Britbike forum Forums British motorcycles in general Bsa gold star forum 1954 BB Gold Star

I’m tempted to ask the question… did you straighten the rod, or realign the crank?… but I might come off sounding like a bit of a [***], so I won’t. It would have been an attempt at humour, that likely would have failed given what John just lobbed into the discussion.
Hard to know what to do not knowing the history of the crank. You’ve got this far…. the rod needs a bushing.. you now have the ability to do that in the bike, so I would do that much. Think about it some more and go from there. You could button it back together and see how it behaves?

You wrote previously “ It came to a stop like good brakes operating against an engine not having the h.p. to overcome the braking.”

That doesn’t sound like it should have upset the crank. If it did, then it doesn’t say much for the pin/ flywheel fit. In fact… if it was that bad, wouldn’t you have known by now? I’m trying to be helpful…because if it was me, the thought of pulling the motor would have me looking around the shop for a piece of rope.

How about a come-along?
You'd have to be serious.

I couldn't agree more, because......

The PO wasn't a very good mechanic.

.005" runout on the double bearing drive side is not very good.

Continued running like that will waller* out the housings even worse than they may be now.

* Etymological question -- "waller" a hole
2014-01-29 18:34:12 UTC
The sons of the family mechanic talked about "wallering a hole", in the
sense of "to wear and enlarge in an uncontrolled manner" ("the bolt was
loose and wallered out the hole, now it don't fit"), or (with contempt
for shade-tree mechanics) "to intentionally enlarge a hole in an
uncontrolled manner" ("that <deleted> just wallered out those holes
instead of using a drill bit of the proper size, now nuthin' fits right").

For one awful moment I thought you were being serious, but now I'm looking forward to the crankcases being split.

Ignoring chaterlea25's suggestion for now, I decided to deal with the head next in order to reduce the number of BB34.GS and BB34 Alloy Clipper parts covering every available surface. This requires cleaning up the face of the exhaust valve to remove welded aluminum bits, making and installing new guides, and cutting the seats.

MM preparing to re-re-face the DeckChecker® in honor of the International Day of Light.

Well… have you resumed yet? Here I am polluting your thread with jigs again. Not sure what to make of this… it looks skookum, but I haven’t tried it out yet. The main round thingamajigger is counterbored to support the collets… it also doubles for a holder if the armature needs to go in the lathe, at least I think it does. The pinch bolts aren’t used during the pulling operation. One end piece gets pulled off and then the armature is flipped around to pull the other end. One collet to go around the wire thingy that pokes out of the coil. Just waiting for someone to tell me my effort is all for naught and they just hold the shaft between their teeth and pull it apart.

4A8982E2-3AEB-4765-878D-CDD9B42669FF by First Last, on Flickr

A389A8BD-07CE-417C-98AC-C0BA70847019 by First Last, on Flickr

Looks impressive. Now what does it do...... exactly?

Oh, I see, it removes the end caps.

So I guess there's no more need for tapping with a small hammer and hardwood blocks.

Myself, I think I would be tempted to use just the collar as the fulcrum to gently pry the end cap to where it is just free of the laminated center. But then, that's just me.

It looks great. You should be able to separate armatures with it even if the sections are welded together.

I would say that 9 times out of 10, I could/would just use it as a fulcrum, but I won’t because I only see about 3 condenser replacements in my future. I’ve had it on good authority that once in a while the end caps don’t want to move easily, hence the complexity. I could lever on the collet to spread the load on the end cap. Using it 3 times doesn’t justify the time spent, but there is a method to my madness….. at least I think there is. I’m assuming the fulcrum can be used to hold the armature in a lathe if anything needs to be cleaned up.

We returned last night from a week-long trip to New York for my youngest granddaughter's 2nd birthday, so now I'll be back at work on the BB. We planned that trip several months ago but, unfortunately, I didn't pay attention to the calendar when I signed up for the Distinguished Gentlemen's Ride this past weekend. I'll try to make amends by riding the route in the next few days.

Because of a catastrophic equipment failure that I discovered… I mean, equipment purchase failure, I don't have the right mandrel in order to use my Honall® to hone the holes in my head … I mean, my BB's head for the valve guides. While I could use my full-size hone for this, for which I do have the right mandrel, it would require holding the head at the correct angle while stroking it back and forth, which is a bit awkward. Anyway, the correct mandrel for the Honall® won't be here until at least Friday, so I reorganized my to-do list

I made a list of the tasks, in approximately the order I intend to do them, that I need to accomplish in order to put a fully-functional BB back on the shelf, so to speak, and return to working on the Vincent:

Loosely bolt the Alloy Clipper engine back together using the bad cylinder from the BB
Rebuild the STD.T and swap it for the present STD
Machine the sealing face of the exhaust valve
Hone the guide holes in the head
Make new guides
Install the guides
Cut the seats
Remove the engine, check the bearings and crankshaft, and rebuild crank if necessary
Make new small end bush and install it in the rod
Put the engine back together using the Alloy Clipper's barrel and piston
Install engine in the frame
Adjust the clutch and time the magneto
Replace the fuel tank and foot pegs
Put tools away and collect any leftover parts (valve keeper, gudgeon/wrist pin circlip, …)

After making the above list I realized AI might be able to organize these tasks better than a mere human could. So, I fed the above list to Google's Bard AI, but it just gave the list back to me in the same order, along with a few comments (like perhaps I should "take it to a qualified mechanic" to do the work for me). I then tried ChatGPT, and it changed the order of a few of the tasks (and made a slight change to the wording):

Remove the engine, check the bearings, and inspect the crankshaft.
Rebuild the crankshaft if necessary.
Loosely bolt the Alloy Clipper engine back together using the bad cylinder from the BB.
Machine the sealing face of the exhaust valve.
Honing the guide holes in the head.
Make new guides.
Install the guides.
Cut the seats.
Make a new small end bush and install it in the rod.
Put the engine back together using the Alloy Clipper's barrel and piston.
Install the engine in the frame.
Adjust the clutch and time the magneto.
Replace the fuel tank and foot pegs.
Put tools away and collect any leftover parts (valve keeper, gudgeon/wrist pin circlip, etc.).

My only comments about ChatGPT's suggestion are, I need to get the Alloy Clipper engine bolted together and moved in order to free up space on the workbench to do the other tasks. Also, if I'm going to rebuild the BB's crankshaft, it would be better to install the new small end bush at that time, rather than later, since it would allow me to accurately align it with the axis of the spindles.

At my peril I decided to ignore ChatGPT, and the Alloy Clipper's engine is back together again and sitting on the floor. So, the next task is to rebuild the STD.T gearbox.

In for a penny…

My P-190 Honall® kit came with a P-180 driver plus four mandrels covering valve guides with IDs from 5/16" through 7/16". These are extended-length 'BL' mandrels, whereas most of the ones I have in this size range for my main Sunnen are 'K' (e.g. a 5/16" BL mandrel has a 2¼"-long stone whereas the stone for a K mandrel in this size is 1¼"). Although BL mandrels work on the main Sunnen, Ks don't work on the Honall®.

Anyway, having discovered I didn't have the necessary BL mandrel for the ~⅝" OD of a valve guide, I decided to be ready for the future by also ordering the additional ones needed to have them in 1/16" steps from ¼" through ¾". It turns out I already had the necessary ¾" mandrel, which is why I stopped there, but the fact that I could have ordered them through 1", but didn't, shows I'm fully capable of moderation where tooling is concerned (for now…).

The "manual" I've written to describe rebuilding a gearbox is 89-pages, not counting the Appendices. However, Bard and ChatGPT are way better, since both of them describe it in just 8 steps:

Question: How do you rebuild a BSA Gold Star's gearbox?

ChatGPT's answer: (Bard's is essentially the same, although using fewer words)

1. Remove the gearbox from the motorcycle: Start by disconnecting any necessary components such as the clutch cable, kickstart lever, and gear linkage. Carefully detach the gearbox from the engine, ensuring you have proper support to handle its weight.

2. Disassemble the gearbox: Carefully remove the gearbox covers, including the primary and secondary covers, to gain access to the internal components. Take note of the arrangement and orientation of the gears, shafts, and other parts to aid reassembly.

3. Inspect and clean the components: Thoroughly inspect all the gearbox components for signs of wear, damage, or excessive play. Check the gears, bearings, shafts, and selector forks for any issues. Clean all the parts using a suitable solvent to remove any dirt, debris, or old lubricants.

4. Replace worn or damaged parts: If any components are worn out or damaged beyond repair, it's important to replace them with new ones. This may include gears, bearings, seals, gaskets, or other small parts. Ensure you use genuine or high-quality aftermarket parts for optimal performance.

5. Reassemble the gearbox: Following the manufacturer's instructions or a service manual, carefully reassemble the gearbox components in the correct order. Apply lubrication to the gears and bearings as you go. Pay attention to the proper alignment and engagement of gears and ensure all parts are tightened to the specified torque.

6. Install the gearbox back into the motorcycle: Once the gearbox is fully assembled, reattach it to the engine, making sure it aligns properly with the primary drive and clutch. Reconnect any cables, linkages, or levers that were disconnected during removal.

7. Adjust and test the gearbox: After installation, check the clutch adjustment and ensure smooth gear shifting by testing the gearbox while the motorcycle is stationary. Adjust the clutch cable and any relevant mechanisms as needed to achieve smooth operation.

8. Replace gearbox oil: Finally, fill the gearbox with the recommended oil to the appropriate level. Refer to the manufacturer's specifications or the service manual for the correct type and quantity of oil.

The above makes me wonder if Haynes has been using ChatGPT to write its workshop manuals starting at least twenty years ago.

Having given AI's recommendations all due consideration, it was time to start on the gearbox. As the first photograph shows, BSA's labeling can cause confusion for the unwary, since there's an STD2 and an STDT.



While a vise can be used, the gearbox stand I made some years ago is more convenient since it provides a sturdy support while also allowing me to easily reposition the gearbox whenever needed to provide better access.



The vertical supports allow the gearbox to be positioned anywhere in a 270° arc, or mounted on the opposing end for another (overlapping) 270°.

Remnants of RTV sealant betrayed the fact it had been apart before, as did loose Fasteners and a missing part from the kick starter ratchet. Also, I found a piece of broken plastic as well as that a chunk of aluminum is missing.



As can be seen, the case managed to collect sawdust during the time the previous owner had it apart.



I have other cases, but the needle and main bearing are both in good shape, so I'm going to see if I can repair it. The final composite shows the broken area before and after I started preparing it for welding.



It has the correct needle-bearing layshaft, but I still need to count teeth to make sure all the gears are the correct ones.

I counted teeth, and it's an STD.T. For what it's worth, a superficial inspection didn't reveal any differences between the three gears with different part numbers on the STD.T and those on an STD.

Turning to the broken case, I tilted it in the stand to have the area for the weld parallel to the outside of the case.



I preheated the case with a rosebud tip on my oxyacetylene torch and use the tank with 50/50 He/Ar for the TIG, although I don't think either was necessary. After filling in the broken area with 4043 I removed most of the excess with a die grinder and several files.



I originally planned to use a sheet of Cu (to which Al doesn't adhere) to keep the fill from extending beyond the gasket face, but decided not to bother once I saw how well things were going with the welder. After using the die grinder and files to profile the lump of weld, I fixtured the case to the mill and machined away the weld that extended beyond the gasket face.



I then reattached the three good studs, placed the middle case over them, and used a transfer punch to mark the location of the hole.



After using a genuine BSW tap, rather than cheating with UNC, the case is now as good-ish as new.



I didn't keep track of time, but all of the above probably took about two hours. Spending the time to fix, rather than replace, is my preferred way to work.

Unfortunately, the needle roller bearing is starting to shed rollers, so I'll have to replace it.



To keep smoke to a minimum, I'll need to de-grease the case prior to sticking it in the oven to remove the bearings. When I removed the lid from the de-greaser pail I discovered a gearbox outer cover that had been in the solution for long enough that I don't remember when I put it there, or why.



While the STD and STD.T have the same ratios, the following chart shows that three of the gears (A, B and H) have different part numbers.



The lesson from this is, the correct number of teeth does not a correct gearbox make. Elsewhere on Britbike, probably in my Spitfire thread in the 'projects' forum, I showed where on a gear I measure the width.



The following chart shows that A and B differ by enough in thickness to distinguish between them.

_____________ width
_____ Bush-type ____ T-type
Gear __ STD2 _____ SCT2
A ____ 0.68" _______ 0.70"
B ____ 0.70" _______ 0.78"
C ____ 0.88" _______ 0.85"
D ____ 0.59" _______ 0.60"
E ____ 0.68" _______ 0.86"
F ____ 0.94" _______ 0.93"
G ____ 0.86" _______ 0.86"
H ____ 0.54" _______ 0.54"

According to the above chart, as well as in reality, Gear H has 27 teeth and appears identical for the two types of gearboxes. Turning to AI to help distinguish between them, Bard says they have 25 and 24 teeth, respectively. Yet again, AI has very authoritatively provided incorrect information (known in the AI trade as hallucination). As long as Skynet continues to get all its input data from the internet it will remain ignorant, albeit completely confident, unable to rebuild a BSA gearbox, let alone take over the world.

Closely inspecting the gears has revealed the gearbox to be a plumber's nightmare (the former owner was a plumber) of mismatched components. So more work awaits me before I'll be ready to start assembling it.

The Dunning Kruger effect hits the big time.

(Note that current efforts seek to disprove it.)

Years ago, after observing a postdoc of mine do something, a colleague commented that "The only thing worse than an idiot, is an idiot with initiative." That pithy observation has stuck with me, having observed it in operation a number of times since then. So, I need to modify what I said about Skynet. While it hasn't developed the intelligence to take over the world, that doesn't mean it hasn't developed the self-confidence to try.

Field Marshall Von Moltke's observation is timeless.

The Dunning Kruger effect. [/quote]


Look it up and there’s a photo of me searching for my magnetic flux reversal.

" No plan of operations extends with any certainty beyond the first contact with the main hostile force."
That certainly applies to my work on the BB, although nearly all the additional work thus far has been self-inflicted.

Gears A and B in a T-type gearbox are wider than in a bush type, and for this reason they have different part numbers (but the same number of teeth) on the STD and STD.T. However, the B gear on the layshaft that was inside the STD.T, on the right in the next photograph, is the narrower bush-type, as can seen when comparing it with a T-type gear B, on the left.



Since B is a press-on gear, someone who didn't understand what they were doing installed the wrong type of gear on a T-type layshaft. I don't know if the previous owner of the STD.T gearbox, who also built the BB, was the one who cobbled together those incorrect parts, but he did rebuild the gearbox on the BB that has the shifting problem. Given everything, I now see in a new light a line from his build notebook about the now-problematic gearbox on the bike "took it apart and it's 'a good one'." Umm, no it's not, and neither was the STD.T from his horde. Or the solid Cu head gasket bodge, or the oil breather bodge, or the loose valve guide, or the clutch springs that were adjusted twice as tight as they needed to be, and let's not even mention the wiring bodge … But, I digress.

Gear A also is wider, as can be seen on the right in the next photograph, in this case with one having the earlier design that was in the RRT2 and SCT2 when they were first introduced, and on the left is the narrower bush-type that was in my STD.T.



The later T-type gear for the RRT2 is like the one on the left (although wider), designed to mesh with the dogs on gear C every 1/12 of a revolution, whereas the earlier type only meshes every 1/6 of a revolution. Presumably the design was changed to allow quicker operation of an RRT2 gearbox on the track.

The next composite shows different views of two mainshafts, with the lower mainshaft the one originally in my gearbox.



As can be seen, the lower one has the narrower A gear from a bush-type gearbox, although of the later style to mate with the dogs on its gear C. The upper one has a wider T-type A gear of the earlier 6-dog design along with the appropriate gear C to mate with it.

As can be seen by comparing with the chart in a previous post, the gears presently on the upper mainshaft would have the ratios of an ASC.T if a matching set of gears were on the layshaft. If I simply swapped shafts, the larger pressed-on gear G of the shaft that came in my gearbox would give the mainshaft the ratios of an SC.T, avoiding the ultra-low 1st of the ASC.T. Again, assuming the correct matching set were on the layshaft.

Gear A is the constraint in the system, with 26T needed for an STD.T and 25T for an SC.T. Aside from two gearboxes that are together but missing the identifying cover, a thorough search turned up four 25T A gears but no 26T, so an SC.T it shall be. However, I prefer the smaller jump an SC.T has between 3rd and 4th for the riding I do, so this isn't a problem.

As can be seen from the next photograph, and comparing with the chart in the earlier post, the gears that were on the layshaft when I removed it from the gearbox aren't appropriate for a gearbox of any type.



Even if I could have found a 26T T-type gear A for the mainshaft, and although changing the 18T for a 17T on the layshaft would result in a layshaft for an STD.T, it still wouldn't mate with the gears on the mainshaft that is presently configured to be correct for an ASC.T gearbox.

Anyway, once I knew it would be an SC.T, I located the required 24T gear D and 21T gear F, so now I have complete layshaft and mainshaft with proper T-type gears.



I have two covers stamped with SC.T (and one with ASC.T) so, before I'm finished, I'll have to decide whether to use the STD.T cover and have it look authentic but confound future generations as they shift through its gears, or use an SC.T cover even though it's age-inappropriate for a 1954 Gold Star.

I glossed over it earlier, but I'll also have to decide between SC.T and ASC.T gearing. The latter has the same upper three ratios as an SC.T, but an ultra-low first. Pressed-on mainshaft gear G determines 1st so, since I have both types of mainshafts, at this point it's the trivial matter of swapping to choose between the two.

As I wrote here or in my Catalina thread, after our trip across Texas in 2017 I changed the Catalina's primary components and drive sprocket to increase its overall ratio to that of a Clubman, which is more appropriate for the roads of the West. I also installed an ASC.T gearbox, giving it a first gear appropriate for any slow off-road riding I might want to do, but the other three gears for streets and highways. Since the BB already has Clubman overall gearing, maybe it should have an ASC.T as well. Or maybe not. Decisions, decisions…