I toyed with the idea of making them floatless, as I did with the Webers on my race car: much more accurate and repeatable fuel levels, more consistent in cornering and acceleration, no worries of sunk floats, stuck inlet needles, and pressure insensitive.
I had to do the same with the H8s in my XK 140. The float level bench adjustments are just an indirect way of setting the fuel level at the jets. Looking at the jets themselves will give you a better indication of whether you are high or low.
I ended up not making any changes to float level, but this exercise gave me a chance to clean up my choke adjustment. With the choke applied, the center carb jet dropped as expected, but neither the front nor rear carbs were getting more than a bit of enrichment. Interestingly, there really isnāt a procedure in the manuals for synchronizing the chokes, but itās pretty obvious. I used a depth gauge to measure the drop in the center carb, and adjusted the linkage so that all three had the same drop. Car starts much better.
I have new UX and UE needle sets coming from the UK, thatās what Iām really waiting for.
The lean carb is still lean. Iām going to take a closer look at the carb pistons. Iām wondering if thereās a difference in weight, fit or spring strength thatās making the back piston rise higher.
In the mean time, try dropping a washers down over the piston stem to simulate a stronger spring.
I did that experiment. FWIW, 25 grams, .85 oz, will make a UM needle work well.
I recieed both UE and UX needles today. I installed the UX and set them up, but the weather was wretched for driving so no road report. I spent a good hour working on the adjustments, and managed to get them adjusted well. But thereās much more cyclic variability that I want. I wonāt know if it leans out at high loads until I can drive it.
Since I couldnāt do a test run, I used the time to set up the Flir camera, and this is what I saw:
At the moment this photo was taken, the A/F ratio was around 13.8 and the engine was idling smoothly. But look at #1. Itās a full 100F cooler than the other cylinders. So thereās my lean condition. Iām beginning to wonder if the variability and lean running are a valve train problem.
I would do a compression test first.
Mike,
If you look at the heat signature on the rear manifold it seems that it is cooler on all rear cylinders. Good use of the flir camera.
Have you tried the different needle you mentioned in an earlier post. If not you have a baseline image to go off.
The good thing about the flir image is that the temperature data is stored on the image.
Matt
This is with the UX needle, but at idle, I donāt think the needle makes any difference. The first couple of stations are identical.
Itās interesting that all the ports are cooler than the junctions. That suggests that the rich mixture results in fuel burning off in the exhaust. Iām going to try to raise the A/F to 14.5 and see what that looks like.
Itās really disturbing that the A/F mixture is unstable. I say 13.8, but really, itās bouncing between 13.2 and 14.5, and much more variability on the last carb. Iām wondering if I might need new valve stem seals, or whether the design of the manifold has something to do with this. Is there an internal balance pipe on the 4.2 manifold?
I donāt think itās disturbing, Mike: itās a carburetor, coupled to a roughly-made intake manifold.
I donāt think youāre going to find it to be any more stable than that, but itāll be interesting to watch your process
This is why I used a datalogger on my WBO2 sensors. Spot values were useless, it was trends that could be observed.
1 Like
Trends: that would be the thing.
Let me be clear: the front bank is relatively stable, fluctuating up or down by about five tenths. The rear has much wider variability.
I had some more time to tinker today, and I set up the carbs \with UX needles. The work extremely well, Iām not sure I will even bother trying the UEās. Along the roads, they keep the A/F in the high thirteens. A bit rich, but the car really likes it. I took a new thermograph after a spirited run:
So the same strange pattern appears with UX needles, and after a hot run at that. The last cylinder is just running cooler. Iāll try a compression test, but I donāt think that even a leakdown test will show a problem. If thereās a problem, I would guess itās a recessed valve, leaking stem seal or scored bucket. I have no intention of tearing into a sweet running motor at the start of the driving season, so any further diagnosis may need to wait for fall. Iām hoping itās just an artifact of the ceramic manifold coating.
Are you sure itās not just the head sucking away the heat? The junctions are always hotter than the ports. More hot gas and flames as they pass through three times as often. I canāt explain the rear one.
The one other thermal image Iāve seen had the front three hotter too but that must be a coincidence.
And the plugs will probably show nothing. If it really runs well, do a leakdown test for peace of mind. Iām sure measuring vibrations would give something away too. Maybe even with a phone app if they exist.
Michael,
It could be because of the reflectivity of the ceramic. Make sure to set the camera for a shiny surface and try looking at it from different angles.
It could even be valve shim gaps.
Matt
Good point, moving it around must help too.
\My inclination is to file this away as a āneeds more dataā project. The first thing Iāll try is to directly measure surface temps with a thermocouple. I think I have one that should work.