FETTLING A FOUR
Bill Bowers’ Ariel wasn’t broken but he decided to fix it anyway… and inevitably it immediately stopped working. What was stopping it starting? The things he hadn’t fixed fixed, or the things he had?
Bill Bowers’ Ariel wasn’t broken but he decided to fix it anyway… and inevitably it immediately stopped working. What was stopping it starting? The things he hadn’t fixed, or the things he had?
My favourite uncle had an Ariel Square Four in the early 1950s. He used to sit me on the tank, driving round Gloucestershire at high speed. So when a 1955 model came up for auction in 1992, I had to have it. Since then it has been well used, and with my wife as pillion we have had seven continental trips and covered many thousands of miles. Over the years the engine has been rebuilt and many minor things have been improved to suit me e – including indicators and a 2ls front brake for safety, a 12V alternator for better lighting and to power heated grips, and a satnav for navigation (especially useful in foreign cities) ).
A while ago someone called the wiring a mess, and that has niggled ever since, although it all worked. I planned to ride to th he Ariel Italian Rally in May 2020; immediately before that we were to organise the Ariel Owners’ MCC Long Distance Run in Cornwall, and near the end of March a local run had been arranged. So at the beginning of February I ignored the old maxim of ‘if it ain’t broke, don’t fix it’, and moved the Square into the workshop.
The wiring did indeed look like a rat’s nest, especially with all the bits added. The polarity was reversed because the negative earth alternator had replaced the old positive earth dynamo. A restored Garrard GP sidecar (found as a total wreck in a field) is attached occasionally to the bike along with its associated electrics. The ammeter had to be changed to cope with the extra output of the alternator.
Off came the seat, tank, headlamp cowling, tool boxes and mudguards, etc, and the old wiring was quickly delivered to the bin. Then I looked at the bits I had removed. It was very apparent that the paint was way past its best and of subtly different shades on different parts of the machine. As the bike was semidismantled, the obvious thing was to get it resprayed. I took the bits along to the paint
shop, and was firmly told that they needed shell blasting. That this and the painting would take ten days or so. Fine, I thought, as long as they are back by the end of February. I needed to use the same paint for the fork shrouds and lower fork legs and give them enough time to dry hard. Meanwhile I replaced the fork oil seals, and took off the carburettor with the inlet tract in order to torque down the head and check the tappets.
Then I got down to the wiring. The first thing to do was draw out a new wiring diagram, which I did on an A3 sheet of paper for clarity. It was based on the original Ariel diagram of 1955, but looks more complex with the additions which include two relays, a flasher unit, and a fuse box.
I ordered different coloured wires, a crimping tool, loads of good quality bullet connectors of different sizes to cope with various thickness wiring, from AutoElectricSupplies.co.uk. I read recently that crimped connectors are better than soldered ones because they do not heat-harden t the wire and t therefore do not s suffer fatigue breaks from v vibration… not t that a Square Four vibrates, of course.
Laying all the wires out along the frame was relatively easy. I made sure they were too long to start with, and then progressively cut them down as it all came together – this is why you need lots of spare bullets. I wondered where to put the fuse box and relays, and eventually bought a commercial domestic junction box which I bolted behind the battery on the old coil mounts. The coil was moved to the top of the extra tool box I have on the right hand side of the bike.
Inevitably, come the end of February, the painted parts weren’t ready, so I ordered a spray can of matching paint. Delivery was attempted while we were away so it went back to the seller, which meant I had no paint for the forks. Fortunately the spray shop used two-part paint and, providing you use it quickly, this can satisfactorily be applied by
hand. So I did this after picking up the shiny parts from the paint shop.
By now there were just ten days to go before the first run, but we had guests coming to stay for most of a week and
I had to make the most of every scarce opportunity to put everything back together. That is where the main problems started. The back mudguard wouldn’t fit because the new fuse / relay box was in the way. It had to be moved forward by a couple of centimetres. Then I connected the battery and nothing worked. The meter showed only 0.6V. A connection had come undone in the headlamp… still only 0.6V. Of course a fuse had blown, but I never did sort out why it showed that 0.6 Volts. At least the lights and horn now worked.
However… now the engine wouldn’t fire. I nearly ruptured myself kicking it over, and then realised that the ammeter wasn’t flickering at all as the points opened and shut. Fuse blown again, probably from when I had been playing around with the ammeter. Still nothing.
Then I noticed that a spade terminal on the ignition switch had come loose at the rivet. It was irreparable, which meant fitting a new switch. I tried again and for a brief moment the engine roared into life… then rapidly cut out. Nothing I could do had any effect. I changed the coil. In spite of using the choke, the plugs seemed dry so I checked the float level. I dismantled and rebuilt the SU carburettor, which had been working perfectly before the rewiring. I cut out a new leather washer for the carb sliding jet and checked the choke mechanism again. The carb is a pain to remove and attach on the Square Four because the nuts securing it to the inlet tract are very difficult to reach when the tract is bolted to the head.
No improvement. Eventually I poured some neat fuel into the carburettor mouth and the engine fired briefly, racing so fast I thought it might blow up. I just had the chance to notice that there was no charge on the ammeter, and then silence fell again. By now I was completely fed up, and went into the house to console myself.
I came back a couple of hours later and happened to notice that the coil was hot in spite of the ignition being turned off. Could it be the coil after all? I changed it for one that I knew worked off an old car. The points were checked. No luck. I thought again. Sometimes circuits can do odd things with a faulty earth, as my friend Pete reminded me. Check, check, and check again.
When I had moved the fuse / relay box
I had failed to replace the earth, which explained why the generator had not charged – it is activated through a relay which needs an earth to function. After that the coil remained cool, but the bike still refused to start. I rebuilt the carburettor again, doubly checking that the choke was working. I felt like putting a torch to it.
At this point I was somewhat relieved when the lockdown came. The March ride was cancelled, which saved me the embarrassment of not being able to join in. I was fairly desperate by now, and my friend Mark offered to lend me his Square 4 carburettor and inlet manifold while he sorted out the cylinder head on his own machine. My carburettor had to come off separately to the inlet tract as the studs for
the tract are seized in the head. So I started the dismantling, and suddenly realised that the nuts securing the inlet tract to the manifold in the head casting were barely finger-tight.
Mark’s carburettor wasn’t needed after all. Huzzah! Problem solved at last. Nuts tightened, petrol on, one kick, and she started!
In retrospect, I realised I had left the bike immediately after torqueing down the head. I had obviously just put the manifold back on, and failed tighten it. The resulting air leak was enough to cause all the problems. Simple really but it took me a week to sort. Thanks to all my Cornish friends who made all sorts of suggestions, mostly helpful, but occasionally frivolous!