SU Float adjustment: float hitting side of bowl?

[Stanley]

I noticed that once the float is adjusted to near the factory setting, which makes the bowl about half full, any further adjustment will make it overflow. Make a very small change of the tang on the float, and there's a small change in fuel level. Make another small change there's a large change in fuel level and the bowl overflows. Using a sight glass to check fuel level.

Maybe the float is hitting the side of the bowl so it can't go higher and close the valve. I straightened up the float so it wouldn't do that, but it hasn't helped. I though of sanding the float, but that might make it soak up gas.

I thought of a way to see if it's hitting, I'll try it today. But I wonder if others have noticed this or have some insight.

[Captain Obvious]

I wonder if others have noticed this or have some insight.

I have. Never figured out exactly what was going on, but I finally got them "close enough" and stopped messing with them.

Blue and I were talking about it a while ago... http://www.classiczcars.com/forums/su-carb-discussions/46778-fuel-bowl-level-bending-tab-not-working.html

Edited April 24, 2014 by Captain Obvious

[zKars]

I had one similiar problem with one SU. The float level would not stay where I put it. In playing with it in my hand, I found there seemed to be friction between the ball on the needle valve and the float tang, like the angles were funny. Move the float by hand, and it would jam, but intermittently.

In the end, I replaced the needle valve. It worked smoothly after that and I was able to adjust it normally. Worth a shot.

[Stanley]

Interesting link. And I ordered a carb tune-up kit just to get the valve, although the valve seems to work OK.

The floats not hitting the bowl, at least not now. Did a little experiment to see how far the float was moving. Ran some dental floss through the vent, around the tang and back up through the vent, so I could pull the float up with the lid on. No gas in the bowl. Got the red straw from a can of carb cleaner and put it in the vent so it sat on the float. Marked the straw with the float all the way down, then pulled the float up with the floss until I felt resistance (the spring in the valve) and marked it there. Then pulled it all the way up and marked it again. Checked with micrometer. About 8mm from bottom mark to middle mark, and 12mm bottom mark to top mark. So looks like about 6 to 8mm of possible adjustment, at least on the front carb (3 screws with longer ears on front carb) with the washers I have under the valve. FWIW.

Took a drive, think one or both bowls still going dry (or dry enough to lean out the mix) at top end when I keep it floored. Never a problem if I ease it up to 5000 rpm. All my adjustments didn't help any. Had to put it back together to get groceries, though. Try again tomorrow.

[Stanley]

Got it adjusted today, no problem, gremlins had a day off. Talked to the mechanic, he said there are some things, other than setting the tang to low, that will cause overflowing after a float adjustment: float hitting side of the bowl, valve sticking open or not sealing, and float sliding off the pin.

[Stanley]

Concerned about running the bowls dry at WOT, and on advice from members, I've been trying to raise the fuel levels without much success until yesterday. I've heard the 19mm down from bottom of lid/top of bowl, so that's what I was going for. More precisely, for 3-screws, 18.5mm for the rear and 21mm for the front, the 2mm difference due to slope of the engine, fudged a little because my car is raked slightly higher in front, to result in equal fuel height at both nozzles.

Last week, finally got the rear one set, but the highest I could get the front one without overflowing was 24mm down. Yesterday decided the float was hitting the side of the bowl and sanded the back edge, where the vertical part meets the sloped part, with 400 grit sandpaper. Just slightly. It worked, it's at 20.5mm down now, measured with sight glass.

Maybe there was a defect in the float. Hope sanding it doesn't damage the fuel resistance of the float.

Noticed that the fuel level goes up and down a little while the engine's running, so the decimal place in the numbers doesn't mean much. The fuel level goes up slightly when the engine is shut down, wonder why?

[240260280z]

Noticed that the fuel level goes up and down a little while the engine's running, so the decimal place in the numbers doesn't mean much. The fuel level goes up slightly when the engine is shut down, wonder why?

The fuel flows through the rail and there is a slight venturi-like scavenging force that draws up on the branches to the bowls. This unloads when the engine is shut off and the fuel stops flowing.

For the fuel level, the tilt of the car is not important, it is only that the fuel level is the same height from the top of the bridge for both carbs. The needle and jet fall in place to match.

[Stanley]

The fuel flows through the rail and there is a slight venturi-like scavenging force that draws up on the branches to the bowls. This unloads when the engine is shut off and the fuel stops flowing.

Makes sense. No return might increase the effect.

For the fuel level, what you said is correct. If you measure the engine angle, measure the distance between the bowl outlet and the center of the nozzle, and do the trig, you see that with equal levels in the front and rear bowls, the fuel is 2mm higher in front nozzle compared to the rear nozzle. For the '72 three-screws, the front bowl was changed to have the ears where the float axle attaches 2mm longer.

But if you set the bowls by level in the nozzles you can forget all that. If you have sight glass with flexible clear tubing, you can just set one bowl, move the sight tube over in front of the nozzle, make a pencil mark on the carb at the fuel level, then mark the other carb at the same distance down. Then set the other bowl, holding the sight glass next to it's nozzle to verify correct fuel height. That's better than marking lines on the bowls at different heights, which is less direct and would introduce errors. Then, like you said, rake of the car doesn't matter. The car should be sitting on level ground, though.

Edited June 4, 2014 by Stanley