Simplicity is good. Few moving parts, a basic electrical layout, black paint, no frills. This could be a description of the Ford model T, but it is not. It is a description of our BMW R26. A 1956 single cylinder, single carb, 6V standard motorcycle. It does not get much simpler, no matter how far you go back. The motto of the Airheads (of which we are members) is "Simple by Choice", and this machine beautifully embodies that motto. It is a beautiful machine built for a purpose, at a time when quality efficient transportation was key. It even has points for sidecar attachment, despite having just 15hp at its' disposal. If you have ever been dragged along the ground by 15 horses at a gallop, you will know that it is more than adequate power. So with such a simple and well-built machine, what tale do you have to tell ? Glad you asked.
It started with the smell of fuel n the garage. It took a while to trace it to the R26, but there was definitely a more pronounced smell around that machine. There was no visible stain or wet spot, just a lingering smell of fuel. The usual suspect on machines like this is the float bowl of the carb being faulty, and failing to shut off the fuel supply leading to a leak. The bottom of the float bowl was suspiciously moist, and the engine casing below it was suspiciously clean, so it seemed like an open and shut case. Upon examination, the float had trapped some moisture, and so a new one was sourced (ridiculously expensive for a brass float compared to plastic, but this bike is nice enough to warrant original). A new float bowl gasket was ordered as well. Once received and installed, I went for a test ride and all seemed well.
Next morning, faint smell of fuel. there was a droplet of fuel forming at the same spot on the bottom of the float bowl. At this point, I began to see if there was a route to the bottom of the float bowl coming from some other part of the carb or the fuel hoses. There was nothing obvious, although at one of the fuel hose connection points, the fabric-covered fuel hose was definitely damp from fuel. Since this motorcycle is just gravity-fed for fuel, there were no clips on any of the connection points. Despite not liking the look, hose clamps went onto every connection point. There was no other place where fuel was evident, so I took a brief test ride and checked. And then I checked again an hour later. The problem looked solved.
Next morning, faint smell of fuel. I laughed the kind of laugh that pokes fun at oneself, but which really indicates that the situation is not really funny anymore. Upon examination this time, there was no longer a droplet at the bottom of the float bowl, but there was a clean spot on the engine case right below where one of the hose clamps now lived. Well I was planning to do a carb rebuild anyway, and so I did. Then, climbing a diagnostic ladder toward the fuel tank, I encountered a moist area right at the petcock lever. Aha ! A notorious spot for problems due to the disintegration of the o-ring gasket. Not content to stop there, I also ordered the petcock gasket for the attachment to the tank. Parts arrived a week later, and took only a few minutes to install. I sat watching the petcok with the fuel turned on, and the machine off. No detectable leakage. I waited an hour and checked again. No detectable leakage. I took a test ride. No detectable leakage. I waited 2 hours and looked again. The petcock was moist with fuel.
From what I could tell, the fuel began right where the petcock threaded on to the tank. But it had a new gasket that I had just installed! I drained and removed the tank and concentrated my attention on the petcock flange. Nothing detectable. I put the petcock on it, plugged the cross connection, and threw in a little fuel. Nothing detectable. I then put the tank in its normal position, and taped some paper towel to the tank encircling the petcock flange. I let it sit. An hour later, Bingo ! The paper towel was moist with fuel. Not much, but certainly enough to form a drip over many hours. I repeated the experiment. Same result, a small fuel leak from the tank itself.
I emptied the tank and began to lightly sand the area around the flange. It was built up with solder, so someone had been here before. I could not find the specific point of the leak, but there were a few suspect areas once the paint was removed. After some days of drying, and then work with a wire wheel and dremmel tool, most of the solder was removed, revealing a hairline crack. It was clear then, that vibration was probably the key ingredient to making it leak and find its way through the solder patch job. Once cleaned up, it was properly welded, and the the paper towel test was repeated. Bone dry.
So what did we learn? A repeated lesson shared before in To Fuel or not to Fuel, and in On Getting Grounded and in To Spark or Not to Spark. Obvious solutions, and the usual suspects sometimes mask the culprit. I did not go to the tank first, because carbs and petcock are notorious for fuel leaks, and I thought I found the problem with the float bowl (which did have an issue, just not the main one). In this case, the simplicity of the machine contributed to a sense that the solution must also be simple. It was, that is once I found the root cause....