Rebuilding an engine will impress upon anyone the need for good lubrication with clean oil. Rebuilding an engine that was damaged by running without oil pressure, like this one, will make that impression even more forcefully. Apparently the car was driven with a bad oil pump, and the owner didn't realize it for some time. It's now my job to undo the damage.
I planned to do the engine work as soon as possible, but I spent a couple weeks working on small bits, like the horns and manifolds, while I waited for back-ordered Whitworth/BSF-sized tools to arrive.
Lacking experience with TD engines, I disassembled this one carefully, took a lot of pictures, and took lots of notes. I have a copy of the factory shop manual, which I followed meticulously. I bagged and labeled all the small parts, so the engine could be reassembled easily and correctly. It's a slow process, but I was confident it would be worth the effort when I put it all back together.
Click on any picture to see a larger version in a new window.
I generally remove as much as possible from an engine before pulling it. Things like the carbs and exhaust manifold have to come off anyway, and the less weight, the better.
I removed the plywood floor panels for easier access to the transmission. It also gave me a good first look at the entire frame cross section.
Once everything was disconnected, the engine was easy to remove.
As soon as it was out, I pulled the transmission and checked the clutch. It looked almost new--it had seemed nice and smooth when I checked out the car before purchase, so no great surprise.
The head looked pretty good. I had checked the compressions before starting the car's disassembly, and they were very good--all within a few PSI of each other. That gave me confidence about the valves, but, as it turned out, the guides still needed replacement. An early look at the underside did not inspire confidence. While the main bearing journals were not too bad, I could feel about 10 mils of looseness in two of the connecting rod bearings. There's the reason for the bearing knock!
The bearings and crankshaft were pretty ugly, but about what you'd expect from an engine that was run without oil. When I drained the oil, there was a lot of metal dust in it; here we can see where it came from. The bearings were badly scored, but the crank's main journals were not too bad. The worst, shown here, had a moderately deep scratch. The rod journals, however, were pretty bad; I measured the worst as 0.006 undersize in the center and 0.004 at the edges. Obviously, it had to be reground. .
The cylinders looked good. They had been resleeved, and, like the pistons, looked almost new; they still had honing marks. I ran my bore gauge through them and found less than a half mil of taper and ovality. I don't think this engine had many miles on it before it was driven without oil pressure; a pity.
I took the crank to a grinder in LA. He thought he could save it, but it would have to be cut to the maximum regrind depth. I hope so, because new ones, or even good used ones, are expensive.
The camshaft was a different story. The lobes were pitted and appeared to have been reground to their limit. Also, the oil-pump drive gear was badly chewed up; that's probably why the oil pump failed. If I left this camshaft in place, the new oil pump's drive gear probably would jam and break again at some point. The camshaft, unfortunately, was trash.
I found a reground camshaft for sale on eBay; it looked pretty good. Here's a close-up of the oil pump drive gear and a couple of lobes, for comparison:
I cleaned the block using naphtha and an engine-cleaning wand. This took off the old oil, but the coolant passages were still pretty bad. I cleaned these by removing the block plugs and alternately rodding it with wire and flushing with lots of water. I wiped the insides of the cylinders with oil so they wouldn't corrode.
The standard way to clean a block is "hot tank" it in a hot, strong, lye solution, which takes off the paint, oil, and rust. You can't do this with a TD block unless you are happy to replace and line-bore the camshaft bearings, which are damaged by the alkali. In this block, the bearings were in perfect shape, so I didn't want to replace them.
One of the oil-pump bolts was broken off, so I drilled it out and chased the threads with a tap. The oil pump's mounting bolts are long and thin, so they are easy to break if torqued too much. Someone obviously torqued this one too much. Fortunately, the hole didn't require any further thread repair. I removed the loose paint on the block with a wire brush.
While I waited for the crank to be reground, I found this beautiful set of NSK micrometers on eBay for a very good price. It should be a big help in getting the engine back together. It's always better to use real measurements than to rely solely on Plastigage. I also bought a few taps and dies for the most common BSF threads used in the car, 1/4, 5/16, and 3/8 inches. These probably won't be needed for the engine internals (its threads are largely metric), but they will be fine for chasing threads elsewhere in the car. Of course, I also have a set of BSF open-end wrenches and six-point, 3/8-inch-drive sockets. I'm optimistic that I won't need more BSF tooling than this.
The engine fasteners are mostly 8x1 mm. This is a standard but uncommon thread size; most 8 mm screws have a 1.25 mm pitch. Fortunately, I have a tap and die in that size; many of the threaded holes in the block and head had to be chased, mostly to remove rust and miscellaneous engine gunk.
While waiting for engine parts, I cleaned and painted the oil pan, installed new freeze plugs, and masked and painted the block.
A week later, my crank came back. The mains were ground 0.020 inches undersized and the rods 0.050. This put the rods at the maximum allowable regrind depth, per the shop manual, so they cannot be ground again. I'd prefer not to go all the way to the limit, but I'm just glad that the crank could be salvaged. I checked the diameters, and they were all within -0.5 to -0.9 mils of the desired diameter. That seemed good to me.
Unfortunately, when I test-fit the connecting rods, I discovered that the big ends were elongated, so they needed to be resized. Another delay in reassembling the engine!
I had the "big ends" of the connecting rods resized at a local shop. They came back looking nice. I checked the size with a bore gauge and they seem OK.
Bearings were installed and the rods were test fit to the crank. All this machining had to be so precise as to leave only 0.001 inch clearance between the bearing and crank journal. I used Plastigage to check this. I also test-fit the crank in the block.
The first task was to install my new camshaft. It went in with minimal trouble. The bearings are tricky; they can go in about six different ways, but only one is correct. This is one of those places where a shop manual is essential. Next, the oil pump was installed; it has to be checked to make sure that it moves freely with no binding. After that, the pistons were remounted on the connecting rods; I used new wrist-pin hardware, as the old looked pretty chewed up. The pistons were then installed in the block; they have to go in from the bottom, because the rod ends are too big to go through the cylinders. A piston-ring compressor and lots of care are essential. In the third picture, you can see that I smeared molybdenum disulfide assembly lube on the camshaft lobes and gears.
The crank was installed, the castle nuts properly torqued, and the crank checked for smooth operation. I then connected the rods to the crank using new bolts and nuts. Finally, I torqued them and again checked for smooth operation. I used assembly lube in all the bearings, of course, and put nonpermanent Loctite on all the connecting-rod hardware. Although they are not shown in the picture below, I installed all the split pins as well.
I installed the timing-chain cover and other stuff on the front of the engine, then the sump gaskets and sump. The rear seal is a little tricky, and the pulley seal, a piece of tar-impregnated rope, is antiquated and easy to damage. The hardware is all 8x1 mm, an odd size in any case. The original hardware, I'm told, had metric threads and BSF heads. Yikes!
Initially, I decided not to touch the head. The compressions were very good, and it seemed best to leave it alone. Therefore, I just primed and painted it.
After some consideration, however, I realized that I should check the valve guides. I took the head apart and found that the valves seemed noticeably loose, and that the guides were inserted too far into the head. (Among other problems, this caused the guides to block the intake and exhaust ports significantly, so replacing them probably gave me a few more horsepower.) The valves looked good and the shaft diameters were well within spec, so I decided to replace only the valve guides.
Removing valve guides was not easy! The shop manual tells us to fabricate a drift with which the guides are hammered out through the combustion chamber side of the head. The head had the old-style, steel guides, which were far too tight for that to work. I fabricated the appropriate drift and pressed the guides out with a hydraulic press. Even with a 12-ton press, it wasn't easy; it required a scary amount of force, and each guide let loose with a disturbingly load BANG.
I used the press to install the new valve guides. They're still tight, as they should be, but they went in easily. I used manganese bronze guides, which are superior to steel ones. It's important to recognize that the intake and exhaust guides are slightly different, and I had to be careful to get them into the right places. Also, the guides have a step in the outer width. Many people think that this step should be flush with the upper surface of the head, but that's not right. It should be about a quarter inch above the head surface. (The shop manual has the precise numbers.)
The guides were a little undersize. I don't know if they were made that way intentionally, with the expectation that they would be reamed, or were just poorly made. In any case, I reamed them to 0.317 inches (8.05 mm), the minimum specified diameter, and checked the valve fit. The guides are hard suckers; reaming them was not easy. In fact, nothing about this job was easy. But it's done.
To make sure that the valves would seal properly, I gave them a light grinding. This is important, as the holes in the new guides could end up a fraction of a mil away from the original ones, preventing the valves from seating perfectly. A light grind compensates for any change in location.
I then reassembled the head. I used a "c-clamp" type of spring compressor, which made installing the seals and keepers very easy.
After installing the head, I checked the compressions, and they were still excellent.
Below is a video of the exhaust with the engine warm and idling. No smoke; just what you want to see.
I didn't do any work on the transmission, as I have documentation of a recent rebuild ("recent" in terms of mileage, anyway; not time). I checked it when I bought the car, and it seemed to shift perfectly, so there was no need to work on its internals. It just needed a clean-up, fresh paint, and, of course, fresh oil.
I used my shop crane to handle the transmission, as it's just a little too heavy for me to lift. The oil didn't have any metal particles, but it was definitely due to be changed. To paint the transmission, I first cleaned it with solvent, then masked, primed, and spray painted its body with engine paint. I painted the bell housing and shifter with a brush. Fortunately, the transmission was relatively clean, so it didn't need any wire brushing or scraping.
The TD transmission contains a certain amount of brass, which can corrode if conventional gear oils are used. GL4-rated oil is required. Most oils today meet the GL5 standard, which is not acceptable. Oils marked "GL4/GL5" also will not do. The only readily available, true GL4 oil I could find was Redline GL4, an oil that is widely used in early MGs.
I mated the transmission to the engine. I wish someone had told me that the flywheel had to be installed before the oil pan. I removed the pan, installed the flywheel (with new bolts, of course, and the correct torque), tapped in a new pilot bushing, reinstalled the oil pan, bolted on the clutch (with my homemade alignment tool), and wiggled the transmission into place. I didn't add oil at this point; that waited until the transmission was installed in the car. I also didn't install the head at this point, so I covered the cylinders with towels to prevent debris from falling into them.
The pivot arm in the clutch linkage needed attention. The pivot bushing was quite worn and the linkage pivot was an ugly, home-made part. I made new parts on my lathe and milling machine and pressed the bushing into the pivot arm. Finally, I repainted the pivot arm.
The universal joints didn't seem bad, at first, just well used. I could have continued to use them, but I decided to replace them anyway. Just as well: once they were disassembled, I realized that they were worse than they had appeared. One had no grease nipple, and its little remaining grease had deteriorated badly. The old U-joints were difficult to remove because the bearing cups were very tight. The last person who serviced them should have honed out the mounting holes a mil or so. I did that, and the cups were reasonably easy to press into place. Some previous owner lost the grease fitting for the splines; I found a bubble-pack of assorted fittings at the local car-parts store and replaced it.
Like everything else, the driveshaft was cleaned and repainted. Removing the old U-joints had an additional advantage: it was easier to clean and paint the yoke, driveshaft, and flanges. I was careful not to paint over the yellow marks on the shaft's flanges that show the alignment with those of the transmission and differential--although it's unlikely that the driveshaft's many previous disassemblers have always been careful to retain the original orientation.
I was not looking forward to the dirty task of cleaning the rear axle; it was covered with dirt and greasy crud. I hung it from my shop crane and scraped it, then followed with solvent and detergent washes. That left it clean enough to work on, but a lot of hard, dry crud remained, mostly on the differential housing. I cleaned that off with the rotary wire brush and a pin scaler. The housing then was primed and painted.
The rear axle housing came apart easily. I marked the axles so they could be reinstalled on the same sides where they previously had been. This is important, as TD axles are notoriously weak, and minimizing fatigue in this way helps to prevent them from breaking.
The differential had too much backlash at the pinion, and I discovered why: one of the mounting bearings for the differential was badly worn. You can see the problem in the third picture; the outer race is so loose that it hangs off center. I decided to replace both bearings, as both have 1950 date codes and they don't owe me anything. The oil clearly had been in the differential too long; it had started to break down, as indicated by its sickly olive color in the pictures below. To remove all the old oil, I cleaned the gears, pinion, and the interior of the entire axle with naphtha and made sure it had time to dry completely. I replaced all the seals and checked the gears and axle splines carefully for wear or damage. The sealing surface of the pinion flange was a little rough, so I put it in my lathe and polished it. I went over the outside of the axle with a wire brush and finally painted it. Of course, it received a full charge of new, top-quality gear oil.
Why did one bearing wear out so badly while the other didn't? The oil level was low, and it clearly had not been changed for far too long. Not a great surprise--in these days of virtually maintenance-free cars, owners often aren't aware that the differential oil has to be checked regularly, let alone replaced. Even though the oil level was low, the bearing in the larger half of the housing was probably lubricated by oil thrown around by the gears. The one that wore out, however, was largely shielded from this "splash lubrication" by the ring gear. Effective lubrication of this bearing seems to require an oil level that is high enough to pool around it, so long operation with a low oil level probably caused it to wear out.
It wouldn't surprise me if this was the first time the axle was opened. The bearings clearly were original and I didn't see any sign of previous work. That's not unusual; differentials are pretty rugged, and it's rare for them to break.
I pulled both old bearings and pressed on new ones. The problem, then, was to make sure that the differential was set up right. This is invariably a tricky business and usually requires special, factory-supplied tooling, all of which has been unavailable for decades. What to do?
There are four adjustments in setting up any differential: the position of the pinion in a front-rear direction, the position of the ring gear in a right-left direction, and the preloads on the bearings of both. I assumed that the pinion was in the right position; since it hadn't been disturbed and there was no looseness, the preload also was probably OK. I based the ring-gear position on the 1.5-3.0 degree pinion backlash specification, which is adjusted by the right-side spacer. The preload is adjusted by the left spacer; although it's not a factory spec, sources say there should be a 10-mil gap between the halves of the rear axle case before they are bolted together. This is typical of other differentials, which usually specify something around 5-10 mils.
The existing right-side spacer gave approximately two degrees of backlash, which seemed ideal. I measured this by marking the pinion flange and rigging up a pointer. (The radius of the flange is 1.7 inches, so 1.5-3.0 degrees is 45-90 mils at the edge of the flange.) Unfortunately, there was no bearing preload--the two halves met when I put them together. I made a 20-mil spacer, added it to the left side, and tried again. To measure the gap, I installed a few of the nuts and adjusted them until the gap was uniform all around the joint. The gap then was 23 mils (there must have been an imperceptible 3-mil gap when I originally mated the two halves). Since the gasket is 15 mils thick and can be expected to compress a bit, that gap seemed fine. I installed the gasket, applied sealer to it, and torqued the nuts to 30 pound feet.
I left the radiator restoration until I had finished the engine. I hadn't taken a good look at it and just assumed it was OK. It wasn't bad, but it did need some paint and general restoration. Interestingly, the tee in the cooling system is home-made. A little sloppy, but it works. The location of the bypass tube is not original; the last picture below shows how it was arranged when I received the car. I changed it to something more correct.
The radiator's lower cradle was very rusty, probably because of a felt pad between it and the lower reservoir, which could get wet and then dry only very slowly. A stupid design, but, of course, the car wasn't expected to last 60 years. I could have patched the rust holes, but there still was a lot of corrosion elsewhere in the part. (In the pictures below, much of the rust has been wire-brushed off!) I decided to make a new cradle out of two-inch steel flat bar. It won't look like the original, but it's not visible, so that doesn't matter. The pieces, before welding, are shown below. I made new bushes on the lathe and cut studs from threaded stock to replace the threaded mounts, which also had deteriorated badly.
I aligned the cradle, welded it together, and checked the fit with the radiator core. The fit was perfect. I wire-brushed the radiator core and sanded the paint that the wire brush couldn't reach. I then masked, primed, and painted both the radiator and the cradle.
I repainted the radiator grille, as it was a little dingy, and the inside of the chromed radiator front. The inside isn't visible, but I thought that some paint might preserve it a little better. I also restored the drain tap at the bottom of the radiator and reinstalled it with a new copper washer. The large hose from the thermostat to the radiator reservoir could be installed only by removing either the radiator or the thermostat housing; I was glad I noticed this before bolting everything into place. Instead of the dinky, rubber washers for the radiator's lower mounts, I used a nice slab of 1/4" rubber. I made new rubber pads for the underside of the radiator mount, as well. A little chrome polish and it will look spectacular.
Originally these cars had no temperature gauge. They really don't need one, because they don't have a pressurized cooling system--if it gets too hot, it just boils over and you know there's a problem. A dual temperature-oil pressure gauge was an option, which later became standard. I decided I could do without the temperature gauge, since the dual gauges are hard to find, expensive, and almost always need restoration. Since I had no plans to use a temperature gauge, I made a plug for the temperature-sensor hole in the radiator. The plug is made of Delrin, which should handle the temperature, and has a 3/8" BSP thread, 0.65" diameter and 19 threads per inch. You don't find those at Home Depot! I had to use the thread-cutting tool on my lathe to make it.
The lower radiator hose and the bypass tube was a problem. I didn't want to buy the standard part ( about $100), and the kludged setup, shown above, which existed when I bought the car, was not acceptable. I made a tap from a couple pieces of brass pipe. I drilled and tapped the large one, then screwed a smaller one into it. I then soldered the joint to prevent leaks and to give it a little more strength. The next problem was the lower hose. With a little on-line searching, I determined that a Dayco 71834 radiator hose, with a little modification, would work. I found one on eBay, shortened it a bit, and it fit very nicely. The cooling system is now complete, and, while not precisely original, still looks good.
Sharp-eyed readers may have noticed that I got the fan blades oriented incorrectly; they move air correctly, but the cupped side should be toward the engine. That's partly because they were installed incorrectly when I acquired the car, and also because the hole pattern and location of the reinforcing plates seemed to indicate the installation I used. After the restoration was completed, I took them off and turned them over. It's not a big deal, but these cars need all the cooling efficiency they can get.
Oddly, the holes on one blade are offset, so the blades are not at a 90-degree angle. I don't know if this is a flaw or a feature, but it probably doesn't matter much.
The engine-block drain tap leaked around the valve pivot. I had tried to restore it, but the nature of the leak convinced me that it couldn't be fixed. I cut it off and kept only the threaded part (1/8" BSP), drilled and tapped it 8-32, and inserted a socket-cap screw. I can remove the screw if I need to drain the coolant, and it should be easy to seal against leaks.
I painted the muffler and exhaust pipes with Eastwood high-temperature silver exhaust-system paint. I painted the intake and exhaust manifolds similarly, as you can see in other engine pictures. I put it all back together with new hardware, including chromed U-bolts. I think it now looks much better, and the coating should extend its life.
I reinstalled the engine with minimal reassembly. This way it was lighter, so it was easier to manipulate and to remove again, if necessary. (There's optimism for you!) Even without the footwell's front panel, it was necessary to remove the shifter from the transmission.
With the engine in place, there was just barely enough weight in the car to install the rear limit straps. The weight also made it possible to connect the rear shock links and the driveshaft. Now the drivetrain is all hooked up. Of course, I still have to wait to tighten the bolts for the shock links, like many other suspension bolts, until there is full weight on the the suspension.
First step is to crank the engine and get oil pressure. The TD's oil pump is notoriously difficult to prime, but once it's pumping, it should need no further attention. I primed it by squirting some oil into the output port. I hooked up the old electrical cables (I'd ordered new ones, but they hadn't arrived), filled the oil filter with oil (to minimize cranking with no oil in the bearings) and cranked until I got pressure. It came up fairly quickly to about 50 psi. I did this before installing the head or other external pieces, to make things easy if I had to take something apart. I did find a small leak in one of the fittings, so I had to take that off and see what was going on. It was just a bad copper gasket.
The engine continues to go together. I think it looks pretty slick.
Here's one change I made unapologetically. The grounded battery cable normally is bolted to the thin metal wall of the battery box. (In my case, this is the negative battery terminal, as my car has been converted from the original positive-ground configuration to negative-ground.) That wall is much too thin for the starting current. Instead, I used a longer cable and grounded it to the point where the engine's ground strap is bolted to the frame. This should provide a better electrical path to the starter and to other electrical components.
Everything went together, although slowly. Here I have installed the distributor and ignition wiring, as well as some electrical stuff on the scuttle. I did an initial valve adjustment and installed the fan belt. I made new mounting brackets for the horns, as the original ones were cracked in several places. The regulator is an old one, but in good shape. I readjusted it and checked the electrical contacts, which were fine. The older regulators, in my opinion, are much better made than the new ones. No surprise, I guess.
I finally got the carbs assembled along with fuel lines and a filter.
Before trying to start the MG, I went through the valve-adjustment procedure a couple of times, then made sure the distributor was set up right and correctly static-timed. I preset the carburetor mixture in the usual fashion, about two turns of the adjuster from full lean. I ran the fuel pump by itself to make sure it was pumping properly and there were no leaks--I did encounter a leak at the front carburetor, which I fixed. I took off their return springs so the choke and throttle would stay in position once I set them. I powered the ignition and fuel pump with clip leads from the battery.
First couple of cranks, the engine kicked backward for some reason, but on the third, it started and ran well.
Here is a video of the engine running when I started it first time after beginning the restoration. I ran it about 20 minutes at 2000 RPM, to bed in the new camshaft.
The oil pressure was about 45 PSI, within spec but a little on the low side, because the oil pump's pressure relief valve was opening a little early. After the car was completed, I added a shim under the spring that sets the relief pressure and raised it to about 55 PSI.
Before spending the money for insurance and registration (the car was on non-operational status and minimally insured), I wanted to be certain there were no major problems that might prevent it from being driven. I pushed it out of the garage, into the driveway, kept a hose and fire extinguisher ready, and started the car. It started easily and ran well. Here is a video of the engine running:
I found a small oil leak and a small fuel leak, both of which I fixed; no other serious leaks. I drove it back and forth in the driveway a couple times to be sure the clutch and transmission were OK. Later that day, I turned on the insurance and changed the registration from non-operational to normal status.
In the picture below, the engine is running. I ran the car for about 20 minutes to warm it up.
Once the engine was warm, I checked the timing and adjusted the carburetors. When all the adjustments were complete, I installed the air filter. Barring a few more tweaks, it's ready to go!
