H8 Float Conundrum

I have a set of original H8 carburetors on my XK 140. I have seen various suggestions for setting the float level, including 5/8", 3/4" and one inch. The float bowls are quite large on these carbs, but the issue is getting the fuel level low enough in the float chamber that it does not flow out the jet flooding the throttle body. I found that one inch worked in one carb, but not the other (the front one if that matters). Anything less than one inch leave the fuel level in the float chamber higher than the jet. Anything beyond one inch submerges the float without changing the volume of fuel displaced, or the fuel level at the jet. The floats are the original brass in good condition.

I am not an engineer, only the son of one, but I suspect many of you are. My uneducated guess is that a fix would be to find a more buoyant float, or one that is taller. As it is, the float only sits proud of the fuel by at most 1/4 of an inch with no load on it. With the cover and fork installed I suspect it is no more than 3/16". As a result, the window of adjustment is very narrow. One inch is also an extreme adjustment for the float fork, creating significant angles which may impair its operation. A more buoyant or taller float could help correct this, but no one seems to make one. Another possibility would be to lower the entire float bowl compared to the throttle body and the jet, but I am hesitant to pull everything apart and start screwing with the original design.

The common lament is that float levels on H8s are a bitch to set, and simply a downside of owning these rare carburetors. But, I need to solve the problem since I have fuel leaking out the front jet no matter what I try. Any suggestions?

Since you have 2 nearly identical units, have you tried swapping parts to see if the problem moves?

My suspicion is the float valve in the front carburetor is damaged to or incorrect. Swap them and see what happens.

I don’t think it is the float valve. It is brand new and passes a blow test going on and off with light pressure. These carbs are iconic, but not very well designed. Despite the very large float chambers, the fuel level has to be low to keep it below the level of the jet. That means most of the chamber is unusable to store fuel. The chamber could have been lowered to increase usable volume, but wasn’t. Also, you see the value of the overflow tubes put on later SUs when fuel starts blowing out the weep holes in the top all over you and the engine.

Hi Bob, sometimes old floats leak fuel into their interiors, which makes them ride too low in the float bowl. If there is fuel inside the float then the fuel level in the bowl must be higher to get the float to close the valve, which leads to fuel too high in the jet.

The float can be checked by removing float from carb. Shake the float and listen for slosh. Compare weight of float with the other float. Compare how far the float sinks in a bowl of water. Unless all your floats are equally good or bad, you can compare them. Leaky floats are annoying but the do occur.

Other issues may include valve not closing even if float is at correct height. One cause for that is fuel pressure too high.

I pulled the float and checked it. Nothing inside and no bubbles when submerged in water. Interestingly, it floats much higher in water than in gasoline with ethanol. I have a new fuel pump with electronic innards, but have not verified the pressure. That is something to check, but I am not sure how I would pull that off. Burlen advertises a “Stay up” synthetic float with a “new adjustable height” but in several emails even they can’t explain to me how that works and I don’t want to find out by spending over $120 just to take a look at a pair.

You probably did this but put it in hot water. The heat expands the air in the float which is then forced out of any leak. Makes leak detection much easier.


The float should float higher in water than gasoline. Liquid water has a mass density of about 1 gram per cubic centimeter and gasoline is about 0.75 grams per cc. As an example, suppose you had a cylindrical float of 5 cm diameter and mass 20 grams. It would displace 20 grams of water to float, with water level approximately 1 cm above the float bottom. The same float would displace 20 grams of gasoline to float, with gasoline level approximately 1.33 cm above the float bottom.

1 Like

What type of fuel pump is it? If it’s an SU pump then you will need the ID number then go to their web-site and check the pressure as there are ‘High’ & ‘Low’ pressure versions. If it’s a non-SU then the same applies for the manufacturers web-site.

Are these the H8’s you have?


The mixture of the main jets are correctly set? It’s a trade off between float height and main jet height. The Service Manual helps for a general set up or you could find a Youtube clip?

Have a look around the SU UK web-site for floats, main jets, fuel pumps.


Scientific fact: Water is significantly denser than petroleum.

Hello Bob,

It’s obvious that the fuel level in the bowl must be pretty much level with the jet for the carb to work. Sounds like in your case, that it quite low in the bowl. The float fork (riding on top of the float) should not have to be bent way down to work right. So of course the float itself must be fairly tall. Is there any chance you have the wrong (too short) float?

Are these the original carbs? I have a Burlen “S.U. Reference Catalogue” which shows that most XK140s had H6 carbs, except some 1955 XK140C variants, which had HD8s. The P/N of the original float for these is AUC1182, which is different from that of the H6s. The AUC1182 floats are 2.5" in diameter, but I don’t see a spec for height.

I assume the floats are soldered brass. They should feel very light, and float quite high. Some floats are available in a newer black plastic material, which is supposed to be better.

You probably know about Joe Curto, back in NY, who is surely the foremost U.S. source of both info and parts. The ultimate source of parts worldwide is Burlen in the UK, but Joe is a bit more accessible. You’d be more than welcome to borrow my S.U. catalogue. I’m right here in Boise!

Bob Frisby
E-type S2 FHC and too many other old Brit cars.
Boise, Idaho

Thanks for the input everyone. I am aware that water is denser than gasoline. My only point, made somewhat poorly, is that the floats sit quite low in gasoline, even though they are in good shape. The carbs are original to the car, even though they are not the usual H6s. They are “Sandcast H8s” as shown in Tim’s post and the photo of them below. The carbs were rebuilt by Joe Curto, so I assume the brass floats are correct for the H8s. The fuel pump is a Lucas electronic unit, AUA152EP. The Burlen web site describes the pump as a “[l]arge capacity 12 Volt with minimum flow of 12.5 gallons per hour,” but does not list the pressure. The carbs are currently set to baseline. I have been trying to fine tune them, but flooding in the front carburetor keeps shutting me down. Curto set the floats to the 5/8" recommended by Burlen, but fuel just poured out. As Bob Frisby points out, gasoline, like water, will seek its own level. If the level in the float bowl is higher than the jet, fuel will leak out. If it is even slightly higher, it will pour out. I don’t believe the XK 140 ever came with HD8s. These are H8s. The HD8s (I believe D is for diaphragm) are also 2 inch carbs, but the float bowl is much smaller than the H8s and the floats are a different size. I agree with Bob that the float should sit higher to operate effectively and be more more adjustable; however, the standard setup does not seem to allow that. Perhaps the newer synthetic floats have more flotation, but they are not advertised as such.

Change the float bowl lids over and see if the fault follows the lid.

According to Burlen (SU) the AUA152EP is an LCS pump. The SU workshop manual says the LCS pumps have “Maximum output lift” of 48 inches. That is in contrast to the L pumps which have 24 inches max lift.

It takes 1.3 psi to support a 48 inch column of gasoline. 24 inches max lift occurs for 0.65 psi. The needle valve stops on the float chambers are vulnerable to pump pressure too high. If the float dimensions are inappropriate for the carb, then valve closing problems are likely as well.

1 Like

I took the bell housings etc. off and took some careful measurements. Both floats weighed 60 grams which is about 10 grams more I recall as their listed weight. I confirmed there is no fuel inside them. I switched the float bowl tops for the hell of it as Phil suggested. As mentioned before, the floats have at most 1/4 of freeboard above the fuel line to borrow some boating terms. You give up a bit of that just to ensure the needle valves are shut. That leaves very little leeway. Instead of futzing around with the bar under the forks to set the float level, I decided to watch the fuel level in the jets and set the forks to raise that level as little as possible when replacing the float bowl lids. I then put a length of rubber fuel hose into the fuel inlet and tried to blow into it while watching the jets. No change in fuel level. I then turned on the fuel pump and at least for now, no leaks. The pump does keep ticking over slowly, so I am suspicious of the fuel pressure. The LCS pumps are supposed to put out 3.8 ft/lbs of pressure which is close to, but under the 4 ft/lbs limit. These numbers are all from memory so let me know if I am off base. If the pressure is too high, is there any way to adjust it from the pump, or is an inline fuel regulator the answer?

It does sound as though the fuel pressure is too high. If you can hear the pump slowly ticking over then fuel is on the move, most likely through the needle and seat, into the fuel bowl and then into the main throat of the carburettor as you have seen.

Interestingly, the H8 has a needle & seat which is unique for it.

These needles and seats are said to perform better than the standard versions.

These have been around for a very long time and look period. You can fit the gauge for a reference then swap it for the brass blanking plug.



Just a thought and you have probably checked but are the fuel bowl vents clear and open. If they are blocked in any way the fuel bowls can pressurise and give the exact same symptom as high fuel pressure.


I am starting to focus in on the fuel pressure myself. I put in an original style Lucas fuel pump rebuilt to run electronically instead of with points. As discussed, the float level has a very narrow band and the flotation pressure of the original brass floats is not impressive. The needle valves definitely close completely, but the flotation pressure keeping them closed might be overcome if the pump pressure is too high. That would fill up the float bowl irrespective of the float level. The front carb might be the problem one because the fuel line takes a straight shot to it, whereas there is a T to the rear carb. The plan is to buy a cheap pressure sensor from a local auto parts store and see if that solves the problem. If it does, I will probably buy a Filter King as Tim suggested.

For what it is worth, I read somewhere that H8s, which were designed for racing, were given the large float bowls to make up for the weak fuel pumps of the day and their inability to deliver sufficient fuel when running at full throttle. Good thought, but I am not sure it worked. When the HD8 came out, it had a normal size float bowl. The smaller and lighter floats float higher, alleviating much of the problem we have been discussing.

What is that? Can you post a photo? An SU fuel pump was the standard fitment. I’ve not heard of a Lucas fuel pump?

The SU pump was mounted on the chassis and looks like this.



Bob, for what it’s worth here’s the receipt from 2010 when I sent my 1950s square body SU pump away to a specialist to be restored and converted to solid state. The restorer provided the test specs for the pump after the restoration. The test specs show an outlet pressure of 2.0 psi, which has proved to be perfectly adequate for my XK120 with the standard H6 carbs.

Ideas in a row here:
1.) Identify make and model of current fuel pump
2.) Identify make and model of fuel pump specified by Jaguar for your car
3.) Measure output pressure of fuel pump. If using a fuel pressure gauge, be sure to get one which works well for the 0 to 3 psi range (modern fuel systems use much higher pressures). If using a tygon tube, take a tube with at least 72 inch rise above the pump and watch to what height the fuel rises when pumping. Be careful to be able to shut the pump off immediately and how the fuel will be emptied if it rises above 60 inches (your car will only need 24 to 48 inches depending on Jaguar specification).
4.) Symptoms pump pressure is too high include no stop in the ticking when the pump is powered with engine off, and also with ticking stop with engine off then has too rich idle mixture because engine vibration jostles the needle valve with pressure too high.