Fuel Level and Float Height

[almarconi]

I was assembling a rack of Keihin CV34 carbs today, setting the float height using a digital calipers. In order to verify my work, I checked the service fuel level using a clear tube. What I observed, has me somewhat puzzled. On two of the carbs, eventhough the float height was measured the same (all floats were set at 21mm), the fuel level was too high on one carb and too low on an other. After taking the floats out of the carbs, I used a tweezers to tweak the tabs a tiny amount. After reinstalling the floats, I checked them and fuel levels were within the service limits of 4mm.

I had always assumed that once you set the float heights using the calipers that this should give you an equal amount of fuel in all of the bowls. It seems that a tiny movement of the float tab makes quite a difference on the fuel level. Has anyone else observed this?

 

[Guest]

It's true the tabs are very sensitive. I'm struggling to get mine right now. Because levels can vary, the clear tube method is the most accurate. Good luck

Paul

 

[rustybronco]

I can't say that I have observed when you change the tab height by a small amount, it makes a great difference in fuel levels. what I have observed is, when setting float height to the factory recommended specification, the fuel levels are quite a few times, different from which is specified.

I believe some of the reasons that may be the cause of differing fuel levels, from setting them at the recommended float height measurement are, production tolerances of the spring tensions in the needles, the spring pressure in the needle weakening over time, floats not being parallel with the float pivot pin, sealing surfaces of the needle and seat assembly requiring more pressure to stop fuel flow, slightly differing float weights and a slight float saturation/deterioration over time.

there may be other causes, but I think this covers the majority of them.

because of what I have observed and shall we say being "bitten" a few times, I no longer only set float height to the recommended measurement, I check the fuel levels.
I would recommend others do the same.

 

[Nessism]

If the float needle spring is weak the weight of the float will cause the needle to compress when the carbs are up side down - such as when measuring float height. You need to hold the float with your hand in such cases and measure the float height at a point where the float tang just barely touches the needle. It's a delicate measurement but doable. It wouldn't be a bad idea to go ahead and change the needles when the springs are weak although they are expensive. Aftermarket needles are half way reasonable but there is always the concern with using non stock parts - I'm not sure what's worse; using aftermarket parts or weak float needles.

 

[bwringer]

I've been bitten before by low-quality replacement float needles that had weak springs (the spring inside the needle that supports the little nubbin in the middle). If this is weak, it will compress under the weight of the float when the carbs are upside-down, making it nearly impossible to set the float levels accurately.

If you have to use these inferior parts, you can support the carbs at an angle to lessen the weight on the springs. Some other models of bikes come from the factory with very light springs in the needles, so you have to use this technique.

Measuring actual fuel level is the most accurate way, of course. If there's a lot of variation, you might have a float that's soaking up gas. However, this is quite rare unless the floats are physically damaged.

 

[rustybronco]

on the '82 xv920j with Hitachi carbs I have been working on, that I had installed new K&l kits. the factory spec'd height of 23mm wasn't even close to getting the fuel levels correct when "properly" set (rotate carb until float just touching the needle).

in order to arrive at the recommended fuel level of 0 (+/- 1mm), I ended up with a 20.5mm float height.

cheep kits, old needles, I don't care, I set them by fuel levels when ever possible.

 

[hp1000s]

You need to use the "tab just touching the spring-loaded pin" method; also, the two round plastic float bodies aren't always exactly on the same horizontal plane, and that can throw off the settings dramatically.