PUBTALK

Each venture into the garage reveals ‘Invicta’ inscribed on the petrol tank of the little veteran – Latin for undefeated, unconquered. But is that message an encouragement for PUB or a warning from the 1914 2-speed belt driver?

Each venture into the garage reveals ‘Invicta’ inscribed on the petrol tank of the little veteran – Latin for undefeated, unconquered. But is that message an encouragement for PUB or a warning from the 1914 2-speed belt driver?

The green shoots of normality are resuming. No, not the lifting of lockdown restrictions but after over a month of hot sunny weather, it is raining outside. At least it feels like England again. Actually the very slight lifting of restrictions may permit a little riding, although possibly still not recommended for us oldies? At any rate PUB has concluded that a short test ride to check out repair work can be added to shopping trips. Such riding keeps her well away from others, and her preferred short test routes remain close enough to push home should a test not work out.

Currently, however, more bikes have been worked upon than tested – due, just as Frank said, to ‘displacement activity’, ie. working on what looks the easiest task, until it stops being the easiest, in which case move on to the next easiest. Of course, parts availability also interferes, but so far electronic orders have come remarkably quickly, save for a Big Port kickstart spring. No reflection upon the Australian supplier (who has a few other new vintage AJS parts available), but presumably international post is affected by border and flight controls. Nevertheless after over a month it seemed necessary to flag up ‘non delivery’ onto the eBay system. No doubt readers can guess what happened next

– the parcel arrived in the next morning’s post. Oh dear, but an acknowledgement and apology was promptly sent, and the eBay procedure terminated.

The spring has now been fitted – ‘fitted’ in the old sense in part because the not-quite-standard fittings on the Big Port required it be ‘adjusted’ a bit. However, that is one of the fixes yet to be tested. Activity may have been a bit ‘scatter gun’, but this month the column will continue with the Invicta gearbox issue.

With little prospect of a Pioneer Run this year, there was little excuse for procrastinating. But since when did PUB need any excuse to procrastinate – it is a core skill? However, the gearbox did need to come apart for investigation. A photocopy about the Mk 4 ‘Jardine’ 2-speed gearbox from the Previous Owner provided

a very helpful diagram of what was likely to be inside, but almost nothing about actual maintenance or dismantling information.

It is a crossover gearbox like modern Japanese items (or should one say that modern Japanese gearboxes are like the veteran Jardine?). With the bigger of its bearings being on the pulley side (offside) it seemed likely that the internals would come out that way. However, the pulley is smooth and the ‘nut’ not hexagonal. So first move was to saw and file the jaws of a redundant large spanner to closely fit the only two flats. On the other side the sprocket was plain (except for the teeth), but at least with top gear selected (which just locks input and output shafts together) a spanner could be put on each side. Given enough heaving, one or other would surely come undone.

It did, and it was the sprocket side. Using a soft drift the sprocket came free before too much force was needed, which, of course, is not good for the bearings inside. To avoid that need in future, and to

facilitate locking it, the sprocket was then drilled with plain and tapped holes. Three 3/8 BSF tapped holes were interspersed with three plain 3/8 holes, and whilst on the drill a further six ½ inch ‘lightening’ holes were drilled just to make it look pretty (it lives under a cover so no-one can see it of course). A piece of scrap angle iron was drilled to match a couple of the tapped holes enabling the sprocket to be locked.

With the sprocket refitted, top selected again and the angle iron locking it from rotating, the far side pulley locknut could be attacked more firmly. It did not budge. Looking at the diagram, and the belt pull, it seemed certain that the pulley would be left hand thread and the locknut right hand. Nothing moved, even with the application of quite a big hammer. Just in case, the opposite hand was tried briefly, without any more joy. What next?

Sometimes a really tight nut responds to being hammered tighter, before returning to the correct direction, so PUB decided to hammer a bit harder the ‘wrong way’, then turn around and continue the ‘right way’ again. However, the final ‘wrong way tightening’ freed it and revealed that the locknut was indeed left hand threaded. This is because the pulley is not threaded at all but keyed to the shaft (ie. not as the Mk 4 diagram). Nevertheless it was undone – and the diagram now has additional handwritten notes. The sprocket was undone and removed again before the gearbox was removed from the bike and onto to the bench.

The threaded bearing retainers came out easily enough, and the bearings too, but not so the gear cluster. Pushed and pulled this way and that, nothing worked, and neither did undoing the gear selector spindle – a mistake that only complicated things. At times the innards were stuck, apparently unable to come out, or to go in again. Tapping this way and that achieved getting the bits back in, but not out. Then suddenly a tap had it all fall out rather randomly, but with no indication of why or how, or, more importantly, how to put it together again!

Eventually the trick was discovered. The first pinion on the mainshaft is a tight and keyed fit on the mainshaft, but has to come free to allow the shaft out, and that then provides clearance to wriggle the rest free. All that is required is a tubular supports for the sliding dog anda gears whilst the mainshaft is tapped free (although the dogs have to be aligned correctly). Similarly, support and tapping were required for refitting (which was done a couple of times whilst investigating the end float and selector travel).

Things did not look too bad inside, in that nothing was broken, and there was only a minor chip or two. Given the amount of harsh action the gearbox has to endure with no clutch, even before bouts of dropping out of and into gear, it could have been worse. However, the state of the dogs was less encouraging. PO (previous owner who, embarrassingly, reads this column) did get in touch to confirm that yes, the gears

had recently been

replaced, or certainly at least three of them had, but not the mainshaft or sliding dog. So the sliding dogs, mainshaft dogs, and possibly the low gear pinion dogs, might all be over 100 years old – and they look it.

Ideally they would all be sent for ‘pulsed laser welding’, a new process that PUB has just read about. This clever, modern technique can apply suitable metals to hardened steel and cast iron with negligible heat and therefore no distortion or de-tempering. The article showed a pinion having a broken off tooth slowly built up, and a gearwheel so repaired. PUB wants a ‘pulsed laser welder’!

However, especially during lockdown, a somewhat simpler (and cheaper) solution has been attempted using a trusty Dremel tool. The gearwheel and mainshaft dogs were easily accessible, and could be ground back a little to square them up, and as long as the parts are hard, which they seemed to be, that should work (case hardening is more problematic according to its depth).

The double-sided sliding dog is harder to deal with because of access to the working faces. It did occur to PUB that one side (RHS in the diagram) engages with the pulley spindle splines, simply sliding along them, but never disengaging, so its dogs should remain in good condition. Turning the sliding dog around might be a good idea, although more generally parts like to remain paired up with those they have run in with, and such changes can be detrimental. That should present better dogs to the low gear pinion, and the poorer ones to the splines they cannot disengage from. Clearly she is not the first to have this idea, because the potentially good side also shows rounded dogs from a past life the other way around. Boo.

Nevertheless, they were still the better side, so after doing as much cleaning / squaring with the Dremel as manageable, it was indeed reversed for reassembly, possibly

back to original? It remains to be seen whether this is sufficient to discourage sliding out of gear. The bike is never going to have to go for many thousands of miles, so if it works that will probably do. If not, then either that new and clever welding, or some remade parts will be required.

Having learned the secret of the one keyedon tight pinion, reassembly was not difficult, and with the simple special tools it was soon refitted to the Invicta. Shopping trips were not suitable for a clutchless bike with negligible carrying capacity, so it has had just one short test ride. So far the result seems encouraging, and the gears stayed put up the steep hill leading to PUB Towers – which the little Invicta plonked up with no trouble at all in low gear. Next time she might be brave and try again without the previously fitted ‘helper spring’, which should be instructive one way or the other.

With the veteran Invicta hopefully OK, save that it may yet be returned to its original handchange configuration instead of cable and spring, thoughts turned to the veteran Triumph. This languishes in the ‘shed of despair’ because the run-and-jumps make it too difficult for its aging rider. Back in 2016/17 she prepared for this day by buying a modern reproduction engine shaft clutch in the style of the period ‘Mabon’ accessory.

When it arrived its taper did not match and it was so large and heavy that it was banished out of sight to the back of a dark and little used cupboard. When the pain in the purse had faded, and riding the Triumph got harder still, it was brought out and attacked with drill and lathe (see RC188 / 189). Thoughts have now turned back to trying it out, for which purpose it has been offered up.

It is still big, but perhaps not too big if it will work. The gearing on its overly large minimum pulley setting looks daunting, although no higher than she used to use many years ago, but can only be tested on the road. This requires the footrests to move back, and be widened on that side by two inches, so suitable little engine plates and a pegged spacer have been made up. There is still a clutch lever to arrange, preferably latching like a cyclemotor control, and then a tickover to consider. A single-speeder without clutch or neutral has never needed a tickover either.

Maybe it will be ridden again, or maybe all this is just ‘displacement activity’, to avoid thoughts of selling a valued old friend on which some amazing travels and memories were made – who knows?

Left: PO (previous owner) had found some Jardine information, which he helpfully supplied, including this diagram. It was a significant asset, but not 100% representative of the Invicta’s gearbox

PUB’s home-made kickstart spring had proved a bit inadequate, so she ordered one all the way from Australia, which appeared to come by slow-boat. It went on without too much ‘fitting’, but has yet to be proven on ride outs

Left: This large ‘King Dick’ spanner had been lying around unused for ages, so was attacked with saw and file until it fitted the pulley nut. It proved to be fairly tough steel, which is good, and the plain shank is ideal for using it like a Villiers ‘hammertight’ spanner

The pulley ‘nut’ is recessed, which makes it difficult to apply a spanner squarely – especially any kind of adjustable. So the spanner needed to be a good fit, and as it proved, suitable for shock treatment (ie. ‘hammerloose’) Three holes would possibly be adequate to provide for an extractor or locking bar, but plain or tapped? Solution: three of each, then some bigger ‘lightening’ holes interspersed between them. It would be quite pretty, if not covered by a chainguard The Invicta plain gearbox sprocket provided little opportunity for easy locking (to undo nuts etc.) With the sprocket drilled it became easy to lock it using this suitably drilled piece of angle iron

A Jardine gearbox in pieces, the internals having finally come out in a rather uncontrolled way. Hopefully reassembly will be reasonably logical The worst is over, and a reassembly method and order found. Here the insides are back inside, and just some final ‘buttoning up’ needs doing POO adviseddid thath theh gears hhadd bbeen replacedld withih new a lilittlel whilehil ago. MMore recentlyl hhe iintimatedid that h maybe it was only three out of the four? If this is the fourth, and thus an old one, then 100 years of wear would explain why the dogs are tapered not square (a newly made item should look better). The taper results in dogs trying to push apart under load. Many modern dogs are ‘undercut’ with reverse tapers to pull them together – but this would not be applicable here because of the dual duty of the meshing sliding dog

The sliding dog is original, and shows considerable rounding from years of use, and harsh gearchanges (there is no clutch). It was hoped that the other side would not show this wear, as it remains permanently engaged with the output shaft splines. But some previous owner had realised this and turned it around in the past – so now both sides are worn (and not easy to access) Having done the best to square up the internal dogs that simple home workshop tools would allow, the Jardine is here refitted to the Invicta. The poor photo angle is dictated by the lack of space in the garage to be 20% or more of the survivors. Yes, the twoA very rare sight! Two Invictas, which is reckoned is a JAP with PUB, but not so the rearmost one, which stroke in front is the 1914 Villiers model now engined model of circa 1922

With one veteran hopefully now fit to ride, attention was turned to the veteran Triumph, which is fit enough but no longer manageable by its rider. The ‘Mabon’ style reproduction clutch, even in trimmed and lightened form, would require footrests two inches wider and further back. An extra pegged spacer on the left, a longer bar (previous one shown below), and some little engine plates appeared to be the easiest way to achieve this

The ‘Mabon’ style reproduction clutch obtained in 2016/17 had been extensively trimmed and lightened, and its taper recut to match the Triumph mainshaft. Here it is undergoing a trial fit, which is looking tolerable although the large pulley diameter still looks worrying. The footrests are widened and moved back a couple of inches