Saturday, March 7, 2020

Crankshaft Truing 2: Checking and adjusting

OK so whether you have tried to replace a big-end bearing or simply have a used crank that's out of true, how do you go about checking and correcting this? You will need a lathe or a truing stand to rotate the shaft and determine the amount and position of any run out. In this description I'm assuming that you are standing in front of the shaft on the truing stand. The dial gauge is positioned on the right and makes contact with the top of the right hand shaft so run out is measured at the top of the right hand shaft.

There are only 4 ways the crank can be out of true and these show at different points in the rotation. Obviously they can occur in combination so they need to be corrected one at a time. So rotate the crank and find the position where runout is at a minimum. Set zero here and rotate again to find the point at which it is maximum. Look at the position of the crank at this position and in particular the position of the big end. This should be in one of 4 pisitions, either vertical, top or bottom, or horizontal, nearest you or farthest away. The position of the crank at max runout tells you the type or distortion you have. These are described below:

Tapering webs
The first type, and the easiest to visualise, is that the webs are not parallel: they may taper in or spread out from the crankpin. Tapering inwards means that max runout will occur when the big end is positioned vertically and down most. Conversely if they taper outwards, max runout will occur when the big end is positioned vertically but uppermost.  Correcting taper-in requires a thick soft metal wedge that can be tapped in between the webs to force them apart a little more: correcting taper-out just needs the webs to be pinched together slightly in a vice (soft jaws). Adjust and check run out again. If you are lucky then such adjustment will eliminate runout completely.

Swivelled webs
The second type of misalignment is harder to explain but occurs when the webs are not in alignment but one is swivelled around the crank pin relative to the other. This defect is manifested by max runout when the big end is horizontal, either forward, nearest the operator or rearward, away from the operator. Correcting this requires a heavy, but soft faced hammer to tap the base of the appropriate web to swivel it slightly around the crankpin. You will need to hold the crank with the crank pin horizontal and away from you, then tap down on the appropriate web nearest you to swivel it round the pin slightly. If it's highest on top right when the conrod is near you, then hold the crank by the left hand web, with the crank pin away from you and tap down on the righthand web. If the runout is highest with the crank pin horizontal but farthest from you the hold the crank by the righthand Web with the crank pin away from you and tap down on the left hand web (see diagram). Retest and repeat.



Runout greatest at top right when big end horizontal and nearest to you. Hold crank as above and strike where shown. 

Runout greatest at top right when  big end horizontal and farthest from you. Hold crank as above and strike where shown. 


If both misalignment types are present, as is likely if assembling a crank without a jig, the worst misalignment will be detected first. Correct that as appropriate until either the crank rotates true or a second misalignment shows itself. Correct that and retest. Keep going until the crank runs true or within 0.1mm of true. I tested a crank I knew to be good and this had a runout of 0.05mm.
I'm taking this approach from the following YouTube video which is very informative.
https://www.youtube.com/watch?v=X34qMzEjOnY

Tuesday, February 25, 2020

Crankshaft truing 1: Making a truing stand

Rebuilding the crank will be necessary whenever the big end develops play. This could be due to wear in the crank pin, rollers or conrod, and as you can't predict which has failed its best to replace the lot. Fortunately there is a rebuild kit available for the Quickly, although at a bit of a price (around £60 at the time of writing). Unfortunately the cost of a professional rebuild then puts the cost of rebuilding your crank close to that of a new replacement. To my mind this isn't green as well as being bad for originality! Disassembling the crankshaft requires a press, but this is useful for a load of stuff and if you're going to do a couple of these then it will soon pay for itself. Luckily I already have one so I'm going to try my hand at a crank rebuild. In theory it's straight forward, but of course it has to be accurate if it's not to shake the motor apart! The new crank has to rotate true to its centre and the webs need to counterbalance the movement of the piston.  This means the crank must be both "balanced" and "trued" . Provided you are not swapping for con rods or pistons of different weight, the crank webs should still balance the oscillating parts and balance can be ignored (I hope, obviously this wouldn't be true of high performance motors but I'm hoping to get away with it here).

Truing however is a different matter and unavoidable in any rebuild. I would also strongly recommend that this be done on any second hand crank you may acquire- even if the threads look great. Obviuously any wobble in the crank will be constrained by the main bearings. This can only lead to rapid wear but further, these bearings wont remove the wobble entirely. Since the magneto pole clearance is of the order of 0.15mm it will not take much movement before the flywheel contacts the magnets as it rotates.

If the crank is out of true then you need a way of rotating the crank and determining whether there is any wobbling movement off-centre (run out). You need to know how much movement there is and also in what crank position. This means making a stand, although I'm told rotating it by hand in a  lathe chuck should also work. I do have a lathe but it seems to me that the whole process would be easier to understand  if I had a stand, so this is how I made a primitive version of one.

The base is made from a 300 mm length of Box section, 75x25mm with a 1.5mm wall. To this I'll fit two uprights to hold the crank and either cut slots or holes in the base  to allow the uprights to be moved so as to accommodate different sized cranks.

The uprights are made in 3 pieces: A base and a top plate, each made from a 60mm length cut from 50mm wide, 4mm thick steel bar and an upright of 150mm length 25mm Square box section. These will be welded together with holes cut in base plate to allow it to be fixed to the base, and in the top plate to allow the bearings to be fitted.  I drilled 8mm holes at 10 and 40mm centres from one side of each plate, 10mm (top plate) or 15mm (base plate) from the top edge as shown.

Top abd base plates as shown, 4mm steel, 50mm wide bar

I drilled corresponding 8mm holes in the box section to match those in the base plate, and continued these through the box section, opening them to 22mm holes on the underside. This will allow me to bolt the bases to the box section and use a small socket or box spanner to tighten the bolts from underneath. The towers are designed to be high enough that the piston can rotate between them so that I can check balance as well.

The top plate carries bearings on which the crank will rotate, but also needs a 10mm square slot cut from the top edge centre (20-30mm from one side) where the bearings meet so that the crank can slip down and rest on the bearings. The assembled tower pieces are illustrated below.

Assembled tower sections for fixing to the box section base.


For rotation I'm using Im using 4 6382RS  8x25x9 rubber sealed roller bearings. The bearings themselves must be positioned quite close to the webs so that the bearings contact the straight and non splined sections of the crankshaft and this is quite short on the clutch side.


This means the bearing fitments must be as flush as possible so I used M8 flanged socket head screws, 20mm in length as these were the least bulky I could find. The bearings were secured using an M8 nylon washer as a distance piece and a nut n the back so that they could rotate (see below).

Bearing fitment using 20mm socket flange capped screws M8 and nut. I used a nylon spacer between the bearing and the top plate to allow rotation.
The length of shaft on the clutch side is short- you could use the narrower section further along but this will mean that the bearing would need to be raised. I found I could rotate the shaft without scraping in this position.

There is more room on the magneto side
Provided that everything is cut true and welded caccurately then everything should fit immediately. In my hands this wasn't the case and I was forced to construct 1 holding tower complete with bearings and bolt it into place. I could then assemble the second tower but without welding the top piece to the upright. I could then experiment with the position of the tower and the height at which the top plate was fixed so that the shaft could lie both horizontally and straight across both bearing pairs. Eventually I got a good orientation but it was untidy on the base. I welded the second tower into position rather than using the bolt holes because its position turned out to need more adjustment than I'd anticipated. It was simply easier to find the right position and tack it into place there. After all, any adjustment for crankshaft length only requires one moving tower.

Assembled stand, note one tower bolted for adjustment and one welded (bit of a bodge!). Bearings fitted with flanged socket head machine screws.



The base was made long enough that a magnetic dial gauge stand can attach next to the towers on each side.
The assembled stad accepted the crank easily and was convenient to use.
Stand in use



Wednesday, January 29, 2020

The transmission

If you've been reading my posts in order then you will by now be familiar with the myriad of gears in the transmission system. I was having trouble getting my head around how it works and how it interfaces with the pedals so this is my attempt to sort it all out. If you're familiar with this stuff then stop here- there will be nothing new and probably some stuff that's wrong.

The clutch

The clutch consists of the clutch base splined onto the crankshaft, the clutch plates located in the clutch basket, the clutch basket itself which bears a drive pinion molded into its top and the spring/release assembly located above the basket. The crankshaft is splined onto the clutch base located in the left hand case. This means the clutch base turns with the crank. The clutch plates fit around the clutch base, but the plates aren't simple discs. They have dog-ear projections. On the metal plate, the projections point towards the centre and on the friction plates they point away from the centre. The clutch base has dog-ear indentations into which the ears of the metal plates will fit. The clutch base and metal plate will therefore both turn whenever the crank is turning. Sandwiched above and below the metal plate is a pair of cork lined friction plates. These have dog ear projections on their outer rim which fit into slots around the edge of the clutch basket. The clutch basket can therefore only rotate if the friction plates are forced against the metal plate such that they are all dragged round together. This is achieved by means of the spring located above the clutch basket and compressed against it by the crankshaft end nut. This forces the basket down and thus pushes the friction plates against the metal plate. Under these circumstances as the crank turns the clutch centre also turns and as the plates are forced together drive is transmitted to the clutch basket and thus to the drive gear molded into the top.

The clutch is released by means of the clutch cup which covers the spring and locates underneath it using a special circlip. The cup can be raised by a cable-operated lever which raises the cup so also the clip in its base. This pulls p on the base of the spring, relieving its pressure on the clutch basket. This allows the plates within to separate. The metal plate then rotates between the friction plates which remain stationary and drive is disconnected.


The gearbox.

The gearbox consists of all the gears and gear change mechanisms  and interfaces with the  pedal system. The first point to realise is that all gears (including those of the pedal system) rotate all the time that the motor is running and the clutch is engaged whether the bike is moving or not. Drive is connected or disconnected to the transmission shaft and its chain sprocket by means of a sliding dog on the transmission shaft. This has three dog teeth on either side and slips along coarse splines on the transmission shaft. When the dog is slipped against  the desired gear the three dog teeth engage with those on the gear itself. Rotating the gear thus rotates the dog and as this is splined to the mainshaft, the shaft is also rotated. Slipping the dog thus locks the selected gear to the mainshaft whilst the unused gear is is left free to rotate idly on the shaft.

Top gear.

The crank rotates the clutch basket and the 15t drive pinion molded on top. This meshes with the large pedal gear (80t) which rotates but as it has more teeth it does so more slowly than the motor by a factor of 15/80. Gearing ratio is calculated by dividing the tooth number if the DRIVEN gear by that if the DRIVING, so in this case 80/15 or 6:1. However in top gear the 80t is dogged directly to the main shaft and this speed transmitted directly to the chain sprocket, although the difference in size of engine and rear wheel sprockets does reduce wheel rpm still further.

For first gear it's necessary to allow more powered cycles per wheel revolution: this means that road speed is reduced for the same motor rpm. This is achieved by means of two features molded into the rear of the 80t gear. The first is a smaller (18t) gear which  obviously turns whenever the large gear turns. This is meshed to the double gear column (24t/17t) on the layshaft. Both these layshaft gears thus rotate whenever the 80t gear is turning. The lower layshaft gear is 17t and this is meshed permanently to a 24t gear that rotates at the base of the main shaft. Consequently all these gears are in motion whenever the 80t gear is turning.  To enable gear change the two gears on the main shaft (that is the 80t double gear at the top and the 24t gear at the bottom) each carry 3 tooth dogs. The main shaft between the two is splined and a sliding dog can slip up and down this shaft, moved by the gear change fork. When the sliding dog is at the top it engages with the dog behind the 80t double gear. Since the sliding dog is splined to the main shaft this means that rotation of the 80t gear rotates also the sliding dog and this is transmitted to the main shaft through the splines. Although all other gears are turning the 24t at the base of the main shaft simply rotates idly on the shaft. When the dog slides to the bottom of the main shaft it engages with the dog on the 24t gear. Again this means that rotation of the 24t will now drive the main shaft through the splines. The 80t gear now acts as simply an intermediary in power transmission; It rotates idly on the main shaft transmitting no thrust to it. Instead power is transmitted to the 24t top layshaft gear. This turns the whole layshaft column, including the 17t  gear at its base. This is meshed with the 24t at the base of the main shaft so this is also driven, and since it's now dogged to the mainshaft this is also rotated. This chain of gears reduces engine speed by  a ratio of 24/18 and again by 24/17. Combining these factors gives a  first gear total reduction of 1.88:1 compared with 1:1 in second.


Friday, December 20, 2019

The minor problem of the Layshaft support or countershaft.


Stripping my second motor showed  a problem with the layshaft or countershaft. The double gear column lacked the lower shim and there were clear wear grooves on the collar of the layshaft with some marks opposite, on the case. 

Layshaft suport (gears removed) there was no shim on this lower collar and wear grooves are evident on the collar upper surface.

Investigation showed the shaft itself, which is pressed into the rhs case, was not pressed in right up to the collar, although it was pressed in flush with the case on the other side. There was a clear gap between collar and case easily accepting a 0.7mm feeler gauge. 

Base of layshaft, clear gap between the collar and case beneath

Base of layshaft viewed from the other side- end of shaft is flush with case.
Test fitting the gear column showed that in fact it could not fit into the cases because when positioned on the countershaft; the gears kept the cases apart. This is probably why the lower shim had been omitted, but even so the gears were still too high for the available space in the case and dragged significantly on the cases when rotated thus accounting for the wear grooves on the collar. The height of the gears must have prevented close fitting of the cases and probably accounts for the signs of gasket leakage on this motor.  

I heated the cases to 100 deg in a water bath and then selected a deep socket to sleeve over the shaft and bear on the collar. 
Deep socket used as sleeve over the layshaft to bear on the collar itself before using the 10t bench press.

I used a large socket beneath so that the end of the shaft could project slightly below the metal of the case. The shaft pressed in to its lowered position with relatively little force in the press although I had failed to move it in the bench vise. 

Shaft pressed home, collar flush against case.


Base of layshaft support now protrudes slightly on magneto side of case. It's base is visible app. 9 O'clock. 


In this lowered position the cases could close over the gears which remained free to rotate. However, when I fitted new shims above and below the gear column (as the old ones were either very worn or missing) and compressed the cases together, the gears did lock up. This I found was alleviated when the gasket was included.  Overall I think its a toss up as to whether it's better to push the shaft down and fit the extra lower shim, or just run without shimming the bottom. I think its even possible that the shaft had been positioned to eliminate the requirement for the lower shim at the factory, although I can find no mention of that approach. In the end I chose to run with the shims and it's worked out fine. 













Saturday, October 19, 2019

Rebuilding the motor 3. Rebuilding my spare motor

My first time rebuilding an NSU quickly motor- for a simple motor it can be an awkward little so and so to get running. I'm very much indebted to the FB groups in both UK and Germany and Shane Woods, Engine rebuilder,  for very patient/tolerant advice. Some of my approach is overly complex, but hey, this is what I did, how I did it and why!

I would add the following guidance gleaned from various things I did wrong.
1. The manual speaks in places of determining clearance and fitting shims to adjust. I found that this was seldom necessary or even feasible, most if not all of the clearance being due to the gaskets! Just refit the shims you find in the positions they were found.
2. Fit new bearings and bushes into preheated cases (up to 160 deg C), fit oil seals into cool cases (or 100 deg C max).
3. If you have time, pre-assemble the motor dry (i.e. gaskets but no sealant). Tighten case screws to 7 lbft and check everything moves freely. Pay particular attention to the crankshaft and layshaft (aka countershaft). I do this test assembly after I've fitted the bearings but before I've fitted the oil seals to avoid damaging them.
4. Do not be tempted to replace any "missing" shims... Or at least measure carefully if you do because they may be missing for a reason. This applies especially to the layshaft shims.  If you must fit new shims then aim for the maximum permitted clearance (ie thinnest shims) as gasket crush will always reduce, or even eliminate, clearance.
5. If using metal cased oil seals from the official suppliers take great care that they do not distort when fitting. I found it was often helpful to fit the shaft first from behind the bearing. I could then slip the seal over the shaft to centralise the seal and tap it down into its recess using a sleeve drift (such as a deep socket) over the end of the shaft and bearing on the outer  rim of the seal
6. New cases may need honing for clutch turret and pedal shaft see below.
7. Small end bushes do differ in internal diameter, although this doesn't seem to be mentioned by the suppliers. My opinion is that these bushes should always be supplied with an ID smaller than the pin. If the bush accepts the pin before fitting it will require a greater crush to generate good fit on assembly. This means its more sensitive to any wear in the con rod eye... Of course there should not be any such wear but its not impossible. See my earlier post considering the piston, rings and small end bush.
8. If you are fitting a used magneto back plate always test the ignition coil before you do so and change the points as they seem to deteriorate in storage.

Preparing the cases- knocks and dents

The first step for me is to prepare the cases. These were dirty and there was a fair bit of corrosion debris so a good clean came first. Then I could set about sorting out any damage, I had clear problems with the LHS case where the cover mounting screw had broken off in its hole and the mating face was chipped; both needed fixing. 

Tackling the screw first, I used a countersink bit to clear away the distorted screw opening and then centre drilled the screw stud

Screw stud broken off in its hole.

centre drilled screw stud
I could then drill into the stud and use a stud extractor to remove it- this is a risky business as the smaller extractors can snap off in the hole. The stumps of these are very hard and usually call a halt to any chance of drilling out the stud. If this happens you can sometimes weld a blob of metal onto the  broken end of the extractor (it doesn't stick well to aluminium), cut a slot in that and unscrew the remnants of the extractor (clockwise obviously).

Once the stump was removed I drilled the hole deeper (5.2mm drill) for tapping to M6. I will then fit an aluminium threaded spacer to butt up against the cover and replace the missing screw pillar. Its also a good idea to check at this stage that all screw holes are not distorted and all threads in the rhs case are clean and functional.

This case was also clearly damaged around the mating surface to the centre case. I needed to resurface the edges and I do this on a sheet of Emery stuck over an engineers flat surface. However this case also had screw sockets pressed into it and these projected below the edge and had to be removed before I could lay the case flat to reface it. They were corroded in and quite stiff- don't hammer them as you will probably snap off the case "ears". Usually, I would just press these out in the vise into a receiving socket, but in this case there 'enough room to slip a socket over the stud. I got around this using a crow foot spanner to surround the stud on three sides
Using a crow foot spanner to surround the socket without touching the case. There is insufficient room to use a socket here.
... before pressing the socket out in the vice.
Pressing out the socket in a vice- complete the process using a push socket
Once the sockets were removed I could flat the case on emery paper.
Flatting the case on an engineers flat surface.
This was reasonably successful and I got a smooth surface over most of the case. Sadly it had suffered some screwdriver leverage in the past and had been slightly bent and burred along the top edge. I was able to remove the burring and get an almost continuous flat edge for the gasket, but removing the dent completely would have meant removing too much metal and affected all the clearances on the shafts inside. I'll have to live with it and I'm hoping that as the damage is at the top it wont be a flooded with oil and so shouldn't leak. The margin on the centre case is perfect and undamaged so I'm going to stick the gasket to that case with gasket compound and put a smear of RTV instant gasket  along the damaged section of the LHS case to fill any space. Hopefully this will work. However there is an element of doubt over this case and it may well end up being replaced!

New Cases
If you are using any n.o.s. cases they may need a bit of preparation. I eventually had to get a new clutch cover but you will need to hone the inside of both the clutch turret and the pedal shaft bore to ensure that the O rings on the clutch cup and pedal locking member will enter and slide smoothly.

Honed clutch turret bore, 3 stone draper small cylinder hone shown.


Preparing the cases; removing Bearings, Bushes and Oil seals

Once cleaned I removed the oil seals by tapping through the bearings and then tried to remove the bearings themselves. In theory heating the case to around 160 deg C should loosen the bearings such that they slip out easily. This temperature is a bit high for convenience without and oven but results can be achieved with a hot air gun or boiling bath. Note here that the case needs to be heated evenly to avoid distortion so a blow torch is best not used- or used with great care.
My main bearings were badly corroded and showed no sign of moving! I had to ignore my own advice and resort to the blowtorch and a Draper blind bearing puller with slide hammer but at last the mains came out. You must heat the case evenly.
Tried a hot air gun first- no joy!

Draper blind bearing puller

Success, bearing removed.
The socket from which the bearing had been removed was still caked in corrosion so another clean up job here. There is an oil hole bored from the cylinder base to the bearing  and on the LHS case this ends in a circular grove extending around the bearing that allows lubricant to move round the bearing. Clean out the oil way with a pipe cleaner and check the groove. I found this was filled with corrosion and needed to be cleaned out.
Socket after removal of bearing- bad corrosion needs removing. Opening of the oil way from cylinder base visible at app. 7 O'clock
I removed the bearing from the RHS in the same way. This bearing has a circlip to separate it from the oil seal on the other side. The oil hole on this side terminates in a transverse groove to distribute the oil rather than a circular one This also was blocked and needed to be cleaned out.
RHS main bearing removed-oil seal visible beneath, and circlip inside
I tapped the oil seal out from the rear and then removed the circlip with circlip pliers- again this was quite dirty and I needed to clean up both the clip and its groove as well as the bearing and oil seal housing.
Removing the circlip... Note the opening of the circlip coincides with the position of the oil way from above so that oil can flow along the groove in the bearing housing- visible here before it was cleaned out at app. 1 O'Clock.



Fitting new Bearings Bushes and oil seals

As I learned later... you need to install new bearings and any bushes before fitting oil seals in case the latter are damaged by the heat used to install bearings- although they should be OK up to 100 degrees. I didn't realise I needed to change the gear change bush until a little later.
Once all corrosion was removed I could fit the new bearings. To do this I boiled the case in a water bath and chilled the new bearings in the freezer for three hours. This did improve the fit but I still needed to tap the bearings home using a socket to press only on the outer race and supporting the case carefully on a wooden block.
Crankshaft main bearing and main shaft bearing fitted into RHS case. Note bush for the gear change shaft (upper centre) which I should have changed first but decided later to replace.
Crankshaft LHS main bearing fitted into centre case
Main shaft LHS bearing fitted into LHS case

Gear change push rod bush.

I found that the gear change push rod seemed to have a lot of play in it. Having measured the shaft, this wasn't appreciably worn along its length so I concluded it must be down to the bush. I tapped the old bush out using a suitably sized socket followed by a drift.
Drifting out the old bush- care not to scratch the case socket
The new bush is chamfered on one side and should be inserted from this direction.

Chamfer on new bush insert chamfered end first.
To keep the new bush square and hopefully prevent it closing up as its pushed in I fed the gear change shaft through the bush and tapped on this and the socket simultaneously to insert the new bush. Note that the top (inner) end of the push rod is chamfered so always insert it chamfered end first. I.e. from the outside towards inside of the motor. Here you can see the end result with the bush and shaft now flush against the case.
New bush installed, shaft in centre to keep it square.

Shaft removed after fitting the bush, but bush now too tight.

Unfortunately this had closed the bush a little and the shaft was now far too tight. I tapped the shaft out again and then reamed the bush using a 6mm twist drill rotated by hand. Ideally a 6 mm reamer would be good but I didn't have one. Got one on order now! This bush is for sliding rather than rotary motion so I don't think this is too critical.
Using a drill bit to hand ream the new bush

Oil Seals

The push rod oil seal is awkward. It needs to be tapped in square, but its very small, access is awkward owing to the gear change fork which I didn't want to remove (its mode of attachment isn't obvious- see below*) and its metal case distorts really easily. In the end I reinserted the push rod through bush and oil seal from the rear and used a section of metal (in this case an old cupboard connector) to spread the load. I could then tap it home using a drift through the gear change  fork.
Tapping in the gear change push rod oil seal- use a metal plate to spread the load and tap evenly

Push rod seal installed
*Nb: Adding a note here! When I tackled my second motor I found that this oil seal hole had been seriously distorted by a previous owner's attempts to insert the seal off-centre. The seal's metal case was distorted and once I'd removed it I could see that the crankcase seal recess was damaged with metal squeezed into the bore preventing any oil seal from seating. The only way to sort this was by cleaning out the recess using either a 13mm flat end mill or, because I didn't have one, a 13mm drill that has been re-profiled to as flat an end as possible. However in order to do this you will have to remove the gear change pivot arm. Its not difficult- but risky as on these later motors, its not intended to be removed! Early versions used a longer pivot pin which extended through the mounting pillars and was secured by circlips on each end- these would be no problem to remove. Sadly in my motors these pivot pins are all peened into place and the only way to remove the arm is by tapping back the aluminium on one side until the pin can be finagled out. Its replaced in the reverse manner- but obviously you can only bash back and re-peen so many times.... Perhaps a conversion to the circlip mount would be possible but access would be very tricky. Anyway once the  arm was out of the way I could mount the case on the drill base plate and centre the drill on the bush using an old gear change push rod held in the chuck and passing through the bush. Once clamped, I could then swap this for a the flat profile drill and clean out the recess as shown below. I found this successful but I think if doing it again I would actually enlarge the seal recess to 14mm using a 14mm flat end mill. This would then accept an easily available 6x14x6 mm nitrile oil seal available for instance from Simply Bearings.

Push rod oil seal recess re faced using in this case a flat profile 13mm drill

I was then able to fit the remaining seals. I have to say that the metal-cased oil seals supplied by the specialist suppliers are a bit awkward to fit because there is nothing behind the metal and it can easily distort. This is especially true of the pedal shaft seal which is so thin it can twist out of round if not struck squarely during fitting. I found that this can be prevented (as in the case of the gear change push-rod above) if you slip the shaft through the seal from the opposite side and use it to centralise the new seal. You can then tap the seal home using a socket as a sleeve to fit over the protruding end of the shaft and bear on the edge of the seal.
Pushing the pedal shaft through the seal...

... and tapping seal home using a socket to sleeve over the end of the shaft.
This particular seal is an odd one with the seal lips projecting obviously backwards beyond the metal case. It would be great if you could fit a conventional nitrile seal here as the temperature shouldn't be a problem. An R23 double lip seal wouldbe good but sadly you need a (21x27.5x4) seal and I couldn't find a manufacturer for this. It would be nice to mill out the housing to 28mm from 27.5 as there's more choice at this size, but there's not a lot of metal in this location so I used the NSU spare part. Maybe as a last resort...

The Internals- Crankshaft. 

Once bearings and oil seals were fitted then I could start reassembling the motor internals. I started with the crankshaft. Here I'm fitting a replacement (a good used item*) crank and this means I'll need to recheck the shimming on the shaft. To do this you have to first work out the size of the crankcase cavity and then subtract the width of the crankshaft itself. I'm showing below how I did this but I would add that having now dismantled three motors, I have yet to find any shims on the crank at all! In this case I was fitting a different crank into these cases, so I needed to be sure but in the event I think no shims were really needed at all.

*I took this as OK to use as it came from a reputable supplier. I think it was just about OK but in future I would always check that a used shaft runs true. See my post on crank rebuilding and truing.

I first measured the depth of each case individually using a metal bar as bridge to support the depth micrometer across the space. I could then measure from the top of the bridge to the inner race of the crankshaft main bearing where the shaft will sit when fitted. This was surprisingly similar on both sides at 0.805 and 0.806 inches. However you need to include the thickness of the gasket which is measured on one side only increasing the second measurement to 0.825. This measurement however includes the thickness of the metal bar bridge two times so this has to be determined and subtracted.
Measuring depth of crankcase half using depth micrometer on a metal bridge. Measure to inner race of main bearing.
Measuring the second case half including the gaskets thickness.

I measured the metal bridge with an outside micrometer as 0.205". Consequently the width of the crankcase internally (with gasket) is 0.805+0.825 - (2 x 0.205) or 1.2 inches.
Measuring the thickness of the metal bridge support
I could then measure across the crank itself using a caliper as 1.19 inches.

Measuring the crank

Subtracting the crank width from the case space available gives a freedom of movement (end float) of 0.03 inches. The manual specifies the measurement as 0.008" to 0.012" maximum so a 0.02 shim will reduce the end float to within this range and still leave a little for any gasket "crush". 

I lubricated the bearings first with red assembly lube which mixes with oil. Using this means that the motor can be stored for years without the lubricant draining from the bearings. Although this lube is intended for 4 stroke motors, it should disperse well here although might make initial running a little smokey.

Red lube grease in bearing
Reassembly obviously involves pushing various shafts into the new bushes. These are a very tight fit and were removed only with difficulty and heat. Clearly you can't warm the bearing once it's fitted without upsetting it's fit in the case, and heating a bearing you intend to use isn't a good idea! You should find that all shafts will fit cold although they will probably need to be polished first and fiddled into place. If you still have difficulty then you can use heat transfer: find a socket that slips smoothly into the bearing, heat it with a blowtorch and then slip it into the bearing for a few mins. This will expand the inner race (probably lock up the bearing temporarily) then remove the socket and insert the shaft. Provided the socket isn't too long it also allows you to warm the bearing centre without damaging a pre-fitted oil seal on the other side. If necessary chill the shaft first but as I was doing this in November in the UK that wasn't necessary!


Hot socket in bearing centre
It slipped easily into the bearing once it had been contact warmed.
Crank fitted

I added the 0.02" shim to the RHS of the crank- this was probably a little larger than required.


Pedal ratchet (Driver and Locking member)

The pedal ratchet system works by means of a driver which is keyed to the cases so it cannot rotate. It bears a screw thread internally which is mated to a thread cast in the pedal shaft.  Once the pedals are fitted the pedal shaft cannot move left or right, and rotating the shaft propels the driver to and fro along the shaft. However take care never to rotate the shaft when the pedals are not attached because the screw action can then dislodge the driver key from its location internally. This movement of the driver brings it into contact with the locking member secured in the case at the rhs of the pedal shaft. Both driver and locking member carry ratchet mating faces so that they can lock together.

The locking member has a ratchet at one end to engage with the sliding driver and a splined section at the other on which the brake lever will be mounted.  The locking member has 4 O rings, 2 are internal and 2 external- and in my case all 4  looked flattened and a little chipped. I decided to replace these, a job I had thought would be quite hard but in the event turned out to be simple.

Locking member segment as removed. Note 2 external rubber O rings. The splined section carries two grooves for the circlips that secure it to the case and hold the rear brake lever in place.

One of the inner O rings shown here,  the ratchet end is visible

Here the internal O ring at the splined end
I removed internal O rings with a sharp 90 degree pick being careful to stab this into the rubber only and not to damage the groove.
New O rings just pressed in internally using a finger, they popped into their grooves surprisingly easily. External O rings were removed by squeezing them along their length between finger and thumb. This stretched them and caused a small bulge which the pick could slip under. Replacing the external rings has the problem that they must pass over the geared end which is sharp and could damage the ring. I shrouded this is a small plastic sheet (a parts bag) and twisted the top together. The rings could then be eased over the plastic and rolled down to their grooves and finally fitted in by removing the plastic.
Shrouding the locking member with a roll of plastic bag

Ease the first (bottom) O ring over the plastic which protects it from the geared end and also stops it slipping into the top  groove

Adding the second O ring.
I think the locking section should be fitted into the lhs case and secured with the first circlip before the cases are assembled. If you do not do this then the locking  member can be pushed inwards along the shaft and can be hard to retrieve.

Layshaft  (counter shaft) or Double Gear Unit

The two speed transmission is achieved by means of a layshaft sleeve bearing two gears (24t/17t). This sleeve is a vital part of the gearbox and to my horror mine wasn't usable; when I'd cleaned it up I discovered to my cost that the smaller of the gears  was clearly damaged. I had wondered where all the metal fragments had been coming from and this at least solves that mystery. In fact I think this could be filed so that the gears could engage but I suspect it will always be stiff to move and noisy. Also since some of the teeth are damaged force transmission may be reduced and the teeth are likely to be stripped if under load. I was therefore obliged to source a replacement double gear (good used item).

Damage found on smaller of the two gears- It will need replacing.

Mainshaft check and renovation.

I did find that there was a fair amount of ground metal debris in the teeth of most of my gears. I cleaned these up as I fitted them and the pics may not always show the cleaned gears as I took those whilst I was fitting various pieces.

Bearing greased before inserting main shaft- note debris in gear teeth of main shaft. The shaft and gears were cleaned before fitting.

I was obliged to clean the main shaft thoroughly. Whilst I was doing this it was clear that the lower gear was quite loose on the shaft. Again measuring the shaft showed that this wasn't greatly worn so I was obliged to change the bush in the middle of this gear. Fortunately this is easily available.
I pressed the old bush out in the vise using a suitably sized socket

Bush emerging from gear
... and completed the job using a larger receiving socket on the other side.
Using a receiving socket to complete the extraction
The old bush did seem thinner than the new one, which wouldn't actually slip over the shaft at all.
Gear plus old (upper) and used (lower) bushes
I pressed the new bush in again using the vise and a small socket, I pushed it just a little too far and then pressed it back without a socket to ensure the front side was flush with the gear.
Flush finishing the bush against the gear face
Lastly I reamed the bush using an adjustable reamer until the gear was a smooth sliding fit on the shaft.
Reaming the new bush

Job done- good sliding fit, washers on both sides of the lower gear.
I also noted that the old Woodruff key had sheared off in its keyway in the shaft, this was removed using a fine punch. Note this key is 2mm x 3.5mm

Woodruff key sheared in slot on mainshaft

Luckily the remnants tapped out easily.

Fitting the Mainshaft

I started reassembly in the right hand case. I used the same process of heat transfer to insert the main shaft into the main shaft bearing- Not forgetting the two washers on the main shaft; the larger one below and the smaller above the dogged gear.

Main shaft inserted, washers beneath the lower gear and above it
The gear change push rod should be inserted from the RHS (ie from the outside) but in my case I had already done this when I swapped the bush. The gears and gear change mechanism is then built  up on these two shafts.

First, add the spring to the gear change shaft, then the fork and then slip the upper dog on the main shaft such that it engages with the gear change fork- its a bit fiddly but easy enough. The only thing to watch for is that both fork and gear dog (claw) have to be fitted the right way round. The tag on the fork has to point downwards at the spring, whilst the dog gear has two sides. The manual isn't very helpful about this because it simply says that the smooth side should be fitted facing the alternator. As the alternator will be beside the shaft the top or bottom can never face it! I think they mean it should face outwards, away from the centre of the engine. I struggled to find a difference between the two surfaces of the dog gear and for this I' m indebted to the folk of the FB group.

Two sides of dog gear or claw. The surface arrowed has a recess towards the centre. This face should  face downwards on the mainshaft i.e. towards the bearing (pic from Bernd Bothe)

Luckily its clearer when the gears are fitted because the dog sits lower the correct way around.
Incorrect orientation, note large gap below the dog gear and the stepped recess against the shaft visibleat the top.


Correct orientation- the dog gear sits flush against the lower gear, the dogs visible on top have no recess against the shaft.

Fitting the Pedal shaft and double gear sleeve.

I added the pedal shaft and the double gear sleeve, 17t gear downwards. You may need to rotate the gears until everything engages in position.



Gear change mechanism

I could then add the spring and gear change fork to the gear change shaft (lug points towards spring). Compressing the spring by hand let me slip the washer (not shown in parts list) and C clip into place to retain the fork. I found this extra washer in this position when I stripped the motor and decided that it may well be useful there. (I should add that I've now stripped three motors and all three had this extra washer, I'm coming to the opinion that its meant to be there).
Gear change fork in position above spring and retained by washer and C clip.
Turning the case over I was then able to add the washer and C clip (missing when I stripped the motor) to allow the lever to move the gear change shaft.
Gear change lever held by washer and C clip on the external side of the case. 

Fitting the crankshaft

I added the 0.02 mm spacer to the right side crankshaft...
Shim added to right hand side

and slipped the crank into place in the right hand case. The shaft was a little burred and needed to be smoothed and polished before it slid in easily.


At this point I coated both right hand and centre case edges with Blue Hylomar gasket sealant and stuck the gasket to the centre case before lowering this onto the right and feeding the various shafts into their corresponding openings. The case slipped down easily and I retained it with the central bolt and nut (anti-shake washer under bolt head and nut). Do not tighten this fully at this stage as it will distort the cases.
Centre case lowered onto right hand side incorporating gasket and Blue Hylomar sealant.
Central bolt inserted to hold the cases together. 


At this point I wanted to insert the various peripheral bolts to clamp the cases together. However this requires the third case to space the bolts. I fitted an old outer case with no bearings or oil seals so that I could clamp the cases together with a few case screws and left it overnight for the gasket sealant to set. I removed the cover the next  day and could then add the driver to the pedal shaft engaging its clip in the case slot
Driver section added, note clip engaged with case slot molding.
Fitting the locking member (gear/ratchet sleeve) has to be done carefully to avoid damaging or displacing the internal O rings. NSU provide a small tapered insert that can be temporarily fixed in the pedal shaft cotter pin slot to prevent the O ring from binding there. I  don't have this so I just used plenty of grease and I think it went on OK. I did manage to fit the case this way but the process pushed the locking member along the pedal shaft and it was tricky to retrieve the splined section from the outside. A better approach is to fit the locking member into the clutch cover fully, and secure it in place with the inner circlip. The locking member can then be pushed down over the greased pedal shaft as the cover is fitted.

Driver slipped onto the pedal shaft and pushed down onto the  locking section. You may need to rotate the shaft to get everything to line up properly.

Fitting the Clutch

Make sure the clutch base is clean- remove any burrs or corrosion from the rear where it will penetrate the crank case oil seal. The base should have an O ring beneath it and a C clip above it. Both were missing when I stripped and so need to be refitted now. Pop the O ring into groove on the back, grease the case seal and then slip the base down the crankshaft engaging the splines and push it home. Secure it with the C clip. It might be better to slip the O ring down the shaft instead of fitting it in the groove to get it to slide more easily over the splines.

O ring in groove on base
Clutch base fitted to crankshaft

There is a special tool to press the clutch base into place, This is a sleeve like fitting that slips over the crank and can be pushed downwards by tightening the nut. This could be replaced with an appropriately sized deep socket- although I found that provided the splines were clean no tool was needed.
Secured with new E clip
One lined and metal plate added, one more lined to go.

I soaked the clutch plates in 2 stroke oil overnight and drained\wiped them before fitting. I had to supply 1 new plate as one had been missing when I stripped the motor. In fact it wasn't clear to me whether this was a 3 plate or 5 plate clutch so I will need to check that before reassembly. For the time being I assembled as a three plate unit. The plate that had been present was virtually unworn. Place a lined plate in first and then the metal centre plate engaging it with the grooves in the clutch base.

Add the second lined plate and swivel both such that the lugs on the top plate  fall into the gaps between the lugs on the bottom plate. At this point I could see that there was a lot more room on the clutch centre. I tried adding two more plates- one metal and then a final oil-soaked friction plate but the stack was then too high for the clutch centre. I concluded that this motor must have been fitted with a two plate clutch as is usually (but not invariably) found on 2-speed Quickly motors..

Three plates in position- clutch centre still stands proud with room for additional plates.  

Add the clutch case, you'll need to rotate it slightly to make sure the plate lugs fit properly. The clutch case also bears a gear for motor transmission which needs to engage with the final gear to be fitted to the box and this is the large dogged gear that slips over the main shaft. This gear also covers the small fixed leaf spring in the case and so will be to some extent spring mounted, The dogs face downwards to engage with the main shaft but make sure they do get pressed down adequately or the dog section will jam on the double gear column and lock the gearbox. Press the large 80t gear downwards and rotate it slightly to make quite sure that its properly engaged.

Clutch cover fitted aligning the internal plates. The gear on top of this engages behind with the large dogged gear fitted to the main shaft to provide drive from the motor when the clutch is engaged.
Fitting the 80t gear, make sure its engaged by pressing it down and rotating slightly.



Close up of gear engagement.
I could then fit the clutch spring which is sandwiched between a top and bottom spring seat. The bottom seat still has the original 6002C bearing from the strip so first step was to remove the old bearing and replace it.


Spring bottom seat with bearing
I used a large receiver socket and a small drift socket to tap the seat out of the bearing.
Bearing on receiver socket

Drift socket in place

Bearing removed
I pressed the seat into a new bearing pressing only on the inner race
and then fitted the spring seat to the crankshaft

Add the spring and top seat and the two wavy spring washers. 

I used the spring compressor again to shorten the spring and add the top nut and both spring (wave) washers, but in order to tighten it you need to lock the motor- no problem stripping but now its been fixed it it will obviously rotate. I fashioned a locking tool using two pieces of bar, cutting a short  section with a cutting disc to engage with the gear.
This was welded to a bridging section drilled to locate over the crankshaft and two bolt holes in the cases.


Assembled tool

Tool fitted to case and locking the large gear.
Once the tool was fitted I could tighten up the clutch nut. Tighten until the clutch starts to slip which is shown by movement of the con rod. Slip the clutch cup over the spring and secure with the base clip.
Tighten up until clutch slips
Before fitting the lhs case the manual suggests that its necessary to check clearance between the case and the large gear. Determining the room in the case was easy enough with a depth micrometer




Determining the amount by which the large gear protrudes is more difficult- and probably meaningless to determine. This is because that this gear rests on the leaf spring and the manual isn't clear whether this clearance should measured with the spring compressed or extended. I found that if I allowed the gear to rest on the spring naturally then it protruded more than there is space in the case! If the spring is to be compressed, then it would always be pressed against the lhs case with zero clearance. Correcting this to a specified value would always need a single size of shim and no measurement would be necessary. If the gear protrusion is measured from the compressed state then the gear is always going to be firmly pressed down removing any flexibility in the spring. I therefore refitted the existing shim at 0.02 mm.
Measuring protrusion of the gear

Shim fitted
Finally I positioned the clutch cup and fixed it with its clip before I assembled the cases, I stuck the gasket to the centre case with blue Hylomar. I then coated its upper surface with Hylomar and pressed the cases together. This completed the case assembly, but I found that after a partial rotation the motor jammed solid and would turn no further.

In order to solve these problems I was forced to replace re-strip the motor and both the lhs cover and the clutch cup. I suspected this problem could have arisen from a distorted clutch turret which had clearly been struck at some time in the past. I decided to fit a NOS clutch cover and since I was changing the case I needed to insert a new bearing and oil seal into the replacement. The case bolts should be  tightened to only 7ftlb (9.5 nm) to avoid distorting them or pinching the internals. However I still found that after a turning a little the motor would jam solid. This I eventually traced to the clutch cup and spring clip which can slip downwards and jam in the 80t gear wheel when the cup is pressed to its lowest possible position. This will happen when the case is fitted as it must push the cup downwards as  its pressed on. The cup then needs to be levered up a little to free the gear. 
Clutch cup at its lowest position jammed in the teeth of 80t gear
I was surprised to find that there is no mechanism to raise the clutch cup and avoid this jamming; the cup clip prevents the cup being raised (without operating the clutch) but cannot stop it from moving downwards. Instead the motor seems to rely on correct adjustment of the clutch cable when fitted to the bike and consequently, jamming like this may be normal when the motor is bench rotated in its motor stand using the pedals. It is also necessary to ensure that the large gear is pressed down onto its spring. This not only helps the gear to clear the bottom of the clutch cup, but also moves the dogs on the back of the gear inwards and out of contact with the double gear column. Its therefore important to make sure there is sufficient shimming on top of this large gear to achieve both of these aims. This confusion caused me to undergo several rounds of dry reassembly but clearance was eventually sorted and rotation restored. I could then apply Hylomar and reassemble the cases.
I changed the little end bearing as described in an earlier post. Note that I do find some new bushes are already too large for the pin, I didn't use those. Adding a piston and then slipping on the cylinder and a refurbed head as described before. 

I could then refit the magneto- no problems in the fitting but I did find that one of the inner flywheel magnets jammed on all 4 coil shoes as the flywheel was rotated. This could have been caused by a misaligned crankshaft (see my comments earlier about the desirability of truing used shafts before fitting) but in fact in this case it was a simpler cause as the flywheel was slightly out of round. I suspect it had been dropped at some point. Luckily I had a spare and this restored free rotation. I then found I had no spark! I removed the flywheel again and checked the  ignition coil. To do this you must unsolder the black wire connection from the HT coil to the capacitor and check the  resistance of both coils. The LT coil (freed condenser wire to earth) should be low resistance around 2.5 ohm. The HT coil from HT lead (cap removed) to earth should be high resistance around 5Mega ohm. I found the LT coil was fine but that there was no connectivity via the HT coil at all suggesting an internal wire break. I confirmed this using a 1.5V pen cell test. This is easy enough to do once the LT capacitor connection has been unsoldered. Refit the HT cap and fit a spark plug, lay it so that the plug can earth well to the magneto back plate. Then flash connect a 1.5V pen cell between the freed unsoldered black wire and the shoe of the HT coil (earth). As you remove the connection you should observe a spark at the plug. It will be very weak because 1.5V is well below the usual operating voltage, but if you get one with 1.5V then you will certainly generate one when the motor turns. There will also be a spark where you flash the pen cell to earth but this is caused by self-induction in the LT coil and means nothing, you must see a spark at the plug. If you don't then the coil, ass in my case, is dud. Luckily I had a spare back plate in which the coil was functional so I was able to swap them and restore spark. Incidentally, if your back plate has been stored for any length of time always change the points as they seem not to last well in storage and in my case the contact surfaces simply fell off! This older motor was fitted with Norris ignition points which seem to be rather a cunning design. I've made a separate post on how to change these.

So at last everything back together and attempts to start are in order. Sadly the motor would give nothing and then suddenly fired and revved at high speed for a second or so before dying. Throttle had no response. This smacks of a mega air leak somewhere and as I was pushing it to try and reuse the crankcase gaskets, it looks like I didn't get away with it! I will separate the cases and fit new gaskets. At the same time I am now in a position to check the truing of the crank which might be contributing to the contact with the magneto pole.