The Sinclair C5 Electric Vehicle
Sinclair Vehicles Sinclair Research
Coventry England Cambridge, England
The Sinclair C5 represents the first commercial attempt to produce an electrically powered vehicle in high volumes. In the electrical and electronic aspects of the vehicle, as in most other areas, minimisation of component and assembly costs was a vital part of the design exercise. In many instances, a traditional EV approach had to be abandoned and design /construction techniques more closely allied to the light electronics industry were adopted.
The paper covers the development of the essential protection systems for battery and motor, the user instrument display and its associated semi-custom integrated circuit, the selection and testing of the cost critical motor power control system, the packaging of the electrical/electronic components into a form suitable for mass-production, and the battery charger design.
Vehicle technical data, including performance and operating characteristics are presented and discussed. Road load is quantified in terms of three constituent parts (mechanical, aerodynamic and gradient resistances) and is compared with the available tractive effort over the full vehicle operation range.
Illustration of the effects of various parameter changes are given with particular regard to the efficiency of utilisation of available battery energy and to operating range.
The Sinclair C5 is a semi-enclosed three wheel power assisted cycle. It features a polypropylene body and weighs 45 Kg including a 15 Kg lead-acid battery. It represents the first attempt to produce, in high volume, a vehicle aimed specifically to comply with the electrically assisted cycle legislation introduced in the UK in August 1983.
The UK does not currently enjoy the same liberal attitude to low-powered motor vehicles enjoyed by many of its European neighbours. Thus, even the lowest power mopeds require registration, road tax, insurance and a driving licence.
As a result of activity by certain interested parties, and after tests and recommendations from the UK Transport and Road Research Laboratory, a new category of electrically assisted vehicle was legally defined in August 1983. The key features were:
In addition, this class of vehicle could be driven by any one of 14 years of age and over. It did not require a licence, insurance or road tax. Furthermore no protective helmet had to be worn.
As a result of the legislation, several companies have begun to offer electrically assisted machines. These are predominantly based upon existing bicycle layouts. The C5 is the first example of a purpose designed three-wheel, semi-enclosed vehicle to appear. (see Figure 1)
Sir Clive Sinclair, as founder and chairman of Sinclair Research Ltd, has designed and marketed a range of personal computer, TV and other electronic products. Sir Clive's long-term personal interest in electric vehicles had resulted by 1983 in the construction of a number of prototype single-seat machines. These demonstrated the feasibility of meeting basic performance requirements, of speed and range, using existing battery technology. By adopting a single seat configuration, minimising weight and maximising efficiency, a realistic performance could be achieved. These developments were the basis of the C5.
Working capital for the project was raised through the sale of a 10 per cent shareholding in Sinclair Research Ltd, to various institutions. The Sinclair Vehicle Project, under the direction of the future managing director, assembled a project team to undertake the development. The team included numerous sub-contract organisations and individuals as well as potential component suppliers. Foremost amongst these was Lotus Cars who undertook chassis and transmission development. A total of some 100 individuals contributed to the development over a concentrated 12 month period culminating in pre-production prototypes and fully committed tooling by the middle of 1984.
The C5 will now be described in terms of its EV design parameters, manufacturing requirements and performance characteristics.
Electrical System Design Philosophy
C5 was designed to be used safely without prior instruction by a spectrum of users from 14 to 92 (the oldest to-date) years old. With the younger age group in particular, an element of deliberate abuse was expected and the electrical system design had to cater for this. In addition, essential component supplier liability could not be imposed without protection. In particular, the motor had to be protected from thermal overload and the battery from over discharge or incorrect charging. Furthermore, a battery master switch, circuit fusing and weather protection were all deemed necessary.
Figure 2 shows a block diagram of the C5 electrical system. It comprises:
The following sections deal with particular aspects of the electrical system in greater detail.
With legal weight restrictions and a target manufacturing price set, it was considered unnecessary to include any form of proportional motor control. A permanent magnet motor was chosen as allowing the legislation on powered top speed to be met simply by having a defined maximum no-load speed on the motor. High stall torque was considered important in achieving good hill climbing performance and rapid acceleration, the latter to give a measure of traffic compatibility for town use.
Given these decisions it was vital to protect the motor from thermal failure caused either through deliberate misuse or excessive overload. Three approaches were considered:
1) A mathematical model of the motor was proposed to using lumped constants (determined experimentally). The model was to be implemented using analogue circuit techniques in integrated form. However, it was found that without some form of real temperature feedback, the technique could generate significant errors when low rates of rise of temperature occurred. It also proved difficult to implement with any accuracy using the integrated circuit elements available.
2) A study and significant practical experimentation was made of a real thermal model using a resistive heater to simulate armature dissipation and a metal block surrounded by thermal insulation. The model included a thermistor whose resistance indicated the expected armature winding temperature of the motor. By mounting the model firmly to the motor case, the required feedback was obtained to eliminate long-term drift problems.
Although were a number of successful models were built, they would have been difficult to reproduce with sufficient repeatability and the effects of ageing on the thermal insulation were unquantified.
3) A still more direct measuring system using an non-contacting thermal probe was built which measured the air temperature inside a hollow portion of the armature shaft. It was considered essential to have a non contacting probe as project timescales precluded an extensive assessment of the wear characteristics of any rubbing seals.
Experiments showed good correlation between the deep armature temperature and that at the practical measurement point (Figure 3). The correlation was good enough to select this technique for the production C5.
A close tolerance thermistor was used in the probe configured as a potential divider. In conjunction with two voltage comparators, the first level warning and a second level power lock out feature was provided.
The probe system was incapable of providing protection in the case of rapid overload such as a deliberate stall condition. To cover these cases an integration of current with respect to time was used. This has a characteristic as shown in Figure 4 and provides stall protection at 140 amps in a few seconds.
The final back up, a resettable bi-metallic trip is mounted on the motor case, breaking the motor relay coil current at 70 º C +/ - 5 º C.
The C5 battery is a flat plate semi-traction lead-acid unit developed and manufactured for Sinclair Vehicles by Oldham batteries. The specification is as follows:
To prevent over discharge, a technique incorporating a compensated terminal voltage measurement under load was used. Circuits employed are part of a semi-custom integrated circuit and these also provide a visible indication of battery charge in the form of a five segment bar graph display. The onset of battery cut-out is preceded by a 20 second audible warning period, in order to provide sufficient time to complete any ongoing manoeuvre and make vehicle and driver safe. A block diagram of the battery condition gauge is shown in Figure 5.
A nominal 18 milliohms of battery impedance is assumed. The gauge can only decrement and contains a 20 second delay which must time out before a display decrement can occur. This eliminates spurious counting. The five equal incremental voltage steps span the nominal voltage range, 12.0 to 10.5 volts. The inherent non-linear voltage profile of the battery (under constant current discharge) gives an non-linear display with time. It therefore provides ample warning of the end of available battery capacity, and in a pedal assisted vehicle is quite sufficient.
Correct charging of the battery is fundamental to achieving a long life. A form of two-stage constant current followed by a constant voltage charging profile is used. The battery can be charged in situ or it can be removed and taken to the charger.
The charging profile is shown in Figure 6 and is implemented by a hybrid circuit with laser trimmed resisters driving 2 SCR's as a phase controlled current source. The whole assembly is potted both to aid heat dissipation and to fully weatherproof the unit. To LEDs indicate charging active and finished. A charge time of typically eight hours is required from full discharge.
Manufacture and Test
The complete instrument pod is shown in Figure 7. It is tested in conjunction with an operator using automatic test equipment on a "bed of nails" fixture. The tests include voltage, current , frequency and time measurements to verify full functionality. The test time of approximately two minutes, which includes one operator adjustment, reflects the extensive functions provided by the semi-custom IC. There is a limit to the circuit speed-up factor that can be employed whilst still maintaining the measurement tolerances. The operator interacts with the test equipment and verifies correct display and of audible transducer operation.
After manufacture and test by a specialist sub-contractor, the circuit boards are mounted in the two-part polypropylene case and shipped to the vehicle assembly plant for installation under the C5 canopy.
The control box contains a mixture of high current motor circuits (up to 140 amps), intermediate duty systems for vehicle services such as lighting etc and low-level signals associated with current measurement (1 millivolt/amp). Traditional electronic components are assembled and tested by a specialist sub-contractor (Figure 8). The power components are then added. These include heavy current interconnections using 4 and 5 mm hardware, with machine crimped ring and 6.35 mm spade terminals. The motor relay is a specially developed unit from Ital-Amec in Italy with a thermal rating superior to that of the motor. Contact arc suppression is vital and provided in the optimum circuit location by a rectifier diode rated at the Quorn amps. It's transient capability is more than adequate and reflects the reappraisal of "the fitness for purpose" approach adopted throughout C5. The relay is specified to withstand 10,000 make/break cycles at 140 amps, in the specified circuit configuration.
A current measurement shunt is provided for driving both the battery condition gauge compensation circuit and the user display of motor load. A simple stamping from a resistive alloy sheet, with a nominal resistance of 1 milliohm is used. Riveted power connections and 6.35 mm push-on sense leads are used. All motor wiring uses 6 square millimetre cross section wire insulated with automotive grade PVC. The continuous rating of this conductor size exceeds that of the motor by a substantial margin and provides the most cost-effective solution as low-cost machine crimped terminations can be used.
The finished control box PCB (Figure 9) is assembled into a two part polypropylene moulding plus access cover. The halves are ultrasonically welded together and combine a tamper-proof cavity with a user accessible portion where the heavy current connections to the vehicle wiring harness are made. Other connections plug-in before the access cover is added. A downward pointing cable entry and splash grommet ensure no ingress of water whilst allowing natural breathing to occur.
The fully assembled vehicle is subject to a functional test at the end of the production line (Figure 10). Operator verification of instrument displays is the only manual intervention in an automated test sequence. The test includes:
An extensive quality audit test is carried out on a sample basis. After test the C5 is packed in its dispatch carton along with charger/toolkit/instructions and loaded directly on to a container truck for despatch to regional warehousing.
Acceleration and Braking
Table I provides some basic dynamic properties with an average weight driver on board, and without pedal assistance. Low-speed acceleration is particularly lively, providing C5 with standing start performance equivalent to some small cars. These figures confirm the viability of the lead-acid battery for this class of vehicle, in which basic battery mass is typically only 15 % of gross vehicle mass.
The 35 Ah (at the 5-hour rate) battery provided with C5 delivers a useful 15 to 28 Ah in service on the vehicle, depending on the particular driving resistance present and the duty cycle imposed. Figure 11 shows how sensitive range is to the level of resistance present (proportional to current drawn) assuming the battery is discharged under steady driving conditions, and no pedal assistance. In order to provide a link with the on-board information available to the driver, current drawn is also depicted in the figure in terms of the motor load indicator information available while driving. Thus, if it cruise the motor load indicator sits in the right most green segment (with perhaps occasional and equal excursions into the left most green and amber segments) then the expected range on a single battery would be approximately 13 miles (21 km). As the capability exists to carry two batteries, also shown in the figure is the range available from two batteries when discharged separately, and when connected in parallel. It can be seen that the value of the second battery, in terms of range, is equivalent to approximately 1.5 batteries when discharged in parallel with the first due to the useful capacity increase provided by halving the discharge rate of each battery. This figure of 1.5 contrasts with a value of approximately 0.9 for the second battery when discharged separately from the first, a figure representing the cost of carrying around an extra 15 kg of ballast.
In line with its lively acceleration from rest, C5 has basically good hill climbing ability. However, this ability is usually limited in practice by motor thermal conditions. Leaving these limitations aside for the present, Figure 12 shows the theoretical tractive effort available for a wide range of operating conditions, excluding the benefit of any pedal power applied. The figure has been prepared using maximum values of battery voltage and vehicle system efficiencies. (worst-case values would produce a tractive effort line of approximately 75 % the level shown.)
The road load curves are for zero gradient and reflect a value for rolling resistance coefficient of 1% together with aerodynamic characteristics derivedmfrom coast down test results. Both resistance components (rolling and aerodynamic) vary widely in practice, and those used for the figure are "worst-case" values. Tests have indicated a "best-case" rolling and aerodynamic drag resistance coefficients of 60 % of these values.
As indicated in the figure any grade steeper than about 2.5 per cent, and of sufficient length, is likely eventually to induce a thermal cut-out. The question of gradability should therefore always be addressed with regard to the length of the climb as well as his grade. In addition to this, motor temperature at commencement of climb also plays a significant role. For the case of an initially cold motor, without pedal assistance, figures presented previously may be combined to generate a chart which predicts the length of any grade of hill negotiable before thermal cut-out intervenes. Figure 13 shows such a chart for the case of a C5 with a lightweight driver.
Starting with, for example, a 10 % grade on the left hand ordinate, the dashed lines indicate the flow of information which may be derived from the chart. Thus the speed up the 10 % grade is 13 kilometres per hour, the current drawn is about 60 amps, the duration of climb before thermal shutdown is seven minutes and the length of hill negotiated in this time is about 1500 metres. Other grades may be similarly analysed.
The discontinuity in the current verses time curve in the figure, at 80 amps, represents the transition from the stall sensing circuit to the thermal protection system. This discontinuity is at its greatest for the cold motor condition depicted, and it reduces with increasing motor temperature. A motor already close to its thermal limit at commencement of climb would display virtually no discontinuity as shown by the chain dashed curve, and the length of climb achieved would be greatly reduced, as shown.
The centre of gravity of the C5 with driver installed is approximately 75 % of the wheelbase behind the front wheel. It is also located very low down compared to conventional tricycle configurations, and therefore permits the generation of quite high lateral accelerations before wheel lift occurs. In practice, after familiarity with the vehicle, it is quite easy to achieve "slip before tip" by suitably shifting driver body mass in turns.
The C5 front brake has been engineered to make it difficult to lock the front wheel. By doing so, steering control is maintained at all times, even during the emergency braking. Furthermore, as only one rear wheel is braked, emergency braking involving locking of the rear brake has no tendency to induce a vehicle spin, as is often the case with cars, etc.
Although vulnerable to impact by a target vehicle, as is the case for any small, lightweight vehicle, the C5 has been developed to provide its occupant with a high degree of protection from impacts occurring within its own performance spectrum. It has been established in reverse acceleration crash simulation studies at the MIRA Hyge s the led laboratory, that no injury is likely to result from frontal impact into a barrier at any speed up to the C5's own level road maximum. We believe that C5 is the only form of passenger-carrying road vehicle, powered or not, for which this claim may be made.