to be somewhat scaled but not so badly as to require treatment before testing. Lister engines of the period are fitted with washable fabric-tube fuel filters; this filter element required a thorough scrub to remove the solid layer of gum that had formed on it but was found intact and refitted. In internal inspection of the dynamo revealed carbon build-up on the commutator where the undercut between the segments had not been cleaned. This was remedied, the brushes were freed in their holders and the machine bearings were greased. A mouse nest was removed from the terminal box but there was no sign of damage to the insulation, which was tested and found excellent. The switchboard is divided between two panels; the general switchgear and meters on the larger and the automatic starting equipment on the smaller. It was apparent that a fair amount of repair work would be required to make the automatic gear operable, so the general switchboard was tidied up and connected to the generator in such a way as to allow manual control of the electric starting. The exhaust silencer was attached; cooling and fuel supplies were temporarily connected up using the original tanks and the injection pump was primed. A 110-volt battery was assembled from wet lead-acids and wired to the panel along with the dynamo.
See Dave and Lucien servicing the plant and setting it up for the test run. [Watch video]. Use your browser back button to return here.
First startup
The main switch was closed and the plant motored swiftly but failed to start, due to some air remaining in the injector line. After a couple of attempts the last bubble was bled out. On closing the compression change-over valve, which had been left loose to reduce cranking load on the machine during bleeding, the engine fired up immediately and swung the ammeter over to charge. Everything worked well; the revolving tappets revolved steadily, the engine governed smoothly, there was no sign of lubricating oil being burnt. After the engine had warmed up, full load was put on for long enough to warm the dynamo windings and prove correct operation; sparkless commutation was observed and the engine continued to perform well under load, just a trace of fuel-smoke being noticeable. It was then shut down pending the completion of the switchboard repairs.
See the start of the test run. [Watch video]. Use your browser back button to return here.
All content © copyright Electrokinetica 2007,2008 except where otherwise stated •
Valid XHTML 1.1
•
Valid CSS