I grew tired of trying to guess the correct fuel settings on my triple SU E-Type. So I installed a fast O2 sensor in each downpiipe, and an Innovate lambda gauge in the dash. I discovered several interesting things, but the one I want to discuss today is that my rear carb has a tendency to run lean when I push the engine. I’ve sprayed the rear carb and manifold with carb cleaner in an attempt to identify air leaks, but there’s nothing obvious. In discussion on the E_Type list, another member reported the identical issue, and eventually solved it by using a UO needle, just in the rear carb.
I’m wondering if this has something to do with the vacuum takeoff for the brake booster. It’s the ony thing I can think of that’s different from the center and front carbs. Anyone else run into this problem?
I have to say, Mike: I never got that anal about the mixture on the carburetors: I’f get them so the plugs all read correctly and it ran well, and left it at that.
I think he’s using UE in the front two. I’ve tried UM and UO down the line with the same result. This isn’t a needle or tuning problem: the rear carb is consistently inconsistent. It leans out under load, the others don’t.
I don’t run the car on the road without vacuum assist. My next step will be to plumb in my vacuum gauge and see if there’s a dip in vacuum when this happens. The purpose of my question here was to see if anyone had seen a similar problem when running in on a dyno.
Vacuum is never a issue, vacuum air is only used when braking, for the rest the vacuum circuit is closed by a valve, and will not draw any vacuum at all ( when there is no leak )
Almost the same as I did 15 years ago, though I use a datalogger and real-time viewing on a laptop. My conclusion at the time was that carbs really only work just well enough, much like the distributor, and that the XK-engine is hugely forgiving of the shambolig state of timing and mixture. Incidentally, my triple carbs were set as well as I could after many years of experience, using the usual measures, and the tailpipe emissions were spot on. The O2 sensors revealed the centre carb was ok, the front massively rich and the rear correspondingly lean.
True statement: at their best–and SUs are not there–carbs are a pretty imprecise bit of equipment. The most accurate were the lat 80s Hondas, and they were crrrrrrazy complex.
They were extremely difficult to rebuild. I tried… and failed.
Was that a 4.2? A third data point indicating the back carb runs lean. Still not proven, but getting there. I could understand if the center carb behaved differently, due to the offset head. But the front and rear cylinder pairs on a 4.2 should be symmetric. I almost feel like reinstalling the equal length header pipes to see if it’s caused by the exhaust manifold.
Is there an internal balance pipe in a 4.2 intake manifold? I don’t have one off the car to examine.
Yes, the brake system is only open when braking , and I have no evidence of leakage. But I can’t think of anything else unique about the last carb. If this isn’t caused by the vacuum system or leaky manifold gasket, then it has to be caused by an error in manifold design.
Yes, a 4.2. But I’d not read much into that, as it could just be that the carb is/was set lean to start with, or that some mystical factor made that carb open differently.
My car is also running a tube manifold, hence how I was able to weld on the WBO2 bungs. It’s also running a Rob Beere modified head and lairy cams, so it’s not an accurate example of original spec. It does have a small claim to fame in being the first 6-cyl Jag to run EDIS and Megajolt
I’m not sure, but I assume, you’re talking about my setup with the mismatched needles. Want to clarify that my 4.2 has a 420 manifold and saloon carbs on it (it’s a Mk2), meaning its 2x HD8 and the rear carb has the float rear of the carb, while the front carb float is forward of the carb. My theory is that under heavy acceleration the fuel drains away from the carb on the rear since it tilts slightly downhill plus the force of acceleration. I ever so slightly raised the float to compensate but a bigger needle solved it. If you have triple carbs and all the floats are ahead of the carb I can’t comment on why you might be seeing the same thing!
I disagree with the comment about the booster only using vacuum while braking. If the diaphragm deal is cracked/broken it will leak air. Best to plug the hose going to the booster and see if the vacuum is affected. Use a vacuum gauge to test.
I had the issue with lack of power on the rolling road test ladt week (see under Saloons),
Lambda was good andeven in the tail pipes and the engine very stable and smooth.
Back home I took 3 plugs out after cooling down. Middle carb gave a very sooty plug. The front carb a very good looking carb with sand colour all over. And the rear carb in between, but on the sooty side
Today the car ran beautifully. I have 2 other 420Gs with the same ZF 4 speed box, and they seem slightky livlier from sliw speeds.
Mike, I found with my SU floats set accurately the spec of 7/16"…I had vastly different fuel levels in the float chambers. I picked a mid fuel height in the bowl(s), went on and off a few times with the banjos and covers until fuel levels were equal. For what that’s all worth!
The inlet needle arm and float system on my carbs was “notchy” in action. SUs make a Stromberg seem like a precision instrument.
My plan for tomorrow is to readjust the float levels by removing the bells and sighting the level in the jets, then adjusting for consistency.
I need to find a way to test the 'fuel sloshing" theory. Thinking about it, the fuel level in the jet shouldn’t change by very much if at all, since the jet passage is centered. If fuel “sloshes”, the level above the orifice is going to be the mean level in the bowl. It might be sloppy in practice, but the theory seems sound.
