Model A - investigating the rear brake drum

 In common with several Ariels of this vintage, the Model A has a rear wheel sprocket which is integral with the brake drum. This in itself would not be a problem were it not for the fact that the sprocket teeth are worn out.

The whole assembly is far too big to go in my lathe so what can I do about removing the sprocket teeth? Some options:

• Can I spin the hub in the rotary table, mounted on the workmate and remove the teeth with a fixed angle grinder?
• Can I build a jig to carry the whole hub in the milling table, and remove the teeth with an end mill?
• Get Draganfly to do it?
• Find a local enthusiast with a bigger lathe?
• Find a better hub?

One thing I have discovered - a BSA A50 chain ring has the required 47 teeth to suit the 5/8" x 3/8" chain I need. An Ariel 4305-32 looks similar.

We'll have to think about that a bit more. In the meantime, we'll have a look at all the other bits in that area so that we can work out what we are dealing with especially since I'm expecting to place an order with the Ariel Klub Slovakia shortly. 

The brake plate is all present and correct.

The brake plate dust exclusion ring is a bit mangled - I have a feeling you can get these new as they are spot welded in place but since the 1930 Model A doesn't have a plated ring here I may repair it and paint it.

In the brake drum we have the 'new for 1929' ring that retains the spoke heads should they break off. Ariel must have had some incidents with brakes seizing or wheels grinding to a halt through loss of spoke heads. Not a pleasant experience to contemplate! It's all very greasy in there.

Equally greasy is the back of the brake plate but at least we have some shoes which are whole and a set of springs and a cam.

The cam bush is very worn indeed and will have to be replaced.

It's fitted through the brake plate and there is a nice little gas weld on the inside:

Here's the cam. I don't think there's any doubt that we will be buying new ones!

Something else that has caused much discussion on the Facebook group is what brake lever this machine should have. This part is definitely Ariel, but it may not be 1930 - however it's not much later, perhaps 32. Checking the parts lists doesn't reveal very much from any of these years - they all show a relatively short stubby brake lever. The various members that contributed to this discussion revealed quite a variety of brake levers including this long curved one - so go figure.

It did have a lot of bends in it. I've taken one out and this will probably sit better with respect to the mudguard stays - we'll have a look when it goes back together.

Moving on to the frame, we'll take the opportunity to clean out the threads for the chain adjusters

There's something odd here - this looks like a braze repair around the sidecar mount. The mounting should exist as a more or less complete ring without that lump of material between it and the frame lug.

The brake plate torque lug is present on the inside and both of the top hat washers are present.

Moving to the other side of the bike, this all looks good and the stand stops are in very good shape.

Missing from this side is the little flattened tube that's used to protect the wheel spindle thread from the chain adjuster. They are often mangled - we will have to make one of those. 

At both ends of the spindle, the wheel nuts are not in the best shape. The left hand side won't run all the way down the spindle - it looks like it's had a hammering at some time.

Removing the barrel

Next stop, remove the cylinder block. Nuts are stuck fast, so we need a dose of Plus-Gas:

This is the 3/8" flange barrel used by Ariel throughout Huntmaster production, and by BSA until the 1/2" barrel flange appeared for the 1955 A10 Road Rocket. Poking around waiting for the Plus-Gas to do its job, take a look down the pushrod tunnels:

That's a leaf, possibly hawthorn... How did that get in there?

Nuts undone, we need to break the red Hermetite sealing the base gasket. This won't take much force and I'm not overly worried about the pistons anyway, so we'll do it like this. Two bolts in head bolt holes, either side of one piston; a suitable bar between them (if this had been remotely reluctant to move I would have made a stiff bar out of a bit of angle or something) and a lump of wood on top of a piston.

A quick tweak on the crankshaft has the block pop up - but it's not all plain sailing. The timing side piston sticks at the bottom of the bore, in the rusty unworn section and takes a bit of persistence to remove it. When it comes off, we find these rather worn cam followers:

A look at the flywheel reveals this is the earlier, smaller, built up crank used until 1958 - perhaps not surprising in that this engine is marked 'LF' and is clearly a 1955 design. The crank has no radial bolts in the flywheel like the later 'large bearing' cranks.

Ariel adopted the large bearing crank in 1958, when BSA rationalised it across all A10 models; the large bearing crank initially appeared in the 1955 Road Rocket.

Here are a couple of general views. The engine has been fairly damp inside for a while:

We'll have a closer look at the camshaft later. I've seen worse.

However... have a look at this video - it's as though someone forgot to put the shells in!

Rule of Nine

Many moons ago, I had a burning interest in classic cars, many with British Motor Corporation 'A' Series engines. One of the useful tips I learned working on these was the Rule of Nine, a way to easily make sure your engine was in the correct position when adjusting valve clearances.

The thing is, when you adjust the clearance in a particular valve train you need to make sure the cam follower or tappet is on the base circle of the cam, or the clearance will be excessive. You do this by checking the clearance on both valves in a cylinder when that cylinder is at TDC. The thing is, wild valve timing or quietening ramps in the cam profile means you have to be accurate when positioning the engine.

But there is an easier way. Have a look at this camshaft:

That's a spare camshaft for my Square Four - a lot less worn than the one in the engine. On the right hand side of the picture, you see the thread retaining the cam chain sprocket, and that first cam is the exhaust cam for number one cylinder - notice it is pointing straight up and the valve will be fully open.

Here's a Stormy Petrel's eye view of the rockers, so you can see what I am talking about:

Move to the left hand end of the camshaft and we are looking at the last or eighth cam, the cam for the exhaust valve on number three. It's pointing straight down, which means the valve is fully closed and the tappet is on the base circle.

Now look at this picture. The cam that is pointing up is the inlet cam for number two cylinder - it's the fourth cam on the shaft.

The one pointing down is the fifth cam on the shaft, the inlet cam for number four cylinder, so

4 + 5 = 9

Remember the first example where we looked at the first cam and the eighth cam? Well:

1 + 8 = 9

Guess what? It works on the other valves too! So, if the valve on the second cam is fully open (which is very easy to discern) then you know that the valve on cam seven (the exhaust valve for number four) is on the base circle and can be adjusted.

Easy. The Rule of Nine.

By the way, I found that after 1000 miles, most of the clearances needed adjusting - maybe because I didn't do it very well last time.

Finish off with replacement of the rocker box gaskets with Threebond. The fibre gaskets specified by Ariel are not suitable for the sorts of temperatures the cylinder head sees.

SQ4 Front Brake

As part of winter maintenance, I have my front wheel off for new bearings, a new tube and a look at the brake. It appears that I still haven't got it right.

Take a look at this, it is the leading end of the trailing shoe:

To me, this is indicating that I haven't got the adjuster tight enough - there is about 20-30% contact on both shoes, at the cam end - in both shoes, there is no contact at the adjuster end. Here is the leading shoe, trailing end:

In both cases you can see where the lining is worn by contact with the drum. The brake is nothing like as effective as the one on my W/NG, which is obviously a much lighter bike but the brake has much more feel. They are bedding in at the cam end (trailing end of the trailing shoe here):

I'm happy to see that the drum is nice and clean, with no trace of grease

Less happy to see the heat damage to the paint, but we will deal with that when the weather warms up.

Setting the drum up with emery tape and double-sided sticky tape in the traditional way results in much better contact. By the way, don't do this with one of the wheel bearings missing since the spindle, and the brake plate and shoes will not be concentric with the drum:

Set the brake plate up such that it is pulled down to the bearing and is held rigid (as far as possible with these pressed steel brake plates) - that way the shoes will be as cylindrical as possible. I've used a short tube spanner on the spindle to take the place of the fork leg, and I am holding the brake plate from turning with a tommy bar. I turn the wheel to sand the linings.

Start with adjuster backed off and gradually wind it in, turning the wheel a few times and then dismantle again for inspection. When your satisfied that the adjuster end of the shoes is ground down enough to fit the drum, put a bit of pressure on the brake cam lever and repeat. Stop when you have contact over most or all of the shoe.

Here's the leading end of trailing shoe:

Leading end of leading shoe

 Lots of dust after the process is complete:

There's always something

Having got the magneto back on, and the bike up and running properly, I went for a little ride... and it was great.

Back in the workshop, it was time to tidy up; while cleaning the oil draining trays out, I found this little chunk of cast steel:

Hmm...

Casting around for help on the AOMCC Forum, Adrie de Graff, from Holland, sent me this marked up picture of my own timing chest and suggested it was a bit of a cam follower.

This was backed up by the ever knowledgeable Nev Hunter, from Australia. I needed to investigate.

The cranked tin funnel came in handy again:

And with it all apart again, we could pull out the followers and have a proper look. This is the exhaust follower:

You can see in this picture that it is very thin around the cup:

It's a brittle fracture, with heavily discoloured material in the break. It's so thin, the case hardening would have gone right through the material. It's been like that for years.

So, Mr. John Budgen to the rescue:

I will be using the two on the left.

Collecting bits

What do we have here? A camshaft, in much better shape than the one in my SQ4:

Valve Timing

So, we had another nice little run out yesterday, on a shopping errand this time. No dramas, Amelia running nicely with her new brake shoes. She attracted some attention from a cyclist who came over exclaiming: "Oh, a Squariel! I haven't seen one of those in years!" He proceeded to tell me about his Triumph TSS and T120R, which apparently was fitted with a Morgo pump as well.

You meet the nicest people on an Ariel.

I'd heard mention of the effect valve timing has on overheating, and it occurred to me that in the joyous response to the problems I had had setting up the valve timing initially I had possibly not been too careful to get it spot on, being happy to go with the factory timing marks. Ariel folk lore tells us this is not wise.

So, when I arrived back from shopping I whipped off the offside rocker cover, the #1 plug and the front crankshaft cap and had a look - and it looked like number one inlet valve was opening near TDC - it is supposed to be 3/16" before TDC.

So, late in the night, and with the head torch on I took off the timing cover. I discovered that if you remove the dynamo sprocket you can reset the valve timing without removing the camshaft sprocket (and hence you can leave the oil pump in place too) After 4 or five attempts I finally got the timing as close as I could - which I think is spot-on. The cam chain tension makes a big difference, and you have to test it with tension on the cam chain.

I may need to change the cam chain at some point - a worn chain affects the valve timing.

So that done, I reset the ignition timing, a job which is very easy in the dark, because you can see when the points open if you turn the ignition on, since there is always a little spark when the points open. However, testing it all again the next day revealed the valve timing was indeed spot-on, but the ignition timing was way to far advanced, as evidenced by several kick-backs and spitting at the carburettor. Realising what these symptoms meant, I reset the points gap and the ignition timing again.

And after all that, as usual, she starts easily; but probably more easily than ever; she runs eagerly (though I could only manage a couple of miles) and she idles very much more smoothly.

We have some errands tomorrow; we will see what happens on a longer run

It's been a while...

About a month actually. Ariel time in the last 4 weeks has been spent on a number of fronts: 

• welding cable guides inside the rear mudguard 
• finishing the wiring 
• messing about with the valve timing 

the latter item has been the most frustrating. I have a good set of guidance notes from Draganfly, written by Bruce Longman which I have followed religiously.

I assembled the head using only cylinder number 1, as I thought I would make it easy on myself and avoid having to turn the cam against four sets of valve springs.

Initially I tried the Waller-style method of aligning timing marks.

Having done this I turned the engine over to see the exhaust valve opening half way down the power stroke. Bruce says that sometimes the factory timing marks don't give the correct timing, so I tried again... and again... and again... eventually getting the timing discs and dial test indicator out to do the job the long way, all the while getting more and more proficient at releasing tension on the cam chain, removing the nuts, drawing off the crankshaft pinion before I could remove the cam shaft pinion... Always the same result.

Eventually I started doing this more scientifically, drawing the valve event charts and calculating overlaps & durations. This is what it is supposed to look like:

Frustrated with Cylinder #1, and confused by the fact that #1 runs backwards, I changed to #3 and drew another diagram:



This revealed that #3 was about 180 degrees out; while #1 inlet valve was correct and #1 exhaust valve was about 180 degrees out too. Confused, I thought about looking at #2, which was doubly frustrating because one of the pushrods would not go in the tunnel.

I set and reset valve timings for about a week, and then one day, looking at the rockers, pushrods and convincing myself that #1 inlet must somehow be on the wrong cam, I started looking at the #2 inlet pushrod problem when I happened to glance down the #1 inlet pushrod tunnel - to see the pushrod going off at a weird angle.

I now know it is possible to get a SQ4 pushrod in it's neighbour's tappet. #1 inlet was in #2 inlet's tappet, so it was operating from #2 inlet's cam! Now that I realise what I am looking at, it's obvious.

Ho hum. we live and learn. At least I found it before I pulled the head off again!