I'm helping my friend get his car working correctly. We are having some carb issues. What is the process for getting the correct idle jets? I have read instructions on the redline website but I'm still having issues. The 914 is a 4cyl 2270 with dual weber 44 Idf's.

I read something about removing the main stacks then driving around. Can somebody please explain this method further.

I don't see how that would help... or even run

There are a number of variables at play other than displacement including, but not limited to: head design/porting, altitude, intake length, cams, etc.

What are your jets currently? We can probably help, but we need a baseline. Have you driven it yet? What specific problems are you having?

Just don't open the throttle very far. I wouldn't go on a public road though.

As far as the idle jets, get a wideband O2 gauge. If it's rich, use smaller jets, if it's lean, use bigger jets. Learning to tune carbs without the benefit of a WBO2 gauge is challenging - to put it lightly.

There currently is an AEM wideband installed. The car seems to drive well. The problem is the air fuel ratio numbers. It seems that the mixture is always way too rich and if I turn in the idle mixture screws then it goes to the other extreme. It seems like there is a sweet spot with the idle mixture screws that I am having a hard time finding or I simply have the wrong jet size. I think my main problem is that I am getting confused because there are too many variables with the main jets ac... That is why I thought the trick with driving with the main stacks removed may help me isolate and help me figure out just the idle jets then I can move on to the others.

Right now I have 45 idles. The mixture is too rich.

When you say it's "always too rich", what specific a/f ratio are you seeing? If it drives well, it may not be as far off as you think (in which case: keep it!)

If it's always too rich and too hard to tune, then I would suspect that you have venturis(chokes) too small for that engine. Especially if you're running 45 idles and it's still rich/hard to tune.

Look down in the top of one of the throttle butterflies with the engine off. Toward the side of the throttle body, but way down inside it will be a ring with numbers stamped on it. What is that number? This is the main venturi which has a huge impact on the rest of the carb's performance.

A/F ratios are about 12.2 idle and 10.5 wot. Vents are 36. The car really does drive quite well- lots of power.

Okay Tornik, that makes sense.

36s should be in the ballpark for that engine. Now let's move back to the simple stuff. These are questions that Kevin, Chris and McMark all taught me to ask first.

What type of fuel pump (recommended psi usage) are you running?

Is it internally regulated at no more than 3.5 psi?

If not, do you have a fuel pressure regulator? A fuel pressure gauge?

Finally, if those things all check out, have either of you opened the carbs up and checked the float settings?

I have one of those inline rotary fuel pumps from cb performance. I have a regulator and gauge and they are set at 3psi. I did set the float levels last week. I set them at 11 without the gasket.

AFR gauges are nearly useless except at WOT with a load.

This is the only time that the guys at the track even bother to look at the gauge.
Most guys with dataloggers even log the throttle position, so they know when the AFR is "valid" (when the data indicates WOT).

So using an AFR to tune at idle will be an exercise in futility.

Follow the process on the redline website for lean best idle:

http://www.redlineweber.com/html/Tech/carb..._lean_best_.htm

"If the mixture screw is more than 2 1/2 turns out then the Idle jet is too lean (too Small). When the mixture screw is less than 11/2 then the Idle jet is too rich (too large). "

What you're describing about the AFR sounds like narrow-band activity. Something is weird with your setup. You need to get that working right as the first step. Most AFR setups can be programmed to act like a narrow band sensor. Lets work on that first. What model AEM gauge do you have?

stugray - Narrowband AFR gauges are nearly useless. Wideband are very useful.

I have an innovate LC-1 which is a wideband.
The other guys at the track that have WB AFR all say that the measurement cannot be trusted unless you are at WOT.

My datalogs back that up.
The numbers are all over the place unless you are basically in a steady pull.

It's still better than not having one, but trying to tune via AFR at idle in the garage will drive you nuts.

Thank you for every ones help. I found a very interesting problem. I had a somewhat cheap fuel pressure regulator (spectre?) and a Mr. Gasket oil filled fuel pressure gauge. I had checked the gauge which showed 3 psi previously. I checked again last night and it was still 3 however when I turned the car off, it stayed at 3! I let the car sit for a while- still 3. I had a extra gauge so I swapped it and turned on the car and the pressure went up to 5.5psi. I also had an extra fuel pressure regulator so I swapped that in and set the pressure to 3 and now the gauge reads 3. GRRRRRRRRR- so I had a busted fuel pressure regulator and a stuck fuel pressure gauge.


Am I correct that an issue like this could cause my rich issues?

Yes. Which is why I suggested to start there first. It may not solve it, but you need to eliminate unknowns before you can tune it correctly.

Try that and see if it makes a difference on your readings. If it's a quality wideband, what you should see should be accurate.

Stu - think of it this way, widebands operate using the same Bosch-design O2 sensors that all modern EFI systems use to dynamically adjust mixture (of course there are other upstream sensors and calculations going on too). Yes, they are most accurate at temps above 1000 degrees, I think, but that shouldn't matter after a few minutes of running as the heater circuit for the sensor itself should be maintaining that. The old narrowband gauges were only accurate at a constant load, like WOT.

Thanks. I understand all of that.

I will try to post a log, but what I observe is that just looking at a meter that is bouncing around from 11.0 to 18.0 at idle tells you nothing.

The data logs show that the meter varys wildly at low RPM and no load, so much so that you cannot even tell what the average value is unless you are looking at a log or a graph.

The only time that I can get an instantaneous meter reading that is even close to what the log says is under a steady load.

Maybe it is the sensitivity of where in the exhaust the sensor is located.
Mine is right after the 4-1 collector.
Also exhaust leaks can screw up the measurements,and can make it vary with flow.

I know that if the probe is mounted too close to the exhaust exit, you will get atmosphere sucked back into the exhaust and skew the readings. This effect is amplified at low RPM.

Hope this doesn't get too hijacked...


But Stu, you hit on several good points. I think there are several things that complicate the issue on our motors. The firing order, exhaust type, sensor location and only 4 cylinders combine to a scenario where getting a "mean" or average reading at low RPMs is problematic with the high-refresh rate of a modern wideband sensor. Combine that with a bigger cam and reversion really screws with things. At least, that's my understanding, so maybe that's what you're experiencing.

Chris made an excellent synopsis of some of this (although for a different problem) in a thread here:

http://www.914world.com/bbs2/index.php?sho...217979&st=0