Fuel air ratio 1932 series 90.

[Steves Buick]

I knew the Marvel ran rich, but how rich. I slipped a fuel air ratio sensor up the tail pipe, softer than last prostate exam. I found it ran between 11.5 and 12.5, the accuracy is most likely off due the location of the sensor being so far aft. Has any one else experimented with this? I'm just curious what others have found. 

[hchris]

Looks pretty good; Stoichometric ideal is around 13:1.

[Bloo]

Stoichometric is 14.7:1 by weight. That is a good place to run for lightly loaded cruising. Few if any cars are going to do that in 1932. No doubt it was built richer. You can run even leaner than 14.7:1, and get better gas mileage, but there is danger of burning a valve or piston if you go too far. Fuel distribution is not always good in old cars, and of course the leanest cylinder is the one that burns up. In more modern things it was common to go leaner than 14.7:1 at part throttle cruise until oxygen sensors came along. Early oxygen sensors work best right at stoichometric, and cars run very well there. Under acceleration or load it needs to be quite a bit richer than 14.7:1, at the very least 12.5:1. Probably a lot richer on a car that old. Maybe 10:1 or more but that's pushing it. Idle mixture works well around 14.7:1 too, but was almost always run much richer back then. For one thing, there was vacuum wiper leakage to cover up. Fastest idle mixture will prove to be about 12.5:1.

 

If a car is old enough to *not* have an enrichment circuit or some sort of variable enrichment in the carburetor, it must run a "power mixture" of 12.5:1 or richer at all times, as if it were accelerating. The engine may not be too happy about that, but I believe it was the case on Model T's and many other cars of that era. Model A's too, but the Model A had a control that allowed the driver to lean out the mixture from the driver's seat when he wanted to, so that wasn't so bad.

 

I don't necessarily think you can trust a gauge like that to be very accurate, but I can't be sure. I'd like to compare one to the results with a 4 or 5 gas machine. Those "air/fuel gauges" got popular a few years ago, but the wideband sensors were expensive and failing a lot at that time, so I didn't buy one. I'm still curious how well they work. Decades ago there were indicators that used simple narrowband sensors. Those worked better and more repeatably than they had a right to, but there were limitations that a wideband sensor could potentially solve.

 

If you just shoved the sensor up the tailpipe, you are probably getting extra oxygen. Gas analyzers that simply have a hose shoved in a tailpipe always read a little higher than theoretical on oxygen. Some of it is due to reversion I suppose, small leaks in the exhaust probably also contribute. There is a lot of oxygen around. It gets in whether you like it or not. The closer to the back you are, the more there is. I suspect that gauge works on oxygen, and if it was just shoved in a tailpipe it probably got more oxygen than it thought it did.

 

Edited February 25 by Bloo (see edit history)

[EmTee]

Just for comparison purposes, what does the A/F gauge say when you shove it in the tailpipe of your daily driver?  Or, your '60s classic...?

[Skidplate]

I have the dual carb setup on a 37 straight 8 that also has the dual exhaust manifold. I went ahead and plumbed in a sensor just past where the duals combine back into one. Then I mounted a gauge on the inside. I think we assume it's a steady reading on the road but it really isn't. You have to kinda average things out. I would watch my O2 on the road and shoot for 11-14 then check my plugs at the house. I'm at a pretty high elevation so I'd also tweak my adjustable jets in the mix. I finally got everything pretty close but it was a process. Oh, and I also used one of those pistol grip heat laser thingies. Kinda watched my engine block heat around the spark plugs. Tweaked a little. Drove a little. Looked at the plugs. Watched the heat around the heads. Rinse. Repeat.

But I will say this... especially with the dual carbs, I believe the O2 reading "on the road" really did help.

Edited February 25 by Skidplate (see edit history)

[Brian_Heil]

Dean Tryon did an in depth air/fuel study on early Buicks. 

[Larry Schramm]

12 hours ago, Brian_Heil said:

Dean Tryon did an in depth air/fuel study on early Buicks. 

 

Dean also has a book out on carburation.

[Mark Shaw]

Fundamental Carburetion Book - Carburetor Calibration Guide

 

[Steves Buick]

Thanks for the replies, I tested my 1950 Chevy truck with a 235, to my surprise it read a solid 12 to 1. Again I suspect the location so far aft contributed to the result. The sensor was inserted only 20” into the tail pipe.. I’ll check a couple other cars for a comparison value. Thanks Mark for the link.  

[60FlatTop]

On 2/24/2024 at 4:07 AM, Bloo said:

Stoichometric is 14.7:1 by weight.

Gauge selection for specific uses is a good indicator of the desired reading. With 15 at the mid-range that would be one of three most accurate points from a petrology standpoint. On a graph gauge accuracy is high mid-range and around 25 and 75% of scale.

 

I like using a tachometer as an example. A dragster uses a 90 degree face with the shift RPM straight up. That is all you are interested in. A road race car has a wide range of 270 degrees because of the variations of driver interest.

 

I have worked on a lot of building mechanical equipment back into the 1880s where operating procedures have been lost. The gauge ranges help conjure up what should be happening.

 

Stoichiometry at 14.7 is an easy one because it is the same number as nominal atmospheric pressure. One of those "magic numbers".

 

Your newer O2 sensors are also looking for 14.7 in their own way.

[Steves Buick]

Got it ...this sensor sets off an alarm at 18.5 :1 no drama..attached is my 1996 Ford 150 cold start ..it leans out as the idle kicks down.

Ill try leaning the marvel just a few notches at a time to see if I can bring it a bit closer to 14:1 and report back. Thanks for the interest..

[EmTee]

11 hours ago, Steves Buick said:

attached is my 1996 Ford 150 cold start .

Interesting - what did it say at warm idle?

[Steves Buick]

Warm idle was 18-19.to one. Im not banking much on the accuracy of the gage as again the sensor is at the tailpipe not closer to the eng. But pretty big dif from the cold Idle. This was a rainy day project and thats about it. Cheers,

[Bloo]

Yeah, there's definitely air in that. Any 1996 Ford runs at 14.7:1 at both idle and cruise once warmed up because it runs in closed loop on a conventional oxygen sensor. Power mixture for hard acceleration will be richer, and the fuel may shut off completely when decelerating until the engine gets close to idle speed. Also, any sensor needs to go before the catalytic converter to deliver useful information.

 

[demco32]

14.7:1 is correct.

 pre war cars had a richer setup due to the poor quality of the fuel in my opinion.

there was a lot of carbon deposits on the valves.

It also makes a difference where one drives the car, at sea level or at altitude.

hot or cold weather.

In modern cars the ECU controls everything, but pre-war cars of course had no ECU.

 

Edited February 29 by demco32 (see edit history)

[Steves Buick]

Well it was an interesting experiment for a rainy Washington day, anybody need a used A/F ratio gauge?

[EmTee]

7 hours ago, Steves Buick said:

...anybody need a used A/F ratio gauge?

Well, I wouldn't totally give up on it.  I would consider installing a bung in the exhaust pipe closer to the manifold so that the gauge could be installed temporarily for diagnosis or tuning purposes.  Together with other tools (e.g., timing light, dwell meter, vacuum gauge, etc.) I think it would be useful.

[Steves Buick]

Thanks I had the same thought,  i might even drive the car with it installed, I’m searching for the perfect F/A setting on the Buick. I’m close, car run great but I have to crack a little choke if I need the power for a hill climb etc. Just fine tuning the Marvel has been a small challenge in its own right. Did you know Bill Antelli Finger lakes area?

[EmTee]

5 hours ago, Steves Buick said:

Did you know Bill Antelli Finger lakes area?

No, but I did get to meet Bill Preston ( @avantey ) at his home and see his '38 Roadmaster last July a few weeks before his passing.  We rode together in my '38 Century to the shop where his Roadmaster was being recommissioned.

 

[32buick67]

@Steves Buick what is the float elevation in your Marvel carb?

 

Something to keep in mind that was recently brought to my attention by expert @Robert Engle regarding fuels and float heights - 

 

 

Older fuels, lower octanes, more dense fuels pushed the floats up high out of the bowl fuel level, hence the 19/64" dimension for our 32 carbs being really close to the top of the bowl cover, which would properly set the fuel vs LS nozzle elevation.

 

Now with modern lighter fuels and higher octanes, the floats at 19/64" will ride too high if the scribe marks are set to the original Marvel specs.

 

I recently re-adjusted my 32 carb down to approx 32/64" from top of bowl to a dry bowl float elevation and it made a difference in running much more appropriately leaner.

 

Per folks like @Str8-8-Dave I have found good and consistent results with a coated a balsawood float in sig nitro coating https://www.hobbytown.com/sig-manufacturing-sig-sd026-super-coat-butwate-dope-fuel-proof-high-gloss-4oz.-sigsd026/p715387

With this modern method, the top of the balsa float sits at the same elevation as the modern less-dense fuel level (regardless of octane and ethanol content), and its too high vs the original cork float buoyancy height in the more dense 50-60 octane fuels of the old days.

 

My conjecture is that if a Marvel is running much richer than the original rich settings, then its likely due to a float which is too high vs the LS nozzle elevation and the float elevation needs to be adjusted down to ensure the fuel level in the bowl is at the same elevation as the LS nozzle per the Marvel specs.

[Str8-8-Dave]

9 minutes ago, 32buick67 said:

Older fuels, lower octanes

which is another subject that needs exploration to make these old low compression engines run correctly.  I currently use Sunoco recreational fuel in my car which is an ethanol free fuel rated at 90 octane.  I still run a fully functioning heat control system and it just won't tolerate all the water produced by burning fuel laced with ethanol.  But 90 octane fuel has a flame propagation rate that is just way too slow for 4 or 5 to one compression ratios our engines run at.  With the initial spark advance set at factory spec, the fuel is still burning when the exhaust valve opens which overheats the exhaust valve because the head of the valve has lost contact with the head and can't transfer the heat of combustion to the head.  The exhaust system, manifold, muffler and pipes are all running hot. You can't just advance the initial timing to fix the problem because at 50mph the centrifugal advance fires at 38degrees BTDC which was a pretty good advance for muscle cars of the 60's before EEC.  Advancing the initial advance will result in excessive full advance and starts pounding rod bearings because the combustion cycle then starts too early in the crank rotation.  In modern cars you would start hearing pinging from pre-ignition. 

 

I would like to know how to get the 90 octane ethanol free fuel down to 60-80 R+M/2 octane by adding kerosene or some other additive and experiment with lower octane and no ethanol.  I know the combustion cycle is too slow with 90 octane, you can hear the exhaust rumbling and it has fried all the high temp paint off the exhaust manifold.  Even the water temp is running around 200 degrees which means things are too hot.      

[Steves Buick]

Beautiful pair of cars there Em Tee, well done. Str8 Dave hit the nail on the head I believe with the slow burn rate of 90 oct fuel. I have also had the same issues with after burn off the exhaust valve with 87 Oct ethanol free fuel, I advanced my initial timing 3 degrees and it did seem to run better but occasionally would back fire on shut down. My water temp never got above 145 until the engine is loaded going up a long hill and my high temp paint is still there. I’ll start experimenting with bringing down the Oct rating to something less than 80. Not sure just how but what could possibly go wrong? Right. I’ll send a shot of my secondary temp set up that shows water temp before and after it enters the engine.

[Lee H]

10 hours ago, Str8-8-Dave said:

I would like to know how to get the 90 octane ethanol free fuel down to 60-80 R+M/2 octane by adding kerosene or some other additive and experiment with lower octane and no ethanol

Pretty easy. Diesel fuel has an octane rating of ≈ 20. Not sure that blending is a linear correlation, but it’s a place to start.

 

Personally, having widely experimented with my 4:1 Model A, I would recommend increasing the ignition lead. I found I could easily run in excess of 60* total advance at cruise. The slower combustion speed of modern gas also means that that you don’t get the very high peak cylinder pressures near TDC (the kind that causes pinging, and wrecks bearings) that you would with old 70 octane gas. You can run a surprising amount of advance with low compression.

[Steves Buick]

Thanks for that, Looking fwd to the challenge. Dean Tryon has a very well done chapter in his book that I may replicate in my car for fun. To me it seems like it may be easier to just raise the compression ratio than blend the fuel. 60 degrees of total advance seems like a lot, what rpm were you turning? Did you get the 60 degrees by going past it until the engine pings under load then retarding it to eventually find 60?