The engines of the MG T-Series

The MPJG: the TA's engine, 1936 to 1939

The TA arrived in 1936 with a 1,292 cc four-cylinder overhead-valve engine, cast iron block and head, bore and stroke 63.5 mm by 102 mm, breathing through twin SU carburettors. Three thousand and three were built, and the engine was unique to the model. No other MG used it.

Two features define it. The inlet and exhaust ports are siamesed, two valves sharing each port, which limits how well it breathes. And the clutch is a cork-lined unit running in engine oil, a Morris arrangement that feels strange to anyone who has only driven later cars.

The crankshaft is not counterbalanced. Combined with the long 102 mm stroke and a heavy flywheel, the MPJG was never going to rev, and the sources are unanimous that it did not. What it had instead was torque, and a great deal of it low down.

Where the record breaks down: the MPJG

Its power output is either 45 bhp or 50 bhp. Wikipedia and the MG Owners' Club give 50 bhp at 4,500 rpm. Neil Cairns, in the engine history the T Register itself recommends, gives 45 bhp at 4,800 rpm. Both the figure and the rpm at which it was measured differ.

Its compression ratio is either 6.5:1 or 7.5:1. Cairns and the MG T Society say 6.5. The MG Owners' Club says 7.5.

And there are four separate accounts of where the engine came from. This is the largest disagreement on the page.

The club sources broadly agree that the base was a Morris side-valve engine converted to overhead valve, with the Wolseley being an application of it rather than its origin. Wikipedia alone reverses that. Nobody names a common ancestor everyone recognises.

No factory torque figure appears to exist. The MG T Society publishes a curve, but states openly that it is back-calculated from published power figures rather than measured.

The MPJG's weak point today is its crankshaft. The MG T Society's own assessment is blunt: many surviving cranks are cracked and already ground to the maximum. The MG Car Club's buying guide, describing the TB that replaced the TA, puts it plainly. The new engine “was capable of far more tuning than the old TA engine and was much more reliable.”

For those who track such things, the block casting numbers run 22001 from 1936, 22231 from May 1937 when the oil filter moved, and 22290 from early 1938, the first time the MG logo appeared on any block. The head, 22002, is marked “10G” for 10 HP MG, and did not change across the engine's life.

What changed with the XPAG

The XPAG began as the 1,140 cc XPJM in the 1938 Morris Ten Series M, designed by Claud Bailey. MG enlarged it to 1,250 cc by moving the four cylinder walls outward to a 66.5 mm bore, and strengthened the internals. The stroke came down to 90 mm. The full story of the engine, its tuning history and its use across the range is on the XPAG engine page; what follows here is what changed from the MPJG, and where the figures are disputed.

Against the TA's engine the XPAG brought a shorter 90 mm stroke, a counterbalanced crankshaft, reduced reciprocating mass, larger valves and improved porting, full-flow oil filtration, steel-backed white metal bearings, double valve springs, and a modern dry clutch in place of cork running in oil. It is a different engine in every respect that matters, and it is why the TB was so much better a car than the TA.

Where the record breaks down: the XPAG

The bore is 66.5 mm in the period tuning article, in Cairns, in the Y Register, in the MG Owners' Club and in the T Register. Wikipedia says 66.6 mm. On the weight of evidence, 66.5 mm.

Compression is roughly 7.25:1 at the start and rises through the run, but the figures do not line up. The period tuning article and the MG Owners' Club's TD page say 7.25:1. Wikipedia says 7.4:1 for the TC. The MG Owners' Club's TC page says 7.5:1, contradicting its own TD page.

For the TD Mark II there are three incompatible compression figures: 8.1:1 from Wikipedia, 8.6:1 from mg-cars.org.uk's production records, and 9.2:1 from Cairns. These cannot all describe the same engine.

The TD Mark II spread of three brake horsepower, and the TF 1250's of one, are within the sort of variation you would expect between a factory claim and an independent measurement. The MG Owners' Club TB page contradicting itself between its table and its prose is a different kind of problem.

Torque is 63 lb ft at 2,600 rpm, 64 lb ft at 2,700 rpm, or 64.5 lb ft at 2,600 rpm, depending which page you read.

On weak points, W.K.F. Wood's XPAG data service records valve guide wear “significantly higher than in most engines”; tappet wear as “a severe problem”, appearing as early as 5,000 miles; big ends with “a marked tendency to elongate”; and crankshafts which, original or reground, “seem to have a short life.” Cairns adds that the long stroke limits revs, the siamesed ports restrict breathing, and the little end retained by a pinch bolt “is not good for continuous high rpm.”

The prefix system: the one thing everybody agrees on

If you want to know what engine is in a car, the prefix tells you, and the T Register's scheme is unambiguous. It encodes clutch size and water-hole shape as much as model.

The number is on an octagonal, circular or square plate with a brass centre, riveted to the left-hand side of the flywheel housing on the TA, TB and TC. From late TCs onward, and on the Y, YT, TD, YB and TF, it moves to a plinth on the forward right-hand side of the block, with the number also stamped into the plinth itself.

Barrie Jones, a long-standing authority on the XPAG and XPEG, makes a practical point about that second stamping. On a genuine original engine the number appears twice: once on the octagonal plate, and again stamped directly into the block just above it. The two should match. It is a quick check that the plate belongs to the block it is fixed to, and it takes a moment to make.

A square plate carrying a letter prefix (A, B, C, E, G or RS) and a longer number is a replacement engine supplied by Morris Motors or BMC, not an original.

With one exception. What XPAG/TD3 actually means is disputed, and the three accounts cannot be reconciled. The T Register says it denotes a TD Mark II with round water holes. The mg-cars.org.uk TD production records say it means Mark II engines after engine 17xxx, from June 1952. Cairns describes TD3 as the 8-inch clutch variant of 1952 to 1953. Cairns and the T Register cannot both be right, because the T Register assigns the 8-inch clutch to TD2. This is the sort of question a factory document or a registrar's records would settle in a sentence, and no published source resolves it.

The XPEG: a bore, and nothing else

The XPEG is 1,466 cc, not 1,500 despite the model name, with a bore and stroke of 72 mm by 90 mm, 8.3:1 compression, and 63 bhp at 5,000 rpm. Three thousand four hundred TF 1500s were built.

The change from XPAG to XPEG was bore alone. The stroke stayed at 90 mm, and every source that gives both dimensions agrees on that. The cylinder walls were moved outward again, which left cylinders one and two, and three and four, siamesed, sharing a wall with no water between them. The round water holes on the non-pushrod side had to move outward by about 2 mm to clear the enlarged bores.

That last detail has a practical consequence worth knowing before anyone starts swapping parts: XPAG and XPEG heads and blocks will physically fit each other, but the moved water holes do not fully line up.

The head itself is the TD2 casting with larger valves: 36 mm inlet against the TD's 33 mm, 34 mm exhaust against 31 mm. The camshaft is identical to the contemporary XPAG. Block and head castings are AEF117 and AEF118, both carrying the diamond “W” of Wellingborough.

Two dates are worth having, because they overlap. The first TF with an XPEG was chassis 6501 in July 1954. The last TF with the 1,250 cc XPAG was chassis 6950, in September 1954. For two months the factory was building both.

Where the record breaks down: the XPEG

Less than elsewhere, but not nothing. Capacity is 1,466 cc in Wikipedia, Cairns, the MG Owners' Club and the T Register. The Y Register's engine-code table alone says 1,488 cc. That looks like an error, and it is recorded here rather than quietly dropped.

Bore is 72 mm almost everywhere; Wood gives 72.23 mm, which may be a measured rather than nominal figure. Compression is 8.3:1 in most sources, while Wood distinguishes a factory figure of 8:1 from an as-measured 8.33:1. The prefix format is given as plain “XPEG”, numbering from XPEG/501, by the T Register; the mg-cars.org.uk TF production records say the engines were marked “XPEG/TF”.

The carburettors are described by the MG Owners' Club as twin semi-downdraught SU HS4s. No other source confirms the HS4 designation, and it sits oddly with the H-type carburettors quoted for the rest of the family. Treat it as unconfirmed.

No source found identifies a fault specific to the XPEG. The siamesed bores are stated as a structural fact, and no credible source attributes cracking or overheating to them. This archive is not going to invent one. Cairns offers context instead, and it is a fair verdict: 63 bhp from 1,466 cc in 1954 was a good specific output, against the BMC B-series managing 60 bhp from 1,489 cc in the 1953 Magnette.

What nobody has written down

Ten things could not be established from any credible published source. They are listed because an archive should be as clear about its gaps as its contents.

1. Whether any TA left the factory with an engine other than the MPJG
2. A factory torque figure for the MPJG
3. The MPJG engine number format, meaning digit count, starting number and any suffix
4. Where the MPJG number is stamped on the block, as opposed to the riveted plate
5. Block and head material for the XPEG, stated explicitly rather than assumed from the XPAG
6. A TB-specific compression ratio, as distinct from one covering TB, TC and TD together
7. Whether the TF 1250's compression differed from the TD Mark II's
8. A club or workshop torque figure for the XPEG
9. Any XPEG-specific reliability or design weakness
10. Exact engine-number changeover points between variants

How this page was made

Every figure above is attributed. Where sources disagree, all the figures are given and none is presented as settled. Where this archive has taken a view, 66.5 mm rather than 66.6, or 1,466 cc rather than 1,488, it says so and says why.

The obvious limitation is that all of this comes from published secondary sources. The T Register holds the factory production records, and its printed Technical History of T Type Engines is not available online. Several of the disagreements above would very likely be settled in a paragraph by someone with those documents in front of them. If you can settle any of them, this archive would be glad to hear from you, and the page will record the correction and the source.

Sources: the MG Car Club T Register, including Roger Wilson's casting-number reference; the MG Car Club Y Register; The MG T Society; Neil Cairns, ENGINES for M.G's Their Story after 1935, hosted on mg-cars.org.uk and recommended by the T Register; the W.K.F. Wood XPAG Engine Data Service; a period tuning article from Motoring, July 1957; the mg-cars.org.uk TD and TF production records; the MG Owners' Club; and Wikipedia.

Read next

The XPAG engine page for the engine's full story and tuning history; chassis and engine numbers for where every number lives on the car; the MG TF 1500 and identifying a genuine TF 1500 for the XPEG cars; MG TA, MG TB, MG TC, MG TD and MG TF for the cars themselves; and the Technical Library for the workshop material.