NGINE - First Attempt
Taking the engine out was a significant step just on its own account. It was like an irrevocable statement of intent: as though this was the step that finally committed us, even though in reality we'd already gone a long way beyond the stage where the only way out was to call the Council for their skip.
There really isn't much for me to say about the engine itself, since the machining tasks are well beyond me, and everyone already knows everything else there is to know. I was inclined to put it into a corner of the garage and leave it till its time came in the normal sequence of events, but James was into Scouts at that time and wanted a major project for his technology badge. So here's a good idea: why don't we take the engine apart, have it machined as necessary, and reassemble it and get it running ourselves, with every stage carefully documented and recorded, so he can get his badge??
Well, the plan worked brilliantly at first.
I took the whole thing to the machine shop and they dismantled it, did what was necessary (which was almost everything), and gave it back to me in pieces. They advised not to put in valve seats to handle lead-free petrol, on the grounds that there was no demand yet for unleaded petrol, and the technology was unproven. Well, I suppose that gives you some idea how long this project has been running. Also they advised against the rear oil seal modification, this time on the much firmer grounds that the existing solution involved a lip-seal oil seal running on the edge of the flywheel mounting flange, and because of the large diameter and the limit on lip speed, engine revs would be limited to about 2000 rpm. There is a much better solution available now, based on the GM split seal, and no doubt if I ever want to do the modification, there will be a better solution yet.
The crankshaft had already been ground several times, and the journals were now at their limit. I was lucky, though, because the engineer said he might be able to source a replacement. And he did, and the price didn't seem too unreasonable. And when I told this to the President of the Car club, he laughed, and said the same man had lost his crankshaft, so he'd had to use his spare. So my car probably has the President's crankshsaft. And I wonder what happened to mine? And later, I came to regret this transaction, and to wish the President's had never been lost!
Anyway, we got the engine back, and I showed James how to
Oil Pressure Although the car was substantially finished - at least finished enough to drive it in good weather - the oil pressure was disappointingly low. I changed the oil pressure relief valve - no difference. I swapped the oil pressure gauge - no difference. I was puckering myself up for the first serious bit of maintenance - taking the oil pump apart and measuring/reconditioning it, followed by the oil filter bypass valve, and then even taking the sump off and checking the suction filter assembly and associated gasket which are located in the sump itself, which would have meant oil leaks ever after because I hadn't done the lip seal conversion and I don't believe it is possible to assemble the cork rear oil seal correctly without taking the whole engine out - when I thought to trace as far as possible oil pressures throughout the system. Now there is an external pipe supplying the rocker shaft from the lower oil gallery, and the feed to the oil gauge is taken off here. The diagram on Page A2 of the manual (Engine Components) shows the feed being taken off the top of this pipe, but Figure A1 on pA4 suggests that the feed is taken from the bottom of the pipe. All the cars I have seen have the feed coming from the bottom, with a union on the lower banjo, but I thought it made a much neater installation to take the feed from the top, so thats what I did. I inverted the pipe so that the banjo with the union was on the top, and took the feed to the oil pressure gauge from there. So the first step in my diagnostic procedure was easy - simply blank off the feed to the rocker shaft, and run the engine just long enough to check the pressure. Bingo: the gauge registered an acceptable pressure (still a bit lower than I would have wished, but acceptable). Evidently the rockers were taking enough oil flow to lead to a significant pressure drop along the external pipe. Solution: blow through the external pipe, replace it in its correct orientation, and take the oil pressure feed from the bottom. Now an acceptable pressure was registered: still a bit lower than when the rockers were blanked off, indicating a significant pressure drop even along the main gallery, and perhaps also indicating that the oil pump is not capable of supplying demand from the entire engine and therefore requires overhaul but good enough for me to put it off sine die.
econd Attempt. I had been looking for an excuse to take the geabox out, because as well as the expected noise from square cut gears on first and reverse, there was an unacceptable noise from the bearings, which in a fit of parsimony I had not replaced, even though they are so cheap, and also there was just a hint of the jumping-out-of-third problem. So I was pleased and relieved when, after only about 3500 miles, an opportunity presented itself.
The engine didn't have quite as much power as it had had, and it started using oil and blowing smoke - so much smoke, in fact, that on club runs they made me drive at the back so others could see. Compression was low on nos 3 and 4 cylinders. And the oil in the sump looked like the stuff they use to repair roads.
I thought, as well as sorting out the smoke issue, I would put in hardened valve seats, unleaded petrol now being an accepted fact of life, and also have Andy Bradley's GM rear oil seal conversion done, since no better solution had been found ( 1, 2, 3) since I originally wrote these words.
So off to the engineers again, this time choosing a different mechanic. He pretended to be dismayed as he took it to bits, but I saw him rubbing his hands and chuckling with glee when he thought I wasn't looking.
He called me to his workshop several times to discuss what he had found.
The first issue, as expected, dealt with the smoke. He showed me that almost all the piston rings were broken, and as he contemptuously flicked bits of piston rings out of the lands onto his otherwise pristine floor he told me that he would never have used that particular supplier of pistons or rings, thus clearly establishing his superior knowledge and credibility compared to the previous engineer. So we had to get new pistons and rings (as expected, of course), and that led to the second visit: the cylinders were at maximum rebore size, and the block would have to be relined, and as he turned the block over, he exclaimed with surprise that it had already been relined once (this was a surprise to me also, because I didn't remember having it done myself, and given that the car was only 17 years old when it came to me, and hadn't had a lot of maintenance, I found it hard to believe that the previous owner(s) had had three rebores and a reline and possibly another one or two rebores in those 17 years!).
The next visit related to the cylinder head. For a reason I can't understand the holes for the waterways in the gasket are much larger than the waterways themselves, allowing corrosion of the head, and in this case, corrosion of the head sideways underneath the gasket reducing the clearance to the cylinders by perhaps as much as 1mm, which was felt to be too much. We could have machined the head to make it flat again, but it was already at 75.5mm (compared to standard 76.75mm), so another regrind would have taken it close to Stage 2 tuning and this was felt to compromise reliability to a greater extent than desirable (on the other hand it would have given me an excuse to use castor oil in the fuel to give that great smell!). So I went to my parts supplier, who showed me where he had what must be all the spare heads in NSW in the excavated space under his shop, but declined to go there himself. So I crawled in, among the spiders and worse, and chose the best head I could find, which looked as though it had had very little use during its life, and which passed the crack test. And he charged me what he would have charged when he put it there, about fifty years ago. I couldn't find casting numbers, but it had oval holes, so it was one of the later heads.
The next problem was the crankshaft. As a matter of routine, he had it crack tested. And it had three cracks, one major one near the rear web which looked as though it was just about to fall off. So it would probably have been better if the President had taken his crankshaft to a different engineer who would have taken better care of his parts, and I could have bought one without cracks.
Anyway, my mate Paul had a spare shaft, which he was willing to sell me and which I bought. So that dealt with that problem. After grinding, the big ends were 0.040" under, and the main bearings were 0.030" under.
We rebuilt the oil pump with a new idle gear and shaft, and now it runs at 45 psi cold and 35-40 hot.
The camshaft and camshaft followers were redressed, and new camshaft bearings fitted. We had hardened exhaust valve seats fitted, new valve guides and new valves and 120lb springs, and modern valve seals, so we shouldn't have a problem from oil passing down the valve guides, the other notorious source of smoke.
The cylinders were relined and machined to standard bore, with new pistons and rings.
And most importantly, of course, the modified rear oil seal. This was well described by Andy Bradley almost twenty years ago, and I'm not going to try to restate his instructions. Suffice to say that the original oil slinger disk was machined away, and the Archimedes screw, the source of all the oil leaks, was built up then machined down to the specified diameter 2.437 inches diameter to take the new two-part seal. In the picture of the crankshaft you can see, on the left end adjacent to the rear main journal and to the flywheel flange, how this has been done, and you can compare it to the original. Bradley's original solution called for some machining to the rear main bearing cap, and to what he and Moss Motors believe should be called the thrower plate but what MG has always called the oil seal cover; the machining to the rear main bearing cap is still required, but today's modification includes a new oil seal cover. And in the picture you can see half the new seal in its place in the cover. The other half fits inside the modified rear main bearing cap, which is obviously not fitted yet.
And finally, the engine rebuild gasket set comes with a teflon impregnated rope seal for the front oil seal. This is intended to be used for landfill, and as a reminder to source from your local bearing shop a modern oil seal dimensions 36 x 47 x 7 mm which will fit without modification to the timing chain cover and the sump and will close off the last opportunity for oil to escape (and if you are in Australia and get it from Consolidated Bearing Co it will be part no TC12532).
Camshaft and Tappets. My tappets were always noisy, and it seemed to me they were noisier than others (but everyone thinks that about their own car, right?). The manual calls for them to be set to 19 thou hot, but I didn't like setting them hot, because it was uncomfortable and messy, and anyway, how hot was hot? It was always a dilemma for me: if I set them cold, what should the clearance be? Should it be the same? - but then the manual wouldn't have specified to set them hot. Of course you have to set them so that, after expansion of the metal as it heats up you end up with 19 thou. But what expands more, the block or the pushrod? - if the pushrod, then you'd expect the tappets to close up on getting hot, whereas if the block, then you'd expect the tappets to loosen up. And what's the difference in coefficient of expansion for the two types of steel? So I had always set them 20thou cold, thinking that if they loosened up when hot, it wouldn't matter much, whereas if they tightened up significantly there was a risk to the exhaust valve seats. Well, after having set them this way, on this car and others, for so many years, it suddenly seemed important for me to find out now.
So I thought I would search the web to find out how they should be set cold; after all, someone must have done the experiment and written it up. But I could find nothing: no-one had set their tappets hot then remeasured them when the engine had cooled down! Or if they had, they'd kept the result secret. Perhaps I should do it myself?
Anyway, absent information about my particular problem, I did find some really interesting things about camshafts.
The first was a general discussion about XPAG engines by Neil Cairns who is by way of being a source of all knowledge about MGs. Here I found, for the first time (where have I been? what have I been doing?), that more than one camshaft fitted my car, and there was a possibility that my camshaft was not the original.
Dave DuBois discusses the difference between the two cams.
XPAG engines up to TD2/24115 (ie mine) were fitted with so-called AAA5776 cam, with valve timing 11;57;52;24 (ie inlet valve begins to open 11 degrees before topdead centre and closes 57 after bottom dead centre; exhaust valve begins to open 52 deg before bottom dead centre and closes 24 deg after top dead centre: the inlet valve begins to open before the piston begins its induction stroke and stays open well into the compression stroke, while the exhaust valve begins to open well before the end of the firing stroke and stays open until well into the induction stroke). From this you can see that the duration of the inlet valve is 248 degrees, and the exhaust valve is 256 degrees; they overlap (ie both valves open, at the end of the exhaust stroke and beginning of the inlet) by 35 degrees. With the specified tappet setting of 19 thou, the valve lift is 0.315 inches.
Later engines were fitted with AAA3096, with valve timing 5;45;45;5, duration 230 degrees, overlap 10 deg, and slightly larger valve lift of 0.327 inches, and tappet setting 12 thou. Note the symmetry between inlet opening and closing, and exhaust opening and closing; note also the much smaller overlap, when both valves are open, and the fact that the exhaust valve doesn't begin to open until much later in the firing stroke.
As well as this information, DuBois goes on to describe a method to work out which you have.
With the engine cold, you take off the rocker box cover, and turn the engine by hand until #1 inlet valve is fully open (easier with the plugs removed). At this point, #4 inlet valve will be fully closed (this would be because, if #1 inlet valve is open, then #1 is on the induction stroke, when of course #4 will be on compression). Set the tappet clearance for #4 inlet to 21 thou, whichever camshaft you think you may have. For the next step, he suggests making chalk marks on the crankshaft pulley. But it's hard to make chalk marks with enough precision for this exercise: I made my marks on a piece of masking tape, which I then lined up with the timing mark. The first mark is to align with the TDC timing mark on the pulley; a second mark between 5/32 and 11/64 to its right, and a third mark 23/64 to the right of the TDC mark. It's not hard to work out that for standard pulley, 3.75 inches diameter, these marks correspond to 5 degrees and 11 degrees BTDC respectively. Now turn the engine over until #4 inlet valve is just ready to begin opening, that is when you just can't spin the pushrod between your fingers. At this stage, one or other of the marks should align with the TDC mark on the timing cover. If it's the further (the 11 degree mark) you have the older camshaft, and if the 5 degree mark, the newer.
Well, I did this exercise, although I was comfortable with the fact that my camshaft had in all probability never been changed, and I was astonished to find that according to this test I had the newer camshaft. I repeated the entire sequence several times, and always came up with the same answer.
But there is only about 3/16 inch between the two marks, which is not much when the consequence of getting it wrong is so critical.
So I'm grateful to Neil Cairns for reprinting David Clark's much simpler method, which relies on the fact that the newer camshaft is symmetrical (that is, not only the obvious symmetry between inlet vlave opening and exhaust closing, and inlet closing and exhaust opening, but also the camshaft lobes themselves are symmetrical), and if yours isn't symmetrical, you must have the older shaft.
With this method, you first turn the engine by hand until #1 is at TDC at the end of the exhaust stroke and (of course) at the beginning of the inlet stroke, when both valves will be partially open (the exhaust valve getting ready to close in 5 degrees, and the inlet valve having begun to open 5 degrees ago). Undo the tappet locknuts, and unscrew the tappets right up, then down until they just touch the pushrods (that is until you just can't spin them between your fingers).
Then you turn the engine by one complete revolution, until #1 is TDC at the end of its compression stroke and beginning its firing stroke, and measure the tappet clearances on inlet and exhaust valves. If they are equal, or within 5 to 10 thou of each other, you have a symmetrical crankshaft, probably the later AAA3096 shaft; if not, you have the older, asymmetric AAA5776 cam.
Well, my tappets came in at 44 thou and 48 thou. So I am comfortable that I have a symmetrical camshaft.
But Cairns points out that more than one camshaft had the 5;45;45;5 profile, but with different valve lift and different shape, for Morris's 1140cc engine. These should have 19 thou tappets, and if you have one of these and set the tappets to 12 thou, then performance will be "awful". And at least one custom aftermarket camshaft had a symmetrical profile. But what are the odds of having one of these? So I'm comfortable, from the results of both tests taken together, that I have the AAA3096, and I should set my tappets to 12 thou hot. DuBois points out that an engine with the later AAA5776 camshaft but with tappets set to 19thou would be slightly down on power but produce a lot of valve clatter, whereas if the tappets are too low you can burn out the valves and valve seats. So being conservative, and because setting them hot is still uncomfortable and messy, I set them to 14thou cold.
What about the power? I don't detect much difference at low revs; but above 2000-2500, watch out: a noticable increase in power. Also a noticable change in exhaust note - much throatier. So now I must set them down to proper 12 thou hot, and see what happens. And then of course I can let the engine cool and I can report on how the settings change when cold. (and having done that experiment, now I can report that cold, the tappets are anywhere between 12 and 14 thou cold).
I believe the car should be good for at least another 30,000 to 50,000 miles (and since I have been doing less than 1000 miles a year since it was rebuilt, that means it should last for more than thirty more years). I anticipate that any problems after that will belong to the next owner!
Metric threads Well as everyone except me already knew, it turns out that all the threads used on the engine are metric threads - mostly 8 x 1 metric fine, but also some 10 x 1.5 - but you use Whitworth or BSF spanners, as befits a classic English sports car. It took me a disappointingly long time to realise this, and probably quite a few stripped threads as I tried to match BSF and metric threads.
I wanted to write a clever pun about threads - it would have involved the interweaving threads, the warp and the woof that make up the rich tapestry of life to make this sad story of non-standardisation - but I wasn't clever enough ....
Anyway, this story begins in Connecticut in 1826 with the birth of Benjamin Hotchkiss. Like many Americans, even to this day, Hotchkiss was obsessed with weapons and armaments. Failing to interest the US government, he moved to France and established the Hotchkiss company there in 1867. After his death in 1885, his company continued to develop armaments, especially machine guns. The first model, displayed in 1892, was adopted by the French army in 1897, and was subsequently developed into the definitive Hotchkiss gun, a truly automatic, gas-operated heavy machine gun. By 1914, it had been adopted as the standard machine gun by British, French and Japanese armies. Whether they anticipated the hostilities which were to come, or whether it was a condition of sale to the British army, Hotchkiss established its armaments division in Coventry, naturally using the same machine tools as its parent, and replicating the metric threads. After the armistice, there turned out to be little demand for heavy gas-operated machine guns (could there be an opportunity here for an MBA thesis about losing your market by being too successful?), and Hotchkiss turned its knowledge and equipment to manufacture of engines.
At much the same time, in 1892 young William Richard Morris, then aged 15, began a bicycle repair business from his father's back shed, and later expanded it to include motorcycles. By 1910 he had expanded again to include motor car hire and repair under the name of WRM Motors Ltd, and soon began selling new cars as well. In 1912 he designed his first car, the two-seater Bullnose Morris, which used a 1018cc engine from White and Poppe. By 1914, however, a coupe and van needed bigger engines, and a 1548 engine was sourced from Continental Motor Manufacturing in Detroit. The same engine was used in the 4seat Morris Cowley and later Morris Oxford. Demand continued to grow throughout the war years, but after the war, possibly due to prohibitive import tariffs, Continental stopped supplying their motors. Morris bought the design rights, and persuaded Hotchkiss to build it in their Coventry works. At about the same time, he closed WRM Motors and allowed it to be absorbed into a new company, Morris Motors Ltd, and in 1923, pursuant to his policy of vertical integration, he bought the Hotchkiss works and incorporated it into Morris Motors as Morris Engines. (While there appears to be pretty broad agreement about these general facts, not all accounts agree with the chronology.) Hotchkiss continued to use their metric threads; but because only BSF/Whitworth hexagonal bar stock was readily available, they cut their threads into Whitworth stock, so that the metric nuts and bolts fit imperial Whitworth and BSF spanners. And a bit more than eighty years later, it's still giving me grief.
please send me an email
Top of Page