Engines : Rover M and T-Series

History will overlook the Rover M and T-Series engines. It’ll record it as the Rover 800 engine, but Rover’s mid-1980s twin-cam four acted out a diverse range of roles, and proved the soundness of its basic design by its ready adaptability when faced with unpredicted emissions legislation.

Robert Leitch tells its story…

Anchored down in Anchorage

There was surely no more significant meeting in the entire Rover/Honda era, than the ‘secret’ rendezvous which took place in February 1981 in Anchorage, Alaska between Michael Edwardes and other senior BL managers, and their Honda counterparts.

Edwardes states:’That meeting gave birth to the XX Project, announced in November 1981. ‘The timings are significant. The Triumph Acclaim was still eight months away from going on sale when the Anchorage meeting took place, yet the two companies’ faith in their working relationship and the future worth of their collaboration was already so strong that they were ready to begin work as what Edwardes described as ‘a partnership of equals’, on a project which was utterly crucial to the reputations and ambitions of both participants.

Honda and British Leyland compare notes

In 1981, Honda’s largest car was the newly-launched, second-generation Accord, enlarged half a size over its predecessor and powered in its largest-engined form by a long-stroke 1829cc straight four. A completely new set of components would be required for a larger car, which would follow the Accord into every major market in the developed world.

It is worth recalling the scale of the task the HX/XX project was expected to deliver. In its Honda and Rover guises the new car would inaugurate two North America-only prestige brands. Honda had only recently entered the C/D sector with the second-generation Accord, and their acceptance into a higher class still was far from certain.

The Japanese large car sector, dominated by the Toyota Crown and Nissan Cedric/Gloria, was notoriously traditional, yet Honda, nearing the pinnacle of their reputation as technical innovators, would never have countenanced emulating the anachronistic engineering of these two ultra-conservative behemoths.

America in its sights

Austin Rover may have had America once again in its sight, with the Sterling brand, and would also be offered an unprecedented doorway into Japan with the Saitama-built Rover 800s, yet Europe was inevitably to be the battleground in which the Cowley-produced 800 would stand or fall.

The sole major XX engineering component for which Rover would have full responsibility was the four-cylinder engine. Honda had no need for such an engine, as only V6-engined Legends were to be produced.

For Rover to face the new breed of European high-efficiency upmarket saloons, spearheaded by the 1982 Audi 100 C3, a new in-line four with far greater refinement and technical sophistication than anything in their existing inventory was an essential prerequisite.

A tantalising glimpse

In November 1985, Austin Rover released details of the M16 engine, and showed pre-production 800s to a select group of journalists invited to the Canley Design Centre. Their hand may have been forced by Honda’s confident introduction of the Legend at the Tokyo Motor Show. With the Japanese part of the project fully in the public domain, the only significant details of the Rover which were not known were its exterior styling and interior design, and the British-developed four-cylinder engines.

That the Legend would be introduced well in advance of the Rover had been known for some months, but Rover must have felt an obligation to allay concerns that Honda were dominating the project, or that the collaboration was being frustrated by dissent over quite fundamental design issues and disparate and incompatible methods of working.

The ‘private view’ emphasised beyond doubt that the Rover body’s visual identity was quite different from that of its Japanese counterpart, in marked contrast to the ‘licensed’ Acclaim and first-generation SD3 Rover 200.

The plan takes shape

Persuading the doubting critics of the effort which had gone into the new engine required the disclosure of a detailed specification, as well as photographs of the new unit, neat and purposeful looking with ‘Twin Cam 16 Valve’ emblazoned proudly on the milled ribbing of its alloy cam-covers, imitating the Japanese style of the time.

Rover M- and T-Series engines: The M-Series team, led by Roland Bertodo (front row, centre) at its launch in 1986
The M-Series team, led by Roland Bertodo (front row, centre) at its launch in 1986

Notwithstanding its O-Series derived bottom end, the new engine was classically elegant rather than conspicuously inventive in its design yet, for its time, remarkably uncompromised in its specification. Highlights revealed at the end of 1995 included:

  • Twin overhead cam 16-valve alloy cylinder head
  • Hydraulic tappets
  • Electronic engine management
  • Lean-burn combustion capable of operating on air/fuel mixtures of 18:1
  • Choice of 120bhp ARG developed single-point fuel injection or 140bhp Bosch multi-point fuel injection versions

Reports at the time speculated that the engine would be available in capacities smaller than the two-litre version first revealed. As well as a 1.8-litre special to take advantage of the British Exchequer’s demarcation line between working tools and status symbols, a 1.4-litre to be used in smaller Rover products was mentioned. Speculation about the smaller K-Series engine was already lively with the expectation of a maximum capacity of 1.3-litres.

O-Series – A rough diamond shows new facets

It is appropriate at this time to comment on the M-Series’ relationship to the O-Series. The first true Leyland engine, the Rover SD1’s PE146/166 six, was acclaimed at its 1977 introduction for the ingenuity of its design and potential for future development. The O-Series, which appeared the following year, had a far less comfortable reception. First appearing in the heavy Princess and then the unsophisticated Marina did not help its cause, and subjective impressions suggested that early versions compared poorly with their B-Series predecessors in refinement and driveability.

Criticism was also directed at the ‘stack’ arrangement of the alloy head and combined cam cover and camshaft bearing carrier, an arrangement considered to be a production engineer’s dream, but a service mechanic’s nightmare.

Further disappointment was expressed at the O-Series’ restricted dimensions, the result of a requirement for compatibility with the Princess’s in-sump transmission and B-Series tooling. An increase in bore from the largest B-Series’ 80.3mm to 84.5mm delivered the required 2.0-litre capacity, but in 1978 the ‘big four’ was making its comeback, with well-regarded contenders in the 2.3-2.6-litre capacity bracket from Citroën, Mitsubishi and Volvo. With its siamesed bores and offset con-rods, the O-Series block was effectively at the practical limit of its capacity.

A firm foundation

Move forward eight years, and the O-Series had established itself as a strong, reliable and efficient power unit. The supposed disadvantages of its long stroke and closely centred bores had turned out to be a positive benefit when used transversely with an end-on gearbox in the Montego and Maestro.

The siamesed bores and heavily-braced cast iron block (which nevertheless, at 108lb, weighed 19lb less than the 1798cc B-Series block) provided immense structural strength, a huge benefit in the turbocharged petrol and diesel variations. A move towards smaller capacity high-efficiency power units, encouraged by tax regimes across Europe, and widespread adoption of forced induction had by the early 1980s stalled the charge of the big-capacity fours.

In any case, necessity, to paraphrase the words of the 18th century Marquis de Vauvenargues, had delivered Rover’s product planners and engineers from the embarrassment of choice. A ‘new from oil filler to sump plug” engine was beyond Austin Rover’s means, and the closure of the engine manufacturing facility at Canley brought to a close production of the Triumph slant-four and PE146/166 SD1 six. In the post-Edwardes infrastructure, only the Longbridge O-Series line had the production capacity to meet anticipated demand for the XX and other M16 applications.

The Rover 800 unveiled

In July 1986, the Rover version of the XX, badged as the 820 and 825, went on sale. Details of the M16 were at last revealed in full. Atop a largely familiar cast iron block the new cylinder head build-up followed the O-Series pattern with the top of the head casting on the camshaft centreline, and ladder castings, separate in this case for intake and exhaust cams, carrying the upper halves of the camshaft bearings. The head itself used the relatively novel low pressure casting technique in preference to gravity casting to minimise the requirement for subsequent machining.

Unlike the O-Series, where the camshaft bearing cover also served as the cam cover, necessitating substantial dismantling to gain access to the camshaft, the M-Series benefited from separate cam covers. The twin camshafts acted directly on hydraulic tappets, the fitment of which had by the mid-‘80s become normal European industry practice. The same could be said of the adoption of a toothed belt for the cam drives which also drove the water pump. The distributor was located horizontally on the opposite end of the engine from the cam drives, driven off the end of the exhaust camshaft.

The combustion chamber shape was the result of rigorous research by BL Technology at Canley. The expectations were high – a power output of 70bhp per litre, compliance with anticipated emissions legislation by means of highly efficient lean-burn design and power and torque delivery appropriate to a large, relatively heavy, luxury car.

Influenced by the Dolomite

The pent-roof combustion chamber shape, with a central spark plug, was arrived at through use of single-cylinder engines on flow rigs, yet was remarkably close to the design used in Lewis Dawtrey’s pioneering Dolomite Sprint engine in 1973. Roland Bertodo (above), ARG’s Director of Powertrain Engineering, said in CAR August 1986 that the M16 design team were aware of the efficiency of the Triumph engine but had to establish the reason: ‘Triumph people stumbled on the fact that it gave very good economy, but they didn’t quite know why.’

The key to the ‘Triumph’ combustion chamber’s effectiveness was found to be the ‘barrel swirl’ pattern it created, allowing more complete mixing and sweep rates than other designs considered.

Largely driven by emissions legislation, fuel injection and engine management technology had advanced rapidly in the years leading up the introduction of the M16 engine, and the first M16 engines offered a choice of systems. The 138bhp multi-point injected version used a Lucas L-type system similar to that already in use in the Montego EFi, but the 118bhp entry-level cars had a throttle-body (TBI) system which was home-grown.

Single versus multi-point

Although only a single injector was used, limiting the maximum power output, the TBI system was more sophisticated than the Mpi in some areas of its design, using a Honda-style speed density system to measure air mass, rather than the more commonly found hot wire or flap. The mapped ECU incorporated idle stablisation, maintaining a constant speed regardless of load, and its knock control had the facility to retard ignition in individual cylinders. The potential for efficiency of the TBI system was demonstrated by its ability to run petrol/air mixtures of 18:1, whereas the MPI ran richer at 16.5 or 17:1.

In both its forms, the M16 impressed on paper – the 820e’s Granada and Carlton competitors offered eight-valve, single-camshaft carburetted units of some antiquity. A moderately informed analysis of the power and torque figures would have stirred some unease. The TBI figures of 118bhp at 5600rpm and 119 lb. ft. of torque at 3500rpm gave a healthy advantage over its Ford and Vauxhall competitors.

The multi-point engine’s figures of 138 bhp at 6000rpm and 131 lb. ft at 4500 rpm immediately suggest an engine more suited to a hot hatchback or sports coupe than a 1290kg luxury saloon. Low gearing giving 21.7mph/1000rpm compensated partly for the torque deficiency, but at the expense of high-speed refinement and fuel consumption. Roland Bertodo was defensive in the choice of torque curve, pointing out that it was flat enough and suggesting that its critics should ‘try it out and see if we’ve picked the right choice for torque. You can always tell me if you don’t agree.’

Weighed in the balance – and found wanting

The critics duly spoke. In August 1986, CAR carried the cover caption. ‘England expects, but Austin Rover struggles to deliver.’ Gavin Green had the following to say about the M16 MPi: ‘The twin problems with the M16, in its more powerful multi-point guise are that it’s not particularly refined. Nor is it specially lively. And the engine’s busy presence is insufficiently isolated from the passenger compartment.

‘Noisier than it should be in the low to medium rev range the motor becomes increasingly and intrusively audible when extended towards its rather optimistic 6500rpm red-line. By 6000rpm. it has had enough. As with the 2.5-litre V6, it’s short of mid-range pulling power, so must be revved hard to deliver real urge.

‘Maximum torque comes in at an absurdly high 4500rpm, but over 4000rpm the M16 does provide reasonable performance. The problem is that the effective band of performance – owing to mid-range lethargy and top-end hysteria – is too small.’

Sold too soon

Much of the poor impression created by the early-production XX cars was the result of being brought to market too early – ARG did not have the resources to match Honda’s development cycle. The comparison chart shows that the M16’s torque curve was not greatly different from its German competitors.

The XX, although of similar size and weight to its SD1 predecessor, was a very different type of car, and the result was a gulf between expectation and reality. Just because the power output of the 820i’s 2.0-litre four slightly exceeded that of the Rover 2600, there was no reason to expect it to deliver that power in the manner in which a softly tuned engine with two more cylinders and over half a litre more capacity would.

Another publication made the highly apposite observation that the XX was more of a successor to the P6 than the SD1, recalling the character of the 2000TC’s peaky high compression four.

The M16 compared

Manufacturer Rover Rover Audi BMW Ford Saab Vauxhall
Date 1986 1986 1985 1982 1986 1986 1986
Designation M16 SPi M16 MPi EA827 M20B20 2.0 EAO B202 Family II
Application 820e 820i 100CD (C3) 520i (E28) Granada 2.0i 9000i Carlton CDi
Configuration 4 in line 4 in line 5 in line 6 in line 4 in line 4 in line 4 in line
Bore x stroke (mm) 84.5 x 88.9 84.5 x 88.9 79.5 x 86.4 80 x 66 90.8 x 76.95 90.0 x 78.0 86 x 86
Cubic Capacity 1994cc 1994cc 2226cc 1991cc 1993cc 1985cc 1998cc
Head/block material Alloy/iron Alloy/iron Alloy/iron Alloy/iron Iron/iron Alloy/iron Alloy/iron
Valve gear DOHC 16v DOHC 16v SOHC 10v SOHC 12v SOHC 8v DOHC 16v SOHC 8v
Main bearings 5 5 6 7 5 5 5
Compression ratio 10.0:1 10.0:1 10.0:1 9.8:1 9.2:1 10.1:1 9.2:1
Injection ARG single-point Lucas L-type Bosch K-Jetronic Bosch L-Jetronic Bosch L-Jetronic Bosch LH-Jetronic Bosch multi-point
Engine management ARG Lucas Bosch Bosch Bosch Bosch Bosch
Power bhp (DIN)
@ rpm
118@5600 138@6000 136@5700 123@5800 115@5500 135@5500 122@5400
Torque lb/ft
@ rpm
119@3500 131@4500 133@4800 125@4000 118@4000 127@3000 129@2600

Comparison of the torque figures suggests that the Rover’s critics may have been harsh in their judgement, at least in comparison with the Audi, BMW, and Ford competitors. Fuel injection has failed to transform the Ford’s end-of-line Pinto unit, and the BMW engine, the only six-cylinder, is also well behind the field, at least in terms of on-paper outputs.

The Saab engine, from 1985 but originating from the Triumph slant four, is remarkably similar to the Rover engine in its principles – both replaced a simple in-line SOHC head design with an all-new 16-valve DOHC top end incorporating hydraulic tappets.

Finding its place

By the time the five-door fastback 800s were introduced in mid-1988 CAR’s attitude had softened, beguiled by the Vitesse with its newly-enlarged 177bhp Honda V6, and their GBU buyers guide, less subversive than it once was, summed the range up as:’Improving, Britain’s best mass-made executive car’. It was a somewhat disingenuous compliment, given that production of the Rover’s Ford and Vauxhall rivals had moved to Germany many years before.

In late 1990, the throttle body injection 820e and Se versions were dropped from the range. Despite the fanfare which had greeted the system’s introduction four years previously, it was deemed unsuitable for development to meet the incoming emissions regulations, and the multi-point injection engine became the entry-level four cylinder. The carburetted SOHC O-Series engined 820 fastback which had been offered since the 5-door’s introduction in 1987 was also discontinued.

Rover 800 Turbo – a spectacular swansong for XX

In July 1991, with the XX replacement, codenamed R17, only a few months from sale, Rover sprung a surprise in the shape of the most powerful production XX. The 820 Turbo used the M16 engine with a water-cooled Garrett T25 turbocharger and intercooler. Development was contracted to Tickford, which had previously developed the limited edition Maestro Turbo.

The production target for the run-out special 820 was the same 500 as the Maestro, but the extent of modification and upgrading of the base engine suggested that this exercise was more than a short-term generator of showroom traffic.

New inlet and exhaust manifolding, valve springs, exhaust valves, fuel pump and injectors, spark plugs and specially designed Mahle pistons were fitted, and, significantly, all 820 Turbos had a new exhaust system with a three way closed loop catalytic convertor.

Extracting the most from the M-Series

The 820 Turbo’s power output of 177bhp at 6100rpm compared well with rival forced-induction engines, but more impressive was the torque curve, a description scarcely appropriate, so linear was the delivery of the 159lb/ft maximum figure from 2500 to 6500rpm. The car’s far from discreet TWR-designed rear spoiler and side skirts (below) may have divided opinion, but the transformation of the engine gained universal praise.

The new-found torque turned an engine which always had to be worked hard into an effortless and flexible ground-coverer, which paradoxically was a far more willing revver than the normally-aspirated version, with the side-benefit of the twin-cam’s high-end harshness being largely muffled by the turbocharger and new exhaust. Top speed was 137mph, with an 8.0 second 0-60 figure but, as with the Maestro Turbo, the turbocharged 820’s forté was its prodigious fourth gear acceleration capability, with sub-seven second figures for the 30-50mph and 50-70mph increments.

The limited edition four-cylinder car’s price of £23,950 undercut the Honda-powered Vitesse and Sterling by over £3500. The cynical might suggest that Rover was possibly making a statement about the price Honda was asking for its less powerful 2.7-litre V6.

Playing by the new rules

When the design parameters were set for the M16, it was widely assumed in Europe that future vehicle emissions legislation would favour lean-burn engines which reduced pollutant levels by ultra efficient combustion, rather than treating emissions downstream using a catalytic convertor, as was required in the USA, Japan, and other more highly regulated territories.

Not only Rover, but also Ford and Peugeot were investing heavily in new lean-burn engine ranges for the 1990s, but could not have predicted the surge in public concern about environmental matters which arose in the late 1980s, prompted by the effect of acid rain on forestation and a greater understanding of ozone depletion.

Burgeoning support for Green Parties across Europe put pollution control at the centre of the mainstream political agenda, and resulted in the adoption of far stricter standards for the Euro I regulations, applicable from January 1992, than had been widely anticipated.

Making it greener

The result was a set of regulations which made catalytic converters the only option. Although the rigorous new standards were denounced at the time as Green hysteria, and resulted in much abortive work on lean burn engine designs, they did have a sound scientific justification. Nitrogen Oxide (NOx) was widely recognised as a major human health risk, leading to respiratory problems and other illnesses. The high temperatures associated with lean-burn combustion increased NOx levels by breaking down nitrogen which then combined with oxygen.

Automotive industry history will record that the cataclysm which lean-burn apologists predicted never happened, and the regulations effectively gave us engines which were generally more powerful and efficient as well as cleaner.

In order to meet the demand of certain European countries, particularly Germany, which offered tax reductions for cars which met the new standards before they became mandatory, Rover produced a catalyst-equipped M16 from the late 1980s. Far from being strangled, power output fell by only 3bhp to 135bhp. The new standards did concentrate Rover’s best minds on the M16’s potential for improvement, and the outcome, revealed at the end of 1991, was the T-Series, or T16.

R17 and T-Series – more improved than new

The Rover 800, in its re-skinned R17 guise, was unveiled in November 1991. The visual transformation provoked mixed reactions – the revival of the Rover grille and the high tapering tail were thought to be well-handled, but the XX doors, re-tooled but unchanged, sat uncomfortably with the curvier shape and bore manifest witness to the financial strictures placed on the update.

In engineering terms, the major story was the T-Series engine. Rover was at pains to state that virtually no major part of the revised engine was shared with its M16 predecessor, yet capacity, materials, bore and stroke dimensions, constructional principles, and the twin-cam 16 valve top end with pent-roof combustion chambers with centrally located spark plugs, all remained as before.

The principal changes which transformed the M16 to T16 are as follows:

  • Three-way catalysts fitted to all engines
  • Motorola MEMS (Modular Engine Management System) with greatly increased processing power capable of 40,000 instructions per second
  • Revised combustion chamber shape based on the K-Series design
  • Longer inlet tracts, smaller plenum chamber, and revised cam profiles to improve low-end torque
  • Internal block and head breathing system based on K-Series principles
  • Eight-counterweight crankshaft in place of four previously used. In combination with reduced weight pistons Rover claimed a 17% reduction in out of balance forces
  • Patented combined drive for water and power steering pumps
  • New cambelt only requiring replacement at 100,000 miles.
  • Redesigned sump to reduce noise and vibration.

Power output was 136bhp at 6000rpm, down 2bhp on the non-catalyst 820i. The figure all eyes were on was torque – 136lb ft at 2500rpm, 2000 rpm less than the slightly lower figure delivered by the M16 in the same installation.

Small changes, big improvements

Roland Bertodo, interviewed by Richard Bremner in CAR July 1988, said that ‘People drive the torque curve, not the power curve’. The revised engine demonstrated that his designers now understood this, with class-leading fourth-gear acceleration figures and a new found flexibility.

The R17 820’s gearing was raised only slightly. Acceleration and top speed were competitive with 2-litre class rivals, as was fuel consumption, although the new engine failed to match the spectacular frugality of its lean-burn predecessor which, given a moderately gentle right foot, could deliver mpg figures in the high thirties. Given that its basic construction was fundamentally unchanged from that of the M16, it should not be a surprise that the 1991 engine’s noise and vibration suppression was not its strongest suit.

At least the re-profiled torque curve ensued that there was far less need to enter the zone where harshness manifested itself. The eight-counterweight crank helped to a certain extent, but it was no substitute for the contra-rotating balance shafts which were becoming an increasingly prevalent feature of competitors’ large-capacity four-cylinder engines.

Vitesse Turbo – T-Series hero

In the spring of 1992, the R17 Vitesse appeared, with a turbocharged T-Series engine in place of the Honda V6 used in XX Vitesses. Power output matched the limited-edition Tickford 820 Turbo at 180bhp, as did the maximum torque figure of 159lb/ft at 1800rpm.

Like the Tickford, the new Vitesse was praised for its effortless ground covering ability, although the Honda heritage of short travel suspension ensured it would never challenge the fastest hot hatchbacks on the twisty back roads. Good value was a strong suit. At just under £20,000, no five-seater could offer anything close to the Vitesse’s performance.

In August 1994, coinciding with the introduction of the 620ti, the 820 Turbo’s output was boosted to the same 200bhp as its newer stablemate. It should be noted that the reason that the Tickford’s output was limited to 180bhp was thought to be uncertainty about the ability of the PG1 gearbox to cope. Two years of experience with the higher output engine in the 220 Turbo must have given Rover the confidence to standardise Turbo outputs at 200bhp.

T-Series compared

Manufacturer Rover Alfa Romeo Audi BMW Ford Renault Volvo
Date 1991 1991 1991 1991 1992 1991 1992
Designation T16 2.0 Twin Spark EA827 M50 B20 I-4 (N-9C) J Series
Application 820i 164 Twin Spark 100 2.0 16V (C4) 520i (E34) Granada 2.0i 25 TXi 850 2.0 20v
Configuration 4 in line 4 in line 4 in line 6 in line 4 in line 4 in line 5 in line
Bore x stroke (mm) 84.5 x 88.9 84.0 x 88.5 82.5 x 92.8 80 x 66 86 x 86 88 x 92 81 x 77
Cubic Capacity 1994cc 1962cc 1984cc 1991cc 1998cc 1995cc 1984cc
Head/block material Alloy/iron Alloy/alloy Alloy/iron Alloy/iron Alloy/iron Alloy/alloy Alloy/alloy
Valve gear DOHC 16v DOHC 8v DOHC 16v DOHC 24v DOHC 8v SOHC 12v DOHC 20v
Main bearings 5 5 5 7 5 5 6
Compression ratio 10.0:1 10.0:1 10.8:1 10.5:1 10.3:1 9.3:1 10.3:1
Injection Lucas
Ford MPFI Bosch
Engine management Motorola
Motronic ML
K Motronic
M3.1 DME
Ford EEC IV Renix Bosch
Power bhp (DIN)
@ rpm
138@6000 148@5800 140@6000 150@5900 118@5500 138@6500 143@6000
Torque lb/ft
@ rpm
136@2500 137@4000 136@4500 140@4700 123@2500 130@4300 136@3800

The Rover’s torque figures demonstrate that, in this respect at least, it had gone from being class laggard to class leader. New engines from Ford and Volvo had raised the stakes in the class, and BMW had substantially re-worked their small six with a 24 valve twin-cam head.

Audi was offering what was effectively a Golf GTI 16V engine in the 2.0-litre class, with their smallest five now a 2.3-litre. Renault had the only three valve per cylinder offering, but the Alfa Romeo engine is possibly of greater historic interest. Directly derived from the 1954 Giulietta’s twin-cam all-alloy unit, in its final incarnation, dating from 1987, Alfa Romeo’s engineers eschewed a four-valve conversion in favour of two spark plugs per cylinder, controlled by a sophisticated Bosch engine management system. A simple form of variable valve timing was also a feature.

The other M16/T16 applications

Rover R8 200/400

In the forever-lost domain of the Honda-Rover partnership, nothing was ever quite what it seemed. Casual observers of the 1989 Rover 200 and its Honda Concerto twin, which had already been on sale in Japan for over a year, could have been forgiven for thinking that the companies’ relationship had reverted to the licensing strategy of the Acclaim and SD3.

Nothing could be further from the truth. Although the Concerto’s detail design and engineering was carried out in Japan, 20-30 Austin Rover engineers were working with the Honda team at any one time, and lessons were being learned from the British partner in packaging, torsional rigidity and visual presentation.

This was manifestly evident in the car’s restrained elegance externally and its very Rover-like interior appearance, albeit achieved with mainly synthetic materials. Externally, every aspect of the Concerto’s appearance, especially its wrap-around flush glazing, suggested a far closer kinship to the Rover 800 than its Japanese range-mates. The Concerto was never more than a marginal product for Honda, and its designers appear to have set to the task of creating their own Rover with enthusiasm and success.

Anglo-Japanese convergence

Below the surface a further twist in the plot emerges. In the course of the 800’s development ARG’s engineers had reluctantly adopted Honda’s idiosyncratic double-wishbone front suspension in place of a simpler MacPherson strut system. With the YY/R8 there were even more compelling reasons for the British car using the MacPherson strut system and, in this area at least, Rover’s demands prevailed. Japanese-built Concertos featured a double-wishbone system, whereas all examples produced at Longbridge, whether Honda or Rover badged, had MacPherson strut front suspension.

The packaging benefits of the chosen suspension were highly relevant to Rover’s needs. As well as their home-grown K-Series engine and the 1.6 litre Honda D-Series unit, there was a supervening need to accommodate bulkier, larger capacity powerplants. This had arisen from the veritable explosion in demand for diesel passenger cars in much of Europe from the mid-1980s onwards, and also the need to have a competitor in the profitable upper hot-hatchback sector.

The reasons for Rover’s choice of the indirect injection Peugeot XUD diesel engine over their own direct-injection MDi unit, a close relative of the M16, are not a matter for discussion here, but the immediate success of the R8 in 1.4- and 1.6-litre petrol form meant that there was no was no need for indecent haste in expanding the range.

Range expansion

In June 1991, the diesels arrived. The M16 engined-R8 era began in August 1991, which was rather late for Rover, between the arrival of 2-litre rivals from Ford, Vauxhall and Volkswagen among others, and the imminent major revisions to the M16 which created the T-Series. The supposition that the M16 was better suited to a relatively light high-performance hatchback than a large saloon was largely vindicated.

Peak power continued to be delivered as the red sector approached, to the accompaniment of considerable noise and harshness, but less bulk meant that good progress could be made without scaling the untamed eminences of the engine’s rev range. In any case, the 140bhp Rover engine was far more refined and flexible than the 216GTi’s twin-cam Honda D-Series unit. In August 1991, the price difference between the two was £645, the premium for the British engine provided an additional 10bhp and an entirely different character.

The three-door 220GTi was never a major player in the 1990’s performance hatchback market – Rover’s brand image and their unwillingness to emulate the brash egregiousness of their rivals’ styling possibly worked against them. M16 and T-Series power continued to play an important part in the R8 story, featuring in every body variant until they were supplanted by K-Series derivatives in the mid-1996 rationalisation of the Coupe, Cabriolet and Tourer variants, which were to continue to be produced for a further two years.

The fastest Rover ever

Ending years of speculation about the car known as Project Tomcat, in October 1992 the Rover 200 coupe was unveiled. The well-proportioned design by Gordon Sked’s team at Canley affirmed Rover’s R8-era confidence and their new-found capability to use flexible production methods to exploit potentially profitable market niches. The engine line-up was topped by two T-Series variants, the familiar 136bhp 2.0-litre, and an enhanced turbocharged variant with power output boosted by 20bhp to a round 197bhp.

Torque was also increased to 171lb/ft at 2100rpm, and the result was the fastest-ever Rover with a claimed 150mph top speed and a 0-60 figure of 6.2 seconds. The additional power was achieved through alteration of the cam timing and re-mapping of ignition and fuel delivery on the engine management system.

Relatively high gearing, and power delivery deliberately programmed to be linear rather than peaky, imbued the engine with an unexpectedly refined character, but criticism was heaped on the chassis’ inability to cope with its prodigious output, despite the adoption of the Zexel-Gleason TorSen differential which was serving the Audi Quattro and Lancia Delta Integrale so well.

In mid-1996 the range of specialist R8 Coupe, Cabriolet and Tourer variants was rationalised and the T-Series option dropped. A 145bhp 1.8-litre VVC variant took the place of the turbo. More temperate in its performance, the K-Series-powered Tomcat was seen as a better balanced offering at the time, and still has an enthusiastic following, but the occasionally intemperate full-blooded T-Series-engined Coupe remains widely regarded as the definitive example of the breed.

Rover 600

The story of each of the cars produced during Honda and Rover’s collaboration appears as a parable about the state of the relationship itself. In 1993, it seemed as if Rover were finally able to offer the car they had always wanted, their own distinctive version of the hugely successful Accord. Unlike the SD3/Ballade and R8/Concerto, which were built on the same Rover production lines, the 600 and Accord were each to be produced in their own makers’ factories, located less than 40 miles distant from each other.

Launched with a Honda-produced line-up of 2.0- and 2.3-litre four-cylinder petrol engines, the 600’s mechanical specification appeared to represent a return to the days of the Acclaim and SD3, with minimal engineering intervention from Rover, yet, paradoxically, Richard Woolley’s masterly re-skinning of the Ascot Innova, which first appeared in Japan in March 1992, was arguably the styling zenith of the Honda/Rover era.

Fears that Rover’s engineering input to future jointly produced cars would be marginalised were partly allayed by the news that the forthcoming L-Series ‘best practice diesel” would be offered in both the 600 and Accord. Appearing in early 1995, this engine was seen as essential to the European success of the Swindon-built Accord, and was the first instance of a Rover engine being used in a Honda-built car.

British engines delayed

The diesel 600 and Accord were not ready for sale until early 1995 but, in August 1994, Rover introduced the T16 turbo-engined 620ti, at the same time upgrading the 820 Turbo to a full 200bhp. The notion of Rover selling a petrol-engined SK2 was probably entirely at odds with Honda’s rigid definition of the project’s parameters, but the logic of using the T-Series, whose block was substantially similar to the diesel was clear for all to see.

The turbocharged 600 was the only version to use the T-Series, all other petrol-engined variants used Honda engines. The 197bhp Rover was accorded an enthusiastic reception. The turbo engine was already well regarded, while upgrading of the springs, shock absorbers and anti-roll bars at last delivered a chassis able to cope with its class-leading power and torque.

Regrettably, the home market buyers were more impressed by prestigious German badges than swashbuckling British engineering, and the self-effacing 620ti sold in relatively small numbers. It remained in the range until the 600 was replaced by the 75 in late 1998. Incidentally, the 620ti’s list price always undercut that of the slower and thirstier Honda-powered 623GSi, sometimes by as much as £3000, a possible case of Rover, under their new masters, biting the hand which once fed them.

Land Rover Discovery

From the autumn of 1993 Land Rover offered the T-Series in an entry level Discovery, designated 200Mpi. Critics were not slow to question the pairing of a relatively highly-tuned 2.0-litre petrol engine with the Discovery’s two-ton bulk and sizeable frontal area.

Justifying the option as more than a tax-break special, Land Rover was at pains to draw attention to the 136bhp engine’s torque characteristics, with maximum torque developed over a wider speed range than the V8 or diesel options. On-road performance, at least on paper, bettered the diesel by a small margin, but the 2.5 TDi’s fuel consumption was 50 per cent lower, easily justifying the £500 list price premium over the four-cylinder petrol alternative.

The buying public were unconvinced by the 200Mpi and it sold only in small numbers, until it was unceremoniously dropped from the range around the end of 1997.

Rover HH-R 400

Introduced in March 1995, a year into the BMW era, the HH-R Rover 400 was the last of the Honda-Rover collaborations, and is notable for being a largely Honda-designed car with (apart from the curious exception of the 1.6 automatic) a completely Rover engine line-up.

The T-Series was first offered as a 420 in February 1996, offering a useful 26bhp more than the K-Series-powered 416. The HH-R was judged a disappointing successor to the R8, and over-ambitious pricing exacerbated the criticism. That said, the powerful 2.0-litre engine justified a £900 premium over the 1.6-litre, lending character as well as performance. The engine itself was unchanged from its 800 installation in state of tune, but Rover took advantage of the 420’s lighter weight to raise gearing, to the benefit of refinement and fuel efficiency.

The arrival of the 45 in January 2000 marked the end of T-Series production. Behind the new number and slightly startled expression of the 75-like frontal treatment, the big engineering story was the replacement of the 2.0-litre four with a KV6 unit of the same capacity. Another name change and facelift was to take place before the HH-R/KV6 combination demonstrated its true mettle but, for the T-Series, the end of the road had arrived.

Morgan Plus Four

Produced from 1950 with Standard-Triumph engines, the Morgan Plus Four in its 1960’s Triumph TR4-powered guise was replaced by the Plus Eight with Rover V8 power in 1968.

No direct equivalent was offered by Morgan until 1985, when they revived the Plus Four designation in with a Fiat twin-cam-engined car which filled the gap in their range between the Ford CVH-powered 4/4, and the Plus 8. The Italian engine was replaced only two years later by a Rover M16.

Upgraded in 1992 to T-Series power, the Plus Four continued in production until 2001, when Rover could no longer supply the twin-cam engine and the designation was dropped from the range, to be revived in September 2004 with a brace of 2.0-litre Ford Duratec-powered two and four-seater models.

Weathering the storms:
an assessment of the M16 and T-series

The M16 engine was conceived by state-owned BL when the Honda era had scarcely begun, and went into production two years before the company’s ownership passed to British Aerospace. It survived almost until the end of BMW’s ownership.

Despite inauspicious beginnings it was, typically of the company, a remarkable achievement – a near state-of-the-art engine on a shoestring budget. The effectiveness of the transformation of the O-Series into M16 was a powerful demonstration of Canley’s ability to design and productionise a modern engine, highly necessary when BL’s Government paymasters would have more readily outsourced engine design to Honda than fund the development of the K-Series.

In mid-life T-Series form, the engine was at its most impressive. Caught out by the unanticipated strictures of the Euro I emissions regulations, Rover’s engineers took the change from lean-burn to catalyst in their stride, helped by the fundamental ‘rightness’ of the design, and the outcome was a better engine with the M-Series’ torque curve flaw decisively repaired.

It is worth noting that, by the early 1990s, there was already unease about the continuance of the Rover/Honda partnership – with their Swindon facility under construction, Honda needed Rover far less than Rover needed Honda. The T and K-Series engines would have served as invaluable lifeboats if Honda ended the collaboration and supply of vital components.

The T-Series’ decline began with the arrival in 1995 of the 80mm bore 1.6 and 1.8 litre K-Series. Surprise and delight that a great small engine had been transformed with such facility into what seemed to be a great medium-sized one was soon tempered as trade hearsay about head gasket failures reached the public domain. The problems of the open-deck damp-liner K-Series did much lay waste to a hard-won engineering reputation, encouraging speculation as to whether the gap in the range could have been better filled by smaller capacity T-Series units.

Death and transfiguration

The January 2000 Euro III emissions regulations, or rather BMW’s unwillingness to adapt the T-Series to comply with them, ended the life of the engine. If all had gone according to plan, they, and the larger K-Series engines would have been replaced by Valvetronic BMW engines from Hams Hall in 2001. Without doubt an updated T-Series, particularly in turbocharged form, would have been a huge benefit to Phoenix-era MG Rover but no plans appear to have been considered to revive it, although half of the engine continued in production as the L-Series diesel.

The strength and adaptability of the engine has to this day made it a favourite of tuners and kit-car builders, as a lower cost and more readily available alternative to the Ford Cosworth YB and Vauxhall C20LET twin-cam fours. With the turbocharging groundwork completed by the manufacturer, 250bhp is readily achievable using the standard internals and Garrett T25 turbocharger. Replacement of the pistons, conrods and crank, along with work to the head and cams and a larger turbocharger running at unspeakable boost levels has been known to provide outputs of 500-600bhp, an achievement probably beyond any K-Series or Valvetronic derivative.

Robert Leitch


  1. It is possibly worth noting that aparently while in BMW ownership that the T-Sereis allegedly produced the fewest warrentee claims per unit than any engine within the whole BMW group

    • Interesting article and from my personal perspective, quite accurate. However, the M and T Series both had a major flaw involving HGF (no need to explain that abbreviation on this site!).

      Once was a party to a fleet order of approx 50 R8 420 SLi’s for a mobile phone co in North London.
      they had a virtually 90% failure rate within 2 years /50k miles.

      As a result, the T Series engine was banned from their fleet in any spec.

      Was a shame as their fleet manager was a huge fan but just couldn’t defend it any further.

      We moved over to K Series with predictable results…

  2. The T series was the best engine Rover designed.

    if they had developed it further rather than relying on the flawed K series then maybe things may have turned out differently for them.

    • Well, the 1.8 K series fully dressed and wet IE: all components like manifolding and fluids weighs around 110KGS (subtract a few kilos of no aircon is fitted.

      The T series being of Iron / Alloy build is considerably heavier, from memory I think it tips the scales at 140kgs.

  3. A little of the increased weight would, however be down to the extra 200cc or so.

    Therefore does anyone have any idea how the T series compares weight wise with a two litre KV6?

  4. Does anyone know what is needed to fit a t series non turbo into a zs 120, ie wiring etc, much appreciated

  5. Find it a bit strange the non-turbocharged M/T-Series was only able to reach 135-138 hp as opposed to nearer 150 hp or 75 hp per litre, so as to match the 2.0-litre Alfa Twin-Spark or the 2.0-litre 16-valve engine in the mk3 Golf GTi.

  6. Reference the turbo 180/200PS engines and the ‘flat torque curve’, it is an entirely ‘artificial’ shape because the electronics were set to chop off the top of what would have been a very tall mountain ! This had to be done to preserve the transmission – though some strengthening of the PG1 box was in fact done before allowing the power to go up to 200 PS.
    If you look at torque curves for many other turbo engines, you will find a similar truncated shape for exactly the same reason.

    The reason why the the HHR 1.6 Auto used a Honda engine and transmission was simply to meet a Honda requirement that at least one model in the range had to exactly parallel the Honda equivalent – it acted as a sort of reference datum. Rover obviously chose the lowest-volume derivative for this role, thus maximising their own engine volumes.

    • I agree that the HHR 1.6 auto was a contractual requirement between Honda and Rover. To such an extent that Honda temporarily shut down all communication on HHR because Rover wouldn’t agreee to the unrealistic volume demands for this rather niche product. Such was the level of trust between the two companies at the time…

  7. One of my first jobs at Canley in the summer of 1981 was to scour the stores to find all the parts that had been delivered for the first prototype build of ‘O 16-valve’.

    They’d been ordered and the programme was then put on hold. Surprisingly, well to me anyway, pretty much everything was still there. Not that the storemen actually wanted to let go of them…

    The reason for introducing these engines in place of Honda ones was almost entirely down to the prices being charged by Honda. Honda were always going to make a profit from their Rover collaboration and did so through licencing costs on the vehicle architecture and the sale of powerunits.

    BMW did the same at the sale of Land-Rover – after all, ‘business is business!

    • Getting anything out of the Stores at Canley was always a challenge!
      I recall the KV6 was justified purely on the short payback generated by no longer having to buy Honda V6s.

      • Re. the cost of Honda V6s: does anyone know if ever a 5 cylinder M/T-series (n.a. and/or turbo) was considered for XX? Technically that must have been quite possible, as LandRover’s TD5 engine from Project Storm was very much related through L- and O-series…

        • Cannot really say whether a hypothetical 5-cylinder M/T-Series was possible for XX, it is worth noting though the 5-cylinder Honda G engine was used in Accord-based models like the Vigor / Inspire / etc whereas the Legend used V6s.

          It seems that while the modular Project Storm engines that eventually became the TD5 were indeed loosely based on the L/O-Series, there is little indication of the planned 4/5/6-cylinder engines spawning petrol variants. However would not have been supersized if such scope existed prior to BMW entering the picture.

          While mention is made of the Project Storm modular 4/5/6-cylinder engines being used in both Rover and Land Rovers, besides potentially replacing the L-Series in 4-cylinder form it is difficult to see how the 5/6-cylinder variants would have fitted into the Rover range.

          • Indeed, especially as XX was a relatively narrow executive car because of Japanese requirements to remain just over 170cm wide, probably hindering the installation of a 5-cylinder engine east-west. IIRC, Volvo had to develop a unique, very compact gearbox when they introduced their 850 and Honda’s Vigor c.s. had their 5-cylinders installed longitudinally.

          • To be fair such an engine family could have worked had the O-Series (aside from appearing earlier) been a fully clean-sheet design that also featured a similar modular capability when the company used RWD cars, with 5/6-cylinder Perkins Prima diesels butterflying away the need for the VM Motori engines.

            Perhaps the 1200cc A40 and B-Series engines plus related O/M/T/L/G/TD5/etc descendants were like the E-Series (albeit fully unleashed) capable of being developed into a more comprehensive engine family then originally anticipated. As demonstrated by the B-Series “Blue Streak” 6-cylinder along with unbuilt pre-BMC 1200cc A40-based 6-cylinder “C-Series” and V8 engines.

          • Considering E-series limitations we discussed earlier elsewhere, this could perhaps have been another realistic alternative route to rationalising BLMC’s (or indeed BMC’s!) midsize engines: to develop B/O-series into a full range family of 1.6-2.0 fours(, 2.0-2.5 fives?) and 2.4-3.0 sixes, petrol and diesel, all with 2 or 4 valves per cylinder, carbs and fuel injection, with and without turbocharging. They could have saved development of E/R/S-series and PE164/144. If only they’d consulted us 🙂

  8. I am a great fan of the Rover 800 turbo diesels. Sadly two of mine were stolen by sociopaths. Allegedly the VM engines are highly sought after and another black marketeer called “Wayne” exports them. I suspect that the engines get reverse-engineered and turned into marine engines but I don’t know who is doing this.

    A problem with the Rover 825SD is that they do not like granny-cycle especially in winter. The problem is the fairly weak 70 amp alternator. On the other hand the petrol powered four cylinder Rover 800 cars, start as soon as one piston goes over compression and these have 100 amp alternators. Presumably the VM engines cost so much money that 70 amp alternators were fitted to reduce the cost. Nowadays there are companies that make 150 amp six-phase alternators that use TWELVE power diodes but getting one to fit the Rover 825SD is likely to be quite a mission. Even the A127i is a problem to fit and most have no “W” terminal.

    For the time being I’m running a frog-moteur but the build quality just isn’t there. The good news is that the HDi models have excellent economy and slightly lower Road Tax.

  9. Thought the 1.4 version was actually an S-Series rather than the M-Series?

    The M/T-Series would have benefited from a 106-124 hp 1.8 as well as an additional 2.0 variant slotting above the 135-138 hp 2.0 DOHC 16v MPi yet below the 177+ hp 2.0 turbo.

  10. The T-Series could potentially have been capable of being made Euro III compliant and there being unrealised plans to replace it with BMW-developed engines.

    The later related stillborn G-Series diesel engine, which itself was intended to replace the related L-Series opens up the question of whether the T-Series itself could have been replaced by a similar petrol equivalent of the G-Series.

    Albeit one where the naturally aspirated variant possibly manages to put out another 10-15+ hp (or competitive to similar 2-litre NA engines of the period), with the turbocharged variants retaining a similar out potential to both the M and T-Series turbos.

  11. The T series was a far more reliable engine than the K series, and it was a shame it was cancelled as it had given years of loyal service in the bigger Rovers. Also the M series diesel used in the 600 endowed this car with 120 mph performance and over 50 mog, important as petrol engined 600s could be thirsty.

  12. massive fan of the site since i found it recently. have been reading all the articles. deeply fascinating. so far though my thoughts:

    > ARG seemed to he hell bent on releasing new models no matter how little budget there was instead of refining models / equipment / seeing if it made business sense.
    > Engine designs seemed to be created in isolation to other needs of the group. Modularity has not been spoken about. Im not a former BL engineer so I cant comment if there was much of it. Across the whole group there seems to be a bewildering number of engine options.
    > ARG seems to be reactive all the time. Lean burn was a fantastic technology, but as some of the biggest markets had already moved on to the cat solution, it was not really a shocker when that became the standard. Also with the diesel coming late to the market. Only after seeing the benefits in other makes cars did they decide that they needed to play catchup. Australia used to have 90% of the 4×4 being Land Rover. No TDi for years.
    > ARG had the technical expertise, the talent & plenty of time to be making world class products.
    > Lastly slightly off topic, friends of mine who studied marketing said that the ARG advertising campaign in the 80s / 90s were so bad they were a used as case studies because every single aspect of their campaigns were a “what not to do”. This was a major point to why the branding suffered.

  13. I remember going around the proving labs at Lucas King street in 1986 when M16 was developed and they were calibrating the ECU on the rolling road. The engines were fitted into MG Maestros as test mules, now there would have been a good hot hatch!

  14. Phil as you may, or may now know, some M16s were fitted into Maestros by enthusiasts (I have a friend who owns one such car and I can confirm that it is not slowl!)

  15. Did Rover ever look at improving the naturally aspirated M/T-Series with some form of early VVC, a system like Honda’s VTEC or perhaps even Nissan’s NEO VVL (also known as Nissan Ecology Orientated Variable Valve Lift and Timing)?

Leave a Reply

Your email address will not be published.


This site uses Akismet to reduce spam. Learn how your comment data is processed.