WELCOME to my blog about restoring vintage overhead camshaft AJS and Velocette motorcycles

 

Apologies to anyone that has come to this page expecting to see exclusively vintage AJS motorcycles .....scroll down the page a bit and you will find plenty of them.

However, I recently sent all of my cammy AJSs to Bonhams and they were sold at the Autumn Stafford Sale. I only know where one of bikes has ended up - Sammy Miller bought the V-Twin for his musuem.

Somebody sent me the above picture with Sammy on the bike and I there's also a movie somewhere (maybe facebook? ....I don't use it so don't really know) of Sammy riding around his Grand Prix circuit at the musuem. I am very pleased that the bike has gone to one of the best motorcycle musuems in the world to join his collection of extremely rare AJSs and to be seen by visitors. 

The URL of "vintageajs" for this blog is now somewhat misleading but it's too late to change it and there is still plenty to read about the AJAY projects. If you happen to be the new owner of one of the bikes that I recently sold then there is plenty here to read about its build. I still have 2 early Big Ports waiting in the wings ....but they will have to wait until I've completed the Velos. 

In 2023 I started the restoration of 2 early Velocette KTTs plus another Mk 1 OHC cammy special - a few details about each of these bikes can be found here and here. 

A lot of work has been done on these bikes over the past 2 years and the INDEX PAGE provides links in chronological order of the project so far.

Earlier this year I acquired a DOHC 250 Velo Engine and now have a Resurrection Plan. There is an INDEX PAGE for this engine rebuild and work is progressing well. The crankshaft has been rebuilt using MOV flywheels and "K" mainshafts and I have completed work on revising the lubrication system to mirror that of the Mk V KTT engine that is in my workshop.

 

I have just completed further work on the "bottom end" to get the crankshaft to actually fit and to sort out the engagement of the crankshaft shock-absorber and the bevel gear plus I have put helicoil inserts into the threads for the cylinder barrel studs. Further details HERE.

During the last 5 years I have posted quite a lot of information and to aid navigation the "Labels" section on the right side of this page lists the various projects.

The labels marked "INDEX" give a link to a page that provides a complete list and links to all of the separate sub-projects related to that main project.

Alternatively, scroll down this page and see what's here...

When I started this blog I owned a 500cc AJS R10

that I've been riding for many years and wanted an early 350cc bike. I bought one at a Bonhams auction; this is what I brought home....

....a bit of work was needed to bring it back to life 

Full details of the restoration can be found here.

During the restoration of the K7 I figured that I could put an early overhead camshaft Velocette cylinder, cylinder head and cambox onto the crankcases of an AJS 350cc engine from 1931, convert it to chain-driven OHC and make an engine that looks like a K7 but has a Velocette top-end. I had a 1928 350cc AJS sidevalve that I had bought on eBay and used that to create the AJcette ....giving credit to both manufacturers.

It looks pretty similar to the K7 and to demonstrate that there really are 2 bikes, here they are both together.

Details of the AJcette project can be found here.

I have quite a lot of early Mk1 OHC Velocette parts and after completing the AJcette I decided to use some of these to make a replica of a one-off bike that AJS built in 1929/1930 for an attempt on the world speed record. The original is a huge V-Twin beast that started out with a naturally-aspirated engine but, having failed to gain the record, was supercharged ...and again failed. The bike ended up in Tasmania for many years and, after being repatriated to the UK and restored, it is now in the National Motorcycle Museum.

This is what the original looked like:

and this is my recreation.

 

 

Like the AJcette, the V-Twin uses Mk 1 OHC Velocette cylinder components. The full story of how this bike was built can be found in the links here.

There is also a 14 minute edited Youtube summary of how these bikes came about here and a longer unedited version here.

In January 2022 I started the restoration of a 1933 AJS Trophy Model

and this was completed in March 2023.

 

The Index Page for this project can be found here.

I also reported briefly on a couple of my other projects ....vintage OHV Nortons

 and putting a Marshall supercharger onto my 1934 MG PA

 

I hope you find something of interest.

The DOHC 250 Velo Engine: Completing the Bottom End

I mentioned at the end of a previous blog on rebuilding the crankshaft that one of the consequences of using “K” mainshafts and MOV flywheels is that the assembled inner distance between the flywheels is about 1/8” less than the standard “K” crankshaft which means that the mainshafts are 1/16” further inboard.

This ended up having a few knock-on effects; in no particular order:

-       The big-end nuts foul the bearing outers

-       The crankshaft shock-absorber meets the crankcase casting before the taper grips on the drive-side mainshaft

-       On the timing side, the woodruff key on the crankshaft barely reaches the slot to engage with the bevel gear

The first of these needed to be addressed before I could make any progress on the others because it was not possible to assemble the crankshaft into the crankcases. 

As I mentioned in previous blogs I have been using solid steel “bearings” that have the same ID/OD/t as the correct bearings

throughout the build rather than continually using the genuine items as the latter have a slight taper to ensure a good tight fit onto the mainshafts – it’s just easier this way. Modifications were first tested using the dummy bearings and then 0.040” was removed from one side of the bearing outers on both the new MC22 bearings that I had purchased

and making a similar modification to the inner side of the bearing housing in the crankcases (shown below with one of the dummy bearings)

and by using 0.015” of shims on both sides of the crankshaft allowed the crankshaft to be fitted without interference with the big-end nuts, gave a slight (couple of thou) oversize on the overall crankshaft length and positioned the connecting rod exactly in the centre.   

The second issue – non-engagement of the shock-absorber on the mainshaft taper - was quite straightforward to fix. The drive-side crankcase was put on the milling machine and material removed from the timed breather extension of the crankcase casting

until the crankshaft and shock-absorber tapers mated (plus a bit of clearance). There is also a consequence of this – the crankshaft sprocket will be further inboard and will be out of alignment with the clutch chainwheel but this can be easily fixed (at a later date – when the engine is in the frame) by making a thicker K-145 shock-absorber fibre washer.

The last “gotcha” was that the Woodruff key barely reached the bevel gear and had minimal engagement. There are 2 main factors that position and hold the bevel gear in place on the crankshaft: the key and the tightness of the LH threaded nut. As this engine has a lot of gears in the cambox (with their associated inertia etc) the drive train will need to be as strong as possible to survive.

A new key was made from an 18” length of ½” x 3/16” key steel. The picture below shows the new key together with a standard Woodruf key.

Fitted to the shaft (and with the dummy bearing in place) the new key provides the best possible level of engagement with the bevel gear.

The material is O1 tool steel and will be heat treated in due course to improve its properties (see eg here). And I have over 17” of this key steel left over!

The bevel gear is designed to be a close fit in the timing-side crankcase casting to maintain oil pressure in the lower bevel chamber. I found that the fit with this particular bevel gear and casting was a bit too tight – the bevel gear would not even fit into the hole. The crankcase was therefore back on the milling machine and the boring head used to open it up slightly.

As the changes to the lubrication system that I have made will not require a pressurized lower bevel chamber the larger diameter hole will not be an issue.

With a 0.020” gasket on the K-46/2 bevel gear housing and 2 K-111 shims (0.023" in total) behind the bevel gear

there is perfect engagement of the lower bevel gears.

Not shown in the above picture, I have made a ½” x 20 TPI LH thread nut (removed here for clarity) to replace the usual K-114 LH ball-valve nut and the crankshaft bevel has been pushed “home”.

In preparation for the next sub-project – making the eccentric studs, I have put helicoil inserts into the 4 3/8” BSF threaded holes that support the barrel. These must be put in very accurately to avoid the studs sprouting out at weird angles like the spines on a porcupine. Each crankcase half was bolted, in turn, to the large angle plate on the milling machine and the digital level box used to set the top face horizontal (with the milling machine table as the reference).

However, when I checked this by using the position of the vertical head when a 5/8” diameter ground bar first contacted the surface

I found that the level box method was not sufficiently accurate and I could get a better and consistent result by using the contact method.

For the X-Y positioning I machined one stud on the lathe with a small “pip” in the centre and used that in conjunction with a small centre in the collet chuck, under a magnifying glass, to locate the on-axis position.

Whilst this might seem crude it has proved to be accurate. At the time of writing I have completed most of the machining of the eccentric studs – and the barrel slides down over them perfectly.

For any particular helicoiling exercise I never seem to have the size I want and invariably end up buying another kit.

I have also helicoiled 2 of the 3/16” BSW threads that hold on the K-45/2 timing gear cover.

That just about concludes the work on the bottom-end for now. The oil pump drive gear alignment will be checked and modified if necessary at the final build (when the oil pump is inserted) and I will need to remove a bit of material from the back of the inner timing case to clear the piece of additional aluminium that I have added for the crankshaft oil feed quill.

The next step is to complete work on the eccentric studs so that the cylinder barrel can be put in place properly.

The DOHC 250 Velo Engine: The Lubrication System – Part 1

The DOHC cambox was designed with 2 oil feeds at the extremities which direct oil directly onto the cams and a drain pipe which would have scavenged the oil by gravity back into either the oil tank or the crankcases, all circled in the picture below.

There are essentially 2 ways in which oil can be provided under pressure to the cambox. The easier way would be to tap into the lower bevel drive chamber (for example, via the tapped hole for the drain plug) and add a pipe with a bifurcated union to provide oil feeds to both sides. However, there is a problem with that, namely that by combining 2 essentially different lubrication systems – pressurized lower bevel/upper bevel chambers (the existing design on the Mk 1 engines) together with a directed and targeted approach to the cambox it would not be straightforward to control the different amounts of oil going to each part.

On the Mk V KTT engine that is in my workshop the lubrication system was completely revised to provide oil under pressure directly from the pump to a quill in the end of the crankshaft for the big-end (see picture below)

and a jet that is directed between the inlet and exhaust cams

plus another feed that provides oil to the upper bevel box and which then drains back to the lower bevel box by gravity. This approach avoids pressurisation of the bevel gear chamber.

The castings for the timing-side crankcase and inner timing case on the Mk V KTT engine were substantially revised compared to previous cammy Velo engines to support internal oil passages and there is an oil take-off in the timing side crankcase, indicated with an arrow below, to provide oil to the upper parts of the engine.

There is also a drain from the lower bevel chamber back to the crankcase. This picture is of the engine “as received”.

My preference is to replicate this approach on the DOHC 250 engine so that key parts of the system are targeted directly. This avoids pressurisation of the internal cavities of the engine and the oil flow rate to different parts of the engine can be controlled by choosing appropriate diameters for the various holes feeding them.

The disadvantage of choosing to modify an existing “K” engine to strategic lubrication is that it is not a simple 5-minute job. As I indicated in a previous blog, a number of modifications are required:

 1)  The small passage on the end-cap of the pump that connects the feed side of the pump to the lower chamber (indicated with an arrow on the picture below)

 needs blocking off. (I have made this modification previously for the Velo oil pumps on both the AJcette and the AJS V-Twin – see here) by building up with braze and machining.

2)  The drive-side crankcase needs modifying with a passage to take the oil directly from the pump outlet to a union to connect to the “outside world” - as in the picture of KTT 581 above

3)  A quill to provide oil into the end of the crankshaft for the big-end and a modification to the timing gear cover, K45/2, for an oil feed needs to be added

4)  The crankshaft must be modified at the end of the timing-side mainshaft to accept the inserted quill

5)  Oil feeds to the cambox (2) and to the upper bevel will be needed. A drain back to the oil tank will also need to be added.

6) A drain needs to be provided to scavenge the oil from the lower bevel chamber back into the crankcase

7) Some means of regulating the flow/pressure will be required.

The first step in making these changes was to modify the K-45/2 cover with a small (5mm) hole aligned with the end of the crankshaft. I have found on previous projects that the easiest and most accurate way to do this is to insert a centre into the end of a piece of solid bar, 22mm diameter in this case,

to pass through the bearings in the assembled crankcases and with the timing gear cover screwed into place, drill a 5mm hole in the cover.

The next step was to make a boss/oil channel to feed the hole (which will support the quill) out of a piece of aluminium. I didn’t have any material of quite the right size but I did have a few lengths of 40mm diameter 7075 (thank you Christer) and used one of these to carve out an appropriately shaped piece

and with a boss on the inside plus a 6mm LH threaded piece of steel

the collection of bits could be assembled

and then laser welded in place.

The piece of 6mm diameter steel is only to aid setup to unsure that the holes are aligned (and a cable tie was also put around the whole assembly to hold it together prior to welding); it will be replaced with the quill. A 6mm LH thread is used because if there is any interference between the crankshaft and quill then the quill would tend to tighten rather than unscrew. There is a 1/8” BSP thread on the end to feed oil to the assembly.

The quill was made with a 1.3mm diameter hole and fits perfectly into the hole in the crankshaft.

The second part of this project was to cut out another chunk of metal in a similar way and attach that to the crankcase

where the boss would be needed for the oil feed to the upper part of the engine and to then carefully drill

a hole through in exactly the right place to meet the exit of the oil pump

and to tap 1/8” BSP for a union.

It’s always a good moment to see the drill emerge in the right place - THIS was another instance, because a “fix” would not be straightforward.

There was insufficient material in the crankcase casting to allow an uninterrupted hole

and so, a 6mm OD/4mm ID aluminium tube was added to transport the oil from the pump to the boss. It’s held in place with JB Weld.

This completed the work on the aluminium components.

The external pipework to supply the oil to the upper parts of the engine – the bevel drive and the cambox and also means of controlling the flow rate - will come later.

The return paths for the oil must also be considered. The cambox oil will go into the oil tank under gravity but the oil that is provided to the upper bevel and which will then flow down to the lower bevel needs to be scavenged from the lower bevel chamber.

On KTT 581 there is a passage that is cast into the timing-side crankcase which connects the lower bevel chamber to the bottom of the crankcase. Here, rather than drill more holes in the structure, I have simply modified the K-163 Drain Plug with a 3/16” OD copper pipe silver soldered inside that will take oil from inside the chamber (at the same height as the scavenge hole on KTT 581) and which then connects with a modified B-38 crankcase drain plug to return the oil to the crankcase.

The last part, for now, was to modify the pump. I indicated earlier the opening that needs to be blocked and this was done by adding braze to the gap

and then carefully machining back to a cylinder.

to prevent the oil discharging into the lower bevel chamber.

 

…..And Finally….. While I was machining one of the pieces of aluminium on the milling machine there was muffled “pop” that appeared to come from the motor. Everything seemed fine and so I kept working until a few minutes later the machine lost power and I was forced to stop. The electric motor was so hot that I couldn’t touch it and, clearly, something was seriously wrong.

The next day, when the motor had cooled down, I found that the electrical junction box on the side of the motor looked like this inside.

It seems that the capacitor had exploded. I am not an “electrics” person but Google tells me that a capacitor can explode for any number of reasons – I have no idea which applies here.

Anyway, I ordered a new motor; this turned up the next day and was easily fitted.

As you can see, it’s made by Polestar – a Chinese manufacturer of electric vehicles. I fitted the last motor (the one that blew up) about 12 years ago so we’ll see how long this one lasts.