For you quadrajet fans / geeks. Way down deep in the cornflakes.

It’s pretty easy to bend the primary rods and/or power piston when reassembling a qjet airhorn gasket. I’ve gone through my rods and tossed my tweaked ones. That leaves the power piston as a possible culprit of unbalanced metering. On the bench, if you have a few (or a bunch) of power pistons you can observe that the primary rod holes in the power piston are sometimes different heights - leading to different mixtures side to side.

For a while now I’ve been watching my dual 02 sensors register different lambda values side to side when at cruise. What's bugging me is if the primary rods are the same depth out of the the jet, the lambda values should be the same. Same for AFR if you play that way.

This morning I took a few minutes to get a reference height from my most recent carb redone by SMI in 2017 using a digital caliper. The reference height SMI used was 1.3350” from the bottom of the cylinder to the top of the arm. Lots of my power piston arms were much taller or bent or otherwise off kilter. So I just went through a bunch & set all the rod arm tops to 1.3350”.

Here's where I was taking the measurement:

Use the depth gauge feature on your caliper to measure the height of your qjet primary rod arms - get 'em equal for consistent side to side fuel metering.

Here's what that depth looks like on the gauge.


Height from bottom of 170 Series qjet power piston to top of rod hanging arms.

This height measurement applies to later (75 & up) qjets only, not the earlier (70-series) qjets. The early power pistons are much, much taller. The problem of uneven rod arm heights applies there to the early ones too, though. I just don't have any early ones around anymore.

I’m hoping that by lowering the arms to this height will get my cars running more consistently side to side, and more correctly. If the primary rods sit were too high, that means the rods were too far out of the jets to be metering properly. This makes sense since a bunch of my qjets are running a bit richer than they should.

This is as yet untested. If anyone has a better reference measurement or better way to do this I’m all eyes. I’m assuming that pro-rebuilder SMI knows exactly how high to set these arms.

Final note, when I set these things, I polish the pistons outside just to be sure they move smoothly within the main casting, at least for a while since free movement is critical to good metering.

Hope this is of interest to some of you and perhaps helps some of you.

Chris

 

That's some good information about the power pistons height.

I'm not making the connection between the left/right offset on the primary rods and the O2 sensor left/right differences, though. For that to happen, each carb primary would have to feed one side of the engine, and that is not the case. You can easily see in the picture below the left side primary feeds cylinders #1 and #4. Am I just missing something here?

 

Kenneth,
You’re not missing a thing.

You’re quite right about each side of the carb feeding both sides of the manifold, happy to stand corrected. I have Edelbrock Performer manifolds and they work the same way as factory, but with different contours.

Where I’m going with this is to get the carbs just as good as they can be, then if I have roughness or a misfire, that’s one less (very complex) problem-source eliminated.

But truthfully it’s also about getting the lambda/AFR ratios correct across idle, cruise and WOT. I’m close, but it’s a fun challenge to see how close to ideal values/ratios you can get. Cuz the closer you get, the better cars seem to run.

Chris

 

This made me think of a few other variables that could effect relative mixture. The piston height is definitely a good thing to check.

Have the jets been checked for accuracy or at least consistency?

Are manifold runners that close to mirror images? My opinion is no, the first thing the dual plane manifold does is different things at the plenum. Not really knowing or having one in front of me I can’t evaluate, I don’t profess to know, just putting it out there for discussion.

​​​​​​…

 

Insightful post, but I question the connection between metering rod height and side-to-side lambda variation. If you have a dual plane manifold, each metering rod feeds two cylinders on each side of the motor, so each of your dual O2 sensors is seeing both primary barrels. If you have a single plane manifold, one can argue that there might be a bias, but I'd counter that the pulses in the intake will mix the flow and likely mask any side-to-side variation.

This would only be an issue if you had one of the goofy Offy single plane intakes with the divider down the middle, and anyone running one of those has bigger problems.

 

Mountain top view...Keep sight of the fact you have an analog engine with a crude but effective gas and air mixer. Variation in combustion and A/F delivery per cylinder can/will affect each side and is expected.

Variation contributions: plugs wires cap ring seal valve seal et al. Nothing is managed with a computer nor designed with max IC efficiency (by modern standard).

I'd expect an analog carbureted ICE to vary either side of stoich per bank(slightly), but within a tight limit as long as each cylinder is healthy.
Your O2s are taking an averaged of the mix point of the downstream exhaust.

Lamens... you're looking too deep, one side will vary from the other by nature of the beast as long as all components are within spec it's an average. Now add in vacuum and exhaust leaks ahead of the O2s...variable goes up.

Read the plugs & tailpipes...old skool style.

 

Gang,
Delighted to generate some debate. This is how we learn.

1st up. Yeah, it’s not fair to measure vacuum controlled carbs with a digital 02 sensor. The precision O2 sensors can measure is well beyond what a 1977 carb was designed to deliver. Where a microprocessor could back off on the mixture while going downhill or coasting to a stop, carburetor springs & vacuum take more time to react & can run richer. So, sometimes it’s rich, sometimes it’s lean for a few seconds. Just look a little less often…

Springs & vacuum take a few seconds to react to throttle and road changes. I expect it to be a bit slower than a microprocessor. But cheaper & simpler too! Or as a Doc friend of mine called it, “Peak Analog”.

I just unfocus on second-to-second variations reported by the O2 sensor and ask myself over a 30-60 minute drive, is it behaving reasonably & staying close to lambda 1.00 most of the time?

If yes, well, o.k. I’m close to optimal tune. If it’s too rich at WOT, well, o.k lean out the 2ndary rods or hanger. If it’s too rich at part throttle cruise, crank down the Adjustable Part Throttle (APT) screw to lean out the primary rods. If too rich at idle, turn them idle screws in… When I have time, the fun is in seeing if I can make idle, part throttle cruise, and WOT just a little closer to perfect. The O2 sensors are a major help in that department.

2nd up (Joe) - Manifold-dependent, both sides of the carb clearly feed both cylinder banks. So my bank-specific 02 sensors are reading both sides of the carb. Not feeling defensive, but in my mind this raises the importance of having the carb be balanced side-to-side. Balanced choke (easy), balanced idle screws, balanced primary rod arms, balanced 2ndary rods. So that whatever mechanical cylinder variances exist, at least they’re not influenced by carb. Or not much.

Anyway, I test drove my 98 convertible today with the rod hangers set precisely at 1.0335 and saw more consistent bank-to-bank lambdas (aka AFR’s). I consider that progress.

What strikes me after all these years messing with these devices, is that tiny differences on the inch scale can make enormous gains in engine smoothness and performance.

No big deal, but I also switched from a 160°F thermostat recently to 180°F. Gassing up today I learned that my around-town mileage went from an anecdotal 10.4 mpg to 11.6 mpg. Happy about that at >$6.00/gallon. My mechanic said if I went with a higher heat T-stat, mileage would go up. Seems to be true. \

Still learning.

Cheers
Chris