The Gearbox

The first problem I had assembling the gearbox was that the primary shaft (Table 8,#54) was not the correct one! I sent this one off to Jerry Kimberlin who very generously machined it down to the correct specs including welding on the tang at the end. Here are the correct dimensions in case anyone needs to recreate a shaft. The trueness of the shaft needs to be checked on V blocks and cannot have a radial runout greater than 0.05mm.

The second problem was the bush inside the shaft (Table 8, #59) was not the right size and had to be remachined which again, Jerry did. Here are the dimensions if anyone wants to make one out of 660 bronze. I have also shown the dimensions of the ring that sits in between the direct gear and the bearing. This is a critical dimension and needs to be set for each individual gearbox (more on this later). The spacer ring (Table 8,#55) is assembled with the flat side against the bearing as shown in the photo below.

The next step is inserting the direct gear and the spacer ring into the RH cranckase output bearing after cleaning the small oil hole with a blast of compressed air. The spacer ring is assembled with the flat portion of the ring facing the bearing. The gear is a tight sliding fit and should not need a lot of force to either insert or remove.

Assembling the gearbox:

All the assembly is done on the LH side of the crankcase. I assembled it without the crankshaft as you have to open and close the cases several times to check if the spacer ring behind the direct gear is the right fit. I first inserted the clutch hub (Table 7 #10) from the outside into the LH side upper bearing. As specified by the manual, I made sure to support the inner race with a short tube while tapping the hub into the bearing. I then inserted into the hub, the tapered end of the mainshaft with the flat side of the spacer ring facing the bearing. After tightening the nut on the outside of the clutch hub, I rechecked the alignment of the mainshaft (max run-out is 0.05mm) and made sure there was no lateral movement between the shaft, the hub and the bearing. I next assembled the gear selector quadrant making sure that the "4" punched in the casting matched the "4" marked on the tooth arc. Next to go in was the secondary shaft with the large gear closest to the LH casting. There are two spacer rings that go on either end of the shaft-the thinnest (15.2 mm ID x 22mm OD x 1.2mm thick) one goes on the side closest to the direct gear (RH) while the thicker one (15.2 mm ID x 22mm OD x 2.3mm thick) goes on the clutch side (LH).

Now, to assemble the shift drum. The gearbox is assembled in 4th gear. The shifter was properly lock wired and ready for assembly. I rotated the 4th gear shifter (the one opposite end to the gear side of the drum - RH side in the photos) close to the top of the slot in the drum. I then rotated the drum till the lower shifter (LH in the photos used for 1st and 2nd gear) aligned with the first and slid the twin first and second gears into the LH lower clevis fork and the single gear into the upper RH clevis fork. It is easier to rotate the drum with the stationary part of the cylinder clamped in soft vice jaws. The whole assembly then drops into the LH case with the bottom gear teeth mating with the teeth on the shift quadrant.

Now for the most critical part of the assembly - ensuring the right amount of end-float between the direct gear and the end of the mainshaft splines. I first inserted the direct gear onto the mainshaft and slid it till it almost touched the right hand end of the mainshaft splines. I had a washer that was 0.29mm thick that I slid onto the shaft and rested on the RH side end of the splines. I then rested the direct gear on this washer. The specified range is 0.20mm to 0.30mm; I used 0.29 since I had a washer that thick. I next measured the distance from the front face of the direct gear hub to the RH end of the mainshaft. I measured this 4 times and averaged the measurements at 49.71mm. I removed the gear (keeping the washer still on the shaft) and with the metal ring (#61) 1.48mm thick in between the direct gear and the output bearing, installed the gear into the bearing and then installed the RH crankcase and tightened up the screws all round the case. This effectively "compresses" the whole gearbox assembly to its operating clearance. I now measured the distance from the face of the direct gear boss to the end of the mainshaft four times. In the blog post "Engine-the Crankcase" you can see the mainshaft protruding from the direct gear outside the gearbox bearing boss. The portion of the mainshaft you see protruding in the picture from the gear boss is the distance to measure. The average external measurement was 50.18 resulting in a 0.47mm decrease in clearance. Since the external measurement had increased it meant that the internal clearance had decreased. If the internal clearance has decreased a thinner spacer ring needs to be used. Since this was the thinnest ring from Jerry's collection, I ground it down to a thickness of 1.40mm resulting in an external distance of 50.10mm. The difference between the internal measurement (49.71) and the external (50.10) or 0.31 is just outside the optimal max of 0.30mm but within the 1.0mm max tolerance range - I would have had to make a whole new ring to get the 0.30mm. This was quite a painful exercise and the reason why you do not have the heavy crankshaft in the crankcase as the cases required quite a bit of manipulation.

The Engine - Crankshaft & Flywheel

The crank on the left is the one that came on the bike. Notice the length of the shaft closest to the bottom of the picture for the flywheel: it is far longer than the original crank on the right. The crank webs are a different shape. Also, the thread is a conventional RH thread onto which a standard nut was used to bolt on the flywheel without the LH threaded retainer ring. The second photo shows the crank with the nut covering the sludge trap removed and the welded up crank pin clearly visible. Heaven knows what happened to this crank to cause such destruction that the pin needed to be welded up! Jerry Kimberlin had an extra crank in excellent condition which is the one on the right in and that went in perfectly.

While I did not have to do this for Jerry's crank, on the old one I removed the safety wire and unscrewed the sludge trap on the flywheel side crank cheek. I cleaned out the inside and then shot compressed air through the hole in the crank shaft on the RH timing side to blow out any residue.

The way to assemble the big end is to clamp the rod in a vice with the big end bottom half of the bearing facing up. Coat the rod bearing with thick oil and lay the rollers into the curved half. Then lay the crank journal on the rod, coat the journal with thick oil and lay the remaining rollers on the journal. Put the con rod bottom cap on the journal and tighten the new bolts. Tighten it just enough to make the mating surface of the bottom cap lightly touch the con rod. With a wooden mallet tap the big end of the con rod to settle the rollers. Move the rod laterally to align the rollers. Then alternatively tighten the two bolts till the two split bearing faces are tight up against each other. Safety wire the crank weights and the sludge trap cover.

The crank tolerances are as follows:

Plain bearing big end : 35.05mm +0.000mm, -0.015mm
Crank pin journal dia : 29.00mm +0.010mm, -0.005mm
Big end rollers: 33 rollers with a dia of 3.0mm
Small end bronze bush: 20.00mm +0.007mm, -0.028mm
The bush extends 0.50mm on either side of the small end faces.

When these tolerances are exceeded and the crankpin & big end need to be reground:

First oversize:
Big end : 35.30mm +0.000mm, -0.015mm
Crank pin journal dia : 28.75mm +0.010mm, -0.005mm
Big end rollers: 31 rollers with a dia of 3.25mm

Second oversize:
Big end : 35.55mm +0.000mm, -0.015mm
Crank pin journal dia : 28.50mm +0.010mm, -0.005mm
Big end rollers: 28 rollers with a dia of 3.50mm

Con rod center to center length : 156mm
Con rod big end thrust face width: 23.8mm + 0.02mm, -0.02mm
Crank pin length across thrust faces: 24.0mm

The dia of the crankshaft journals that sit in the main bearings: 35.0mm, +0.01mm, -0.000 mm.


The Flywheel

Though this part came later after the gearbox was assembled and the crankcase closed, I am including the flywheel here as it is in the parts diagram above. My flywheel saga will take up a number of blog posts but it is not of any interest to anyone who has an intact flywheel. My flywheel seems to have suffered catastrophic damage at some point in its life and the taper had been rewelded with a much larger boss. This resulted in the flywheel protruding far beyond the crankcase, fouling the flywheel cover. The LH tread had long gone. Jerry Kimberlin's flywheel measurements were emailed and after a trip to two machinists - one to cut the taper and the other to cut the large LH metric thread for the retainer ring nut- the flywheel was ready for installation. I used Loctite Clover 280 fine grit lapping compound to lap the taper onto the shaft rotating the flywheel till I got a smooth finish on both the shaft and the taper. A dab of Loctite 609 retaining compound to take up any remaining slop between taper and shaft and the flywheel was mounted on the shaft.

The conventional RH threaded nut was bolted on and then the LH retainer ring spun on using a special tool from Guzzino. The tool is basically a pipe with tangs cut at one end to match the slots on the flywheel retainer nut. The way to tighten it is to hold the flywheel with one hand and tighten with the other using a foot long rod inserted through the special tool. Then, insert a piece of rope about the size of the spark plug hole into the spark plug hole and slowly bring the piston up to TDC. Somewhere along the way, the piston crown jams up against the rope effectively locking the engine. Then, as they say in the old restoration guides, fetch the rod a couple of whacks with a hammer to finish tightening and you are done. To remove the flywheel, first loosen the conventional threaded nut and then using the special tool, turn it clockwise to loosen it - it acts as its own puller to detach the flywheel from the taper.

The Engine - Bearings

Main Bearings

The LH case main bearing, which has to take the load of the offset flywheel, is a 35mm x 80mm x 21mm single row, cylindrical roller bearing. Look for SKF N307.

The RH side main bearing is a 35mm x 80mm x 21mm deep groove single row, unsealed ball bearing. Look for SKF 6307.

The LH flywheel side bearing has next to it a metal plate and a circular felt oil seal that sits in its cavity within the bearing housing. The felt oil seals were from Stucchi but can be cut from a felt sheet. I have used a circular saw but you end up with ragged edges. Jerry Kimberlin very generously agreed to make a batch of felt seals from me using a special cutter he made. He says that "it is not too hard to make a cutter. They are aluminum and on a mandrel that I put in the mill (or drill press if you have one, I don't). The cutting edges are concentric and as deep as needed for the felt used. In use, the cutter is put in the mill and spun against the felt which is on top of some wood for backing. Felt is easy to cut and is similar to other soft stuff like rubber, teflon, gasket material, etc. The difference between this cutter and a hole saw is that the cutter doesn't have any teeth but is dead smooth and sharp. Also the cutter has an inner and outer cutting edge so the felt ring seal is cut in one pass." Thus spake the oracle!

The RH side main bearing just sits in its cavity without any plate or felt seal. All the bearings except for the special size gearbox clutch and output bearing were from MSC Industrial Supply.

The clutch bearing in the LH case is a 30mm x 62mm x 13mm and the gearbox output bearing on the RH is a 33mm x 66mm x 13mm. Both these bearings were from Peters-Bearing in Germany. The two smaller gearbox bearings in the LH and RH cases are identical and are standard 15mm x 35mm x 11mm SKF 6202 deep groove, single row, unsealed ball bearings from MSC Industrial Supply.

The photo of the LH flywheel case shows the N307 roller bearing, plate and oil seal next to the cavity in the bearing housing and the oil seal on top of the incorrect clutch bearing. The second photo of the LH case shows all the bearings pressed in with the new original spec clutch bearing. The photo of the RH case shows the 6307 main bearing, the smaller 6202 gearbox bearing at the bottom and the special size gearbox output bearing. The metal disk in between the rear gearbox bearing and the felt oil seal has a step which should be facing the outside and be adjacent to the oil seal. I have reversed it in the photo to show the step.

The RH case in the blurred photo shows the special gearbox bearing with the four screw holes welded up and the bearings installed.

To check if the main bearings are correctly seated, push and pull the flywheel after it is mounted on the crankshaft. A radial movement between 0.03 to 0.05mm is tolerable but should not exceed 0.10mm - you should feel no or nearly imperceptible in and out movement of the flywheel.

Some quick notes on the 3 felt oil seals and the two metal plates. The felts are just as big in diameter as the cavity in which they fit. The ID of the main bearing felt is the same as the OD of the crankshaft that goes into the LH main bearing. The felt to seal the RH gearbox bearing has an OD equal to the diameter of the bearing housing but the ID equals the diameter of the spacer that fits on the gearbox shaft and the kickstarter. The metal disk for the LH main bearing is 1.5mm thick and has the same OD as the cavity for the felt seal and ID equal to the crankshaft. The second metal disk is for the RH gearbox bearing and is 1.5mm thick with the OD equal to the OD of the bearing while the ID is the same as the OD of the spacer that fits behind the sprocket. This disk, which goes in between the bearing and the oil seal has a 0.75mm deep recess cut into it on the RH side facing the seal with an OD equal to the size of the felt seal.

There are two brass bushes in the gearbox on both the RH and LH crankcase. On the RH side case, the upper bush is for the toothed gear selector shaft and has an ID of 19mm+-0.02mm with max wear clearance of + 0.04mm. The lower bush is for the gear selector drum and has an ID of 14mm+-0.02mm with max wear clearance of + 0.03mm. (The Super Alce, which was designed for military service mostly in the the Africa campaign, has 0.01mm more max clearance than the dimensions I have given which are for the GTV). On the LH side crankcase the upper bush for the toothed gear selector shaft has an ID of 15mm+-0.027mm with max wear clearance of + 0.04mm. The lower bush for the selector drum has the identical ID to the RH side of 14mm+-0.02mm with max wear clearance of + 0.03mm.

Next, installing the bearings. Since the key is to get uniform heat rather than a local hot spot that can cause distortion put the cases in an oven. For alloy 100 deg C (212 deg F) is a good temperature to aim for (to install any valve guide into a cast iron head use 150 deg C or 300 deg F). Classic Bike quoted former AMC metallurgist Don Hewitt advising that the melting point for typical British motorcycle alloy is 645 deg C (1,193 deg F) and 1,200 deg C (2,192 deg F) for typical British cast iron. If you don't have one of those fancy point-and-shoot thermometers the best way to gauge when the case is the right temperature is to look to see if Ms. Manners is watching and then spit on the case. If your spit sizzles then it is the right temperature. The previous day I had put the bearings in a zip lock bag and tossed it into the fridge. I took them out and put Loctite 620 bearing retainer on the outer ring. Then I gingerly lifted the hot cases out of the oven and put them on the workbench and with my heat gun reheated the bearing boss till it sizzled. I then dropped the bearing right in. With the exception of the LH side roller main bearing they all went in smoothly. The roller bearing (the inner race separates from the outer ring and stays on the crankshaft while the outer race is held by the crankcase bearing housing) needed a light tap (obviously on the outer races) with my bearing drivers to settle in. The tops of the races were all flush with the sides of the cases. If the RH gearbox output bearing needs to be tapped in do it gently as there is not a lot of metal behind and the entire housing could detach.

The Engine-Crankcase

The engine seems a real basket case: the flywheel was held on by a single nut and when the cases were split I found that the studs and bolts holding the two halves together were bent and had virtually all their threads stripped. The inside was no better: all bearings were rusted and barely moved. There was a strange plate covering the gearbox output shaft bearing in the RH case that seemed to be some kind of an oil seal. The outside of this bearing was even worse: the case had broken off when, I assume, the chain smashed into the cover. It later turned out that the bearing and housing were not standard and needed machining of which, more later!

First, I heated the cases in the oven and with a little help with my Makita HG1100 heat gun and knocked out the old main bearings. I say "knocked out" but after warnings from Patrick Hayes and Jerry Kimberlin about entire 3" disks of cranckase casting getting torn off, I amend knock to light taps on the bearing. The gearbox bearings, however, needed my Motion Pro blind bearing puller. I next sprayed the cases with Gunk engine degreaser. With the oil and grease removed it was time to head to the blast cabinet. I used "First Choice abrasive" from TP Tools & Equipment. It removes rust and scale quite rapidly but provides a smooth finish. The makers claim that the rounded beads prevent embedding in aluminium.

Once done with the blasting, it was time to weld the RH side case where chunks of casting had broken off just above the chain sprocket. I tried to degrease it again by using mineral spirits and Gunk degreaser since aluminum welding needs extremely clean surfaces. Using my Lincoln MIG welder and a roll of 0.035" ER4043 aluminum wire I slowly built up the surface. It was really messy because of the oil impregnated in the material. I then smoothed it out with a grinder to get the curves.

On the SA, the gearbox output shaft bearing that sits behind the chain sprocket is not a standard size and was not available till recently. Given the mayhem that occurred at some point and smashed the cases - locked gearbox(!), thrown chain - the bearing housing and the shaft must have been damaged. Someone added a metal ring to the bearing housing to allow the smaller diameter standard size bearings to be fitted (retained with a thin metal disk and four screws), machined away a recess on the outside that is used to keep the felt seal ring, and added a modern lipped oil seal (see photo above). To bring it back to original, I machined the outside boss away to the face of the bearing. I removed the metal ring inside the bearing housing to allow the use of the original larger OD bearing (a-a 66mm) which is now available and machined two aluminum disks, one to seal the outside of the bearing, and the second to hold the steel plate next to the bearing and felt oil seal (b-b 50mm) and the second disk (c-c 41.05mm) to hold the felt seal in place. I welded these two plates to the machined case effectively creating a new cavity for the seal. The photo shows the new housing with the shaft protruding. Lots of measurements of the distance from the inside of the sprocket, the various spacers, the length of the shaft and the width of the bearing were needed before I could figure out how thick to make the spacer. Too thick and it would rub against the face of the chain sprocket, too thin and the seal and metal plate would not fit. Patrick Hayes and Jerry Kimberlin's emails provided the answers of what the original looked like.

If anyone accidentally knocks out this boss the attached dimensions should help in the reconstruction.

I wanted to make sure that the faces of the crankcase were flat. I bought some plate glass and then taped 1000 wet/dry sanding paper to this glass and then rubbed the cases gently over the sanding paper till I got a shiny surface at all points of the case. The first couple of passes over the plate glass showed that the cases were not flat. Both cases needed quite a bit of smoothing before I managed to get an even surface on both cases.

Let me jump ahead a couple of months. When I finally assembled the crank in the cases and tightened the case up I found the crank binding. The cylinder did not fit into the mouth of the case. Clearly, I had removed little too much from the faces of the crankcase. I put a level on the face of the case and measured from the case face to the face of the bearing inner race. On the RHS it was 37.05mm and on the LHS 36.25mm for a total width from bearing to bearing when the cases are closed of 73.75mm. I then measured the crank at the hub which butts against the bearing races at 74.78mm. The difference is 1.03mm. I had to make a gasket which compresses to 1.02mm. I ordered 0.4" 6061-T6 aluminum sheet from onlinemetals.com, a place that Jerry Kimberlin suggested. I placed the case on the sheet and with a Sharpie carefully outlined the case. I cut the outline out with metal shears and punched out the bolt holes with my gasket punch set and smoothed down the edges. When I bolted the cases back up the cylinder slid smoothly into the cavity and the crank turned freely.

The various screws, bolts, and studs were acquired from MSC Industrial. The Italian verion of the parts manual has the thread sizes and dimensions for each fastener. The screws were 45deg slotted oval head stainless steel screws.

The cases were put together without any paper gaskets. Instead, on Jerry Kimberlin's recommendation, I used a thin smear of Dirko, an RTV sealer originally developed for the air-cooled VW engines all round the cases. It is expensive and can only be found on the internet but it works!

Next, the bearings....