[DeTomaso] Optimal Air Fuel Ratio Question

[Daniel C Jones]

>   If memory serves me correctly the optimal ratio is 14.7 to 1.

That's the stoichiometric ratio, the "chemically ideal" ratio
where there is no excess fuel or oxygen left after combustion.
Leaner means there's excess oxygen left after combustion.
Richer means there's excess fuel left. Generally, you want to
run rich of stoichiometric at wide open throttle (WOT) for best
power and leaner at cruise for better fuel economy.

> If he can be reached, I would run this question by Dan Jones.

See the attached post I made about tuning my Triumph TR8.
Look for the section on "A/F  Characteristics".

Dan Jones

> Also, I bet you aren't lean under 2500 rpm.

If it's very lean, it should surge at constant throttle.

> My 1980 8 has an Edelbrock 500 carb with K and N filter. The car runs like
> a dream over 2200 rpm but is running lean at lower rpm, no "get up and go".

I've tuned a number of carbs with a wide band O2 sensor and seat of the
pants can be misleading.  I've had cars that drove like a dream that
were very rich and others that were dangerously lean.

> Can anybody recommend a combination of jets ,springs, and metering rods
> that might get me back on track?

I went through a tuning exercise on my TR8 that the previous owner fitted
with an Edelbrock 500 CFM carb.  My tuning notes follow but first make
sure the float level is set correctly.  The float level sets the pressure
the metering passages see so it's important to get the level right before
tuning.  Also, your engine likely differs from mine in other details
(camshaft, intake manifold, exhaust), so what works for me may not be ideal
for you.  If you don't already have one, you can also download the tuning
guide from Edelbrock's website but be aware that not all combinations are
on the tuning charts.  I purchased the tuning kit which has a selection of
rods, jets and springs but to fine tune the carb, I needed to purchase
additional parts for the combinations that are not on the Edelbrock tuning
charts.  Anyway, here are my notes.

Dan Jones

I recently went through a tuning exercise on my Triumph TR8 using a wide
band O2 sensor.  I highly recommend using a wide band O2 sensor to check
your mixture.  Besides making the process of tuning a carb go much faster,
it's found problems I didn't know I had.  When I started tuning the Rover
3.5L in the TR8, it was rich in 1st, 2nd and 3rd but would go lean at the
top of 4th gear and in 5th.  I figured there were 3 possible causes:

 1. Fuel pump capacity insufficient
 2. Carb inlet needle and seat too small
 3. Fuel filter too restrictive

Turns out the fuel filter I had (one of those metal mesh filters in a glass
tube) was too restrictive.  Holding it up to the light, I could see through
the mesh so it didn't appear clogged but the wideband knew better.  By the
time I hit fourth, the fuel bowls had drained enough to cause the mixture
to go dangerously lean.  A larger diameter clear plastic fuel filter
(Purolator F21111) fixed the problem.  Since the Purolator filter has a
plastic body, I moved it to the firewall to keep it away from direct engine
heat.  The fuel filter originally on the car was maybe an inch diameter with
a glass body, a metal mesh filter element and 3/8" inlet and outlet.

I kept a detailed log of my tuning.  Here's a summary:

The engine is a stock low compression (8.1:1) Rover 3.5L V8 engine with
the following modifications:

 Edelbrock triangular foam air cleaner (needs to be replaced with a real filter)
 Edelbrock 1404 500 CFM four barrel carb (Carter AFB clone)
 Offenhauser/JWR Dual Port intake manifold
 Crane hydraulic flat tappet cam (unknown specs)
 New but stock rate TR8 valve springs
 Rhoads fast bleed rate lifters (not a smart move with non-adjustable rocker
 arms, soon to be replaced by standard lifters)
 Rimmers tri-y headers and dual exhaust (H-pipe, no catalytic converters)
 Mallory Unilite distributor with vacuum advance
 MSD 6AL spark box
 MSD Blaster 3 coil
 NGK BPR5EY11 spark plugs gapped at 0.040"

I installed the Mallory, MSD and plugs.  Everything else was installed by
the previous owner and, if I were buiding the engine from scratch, I'd have
picked different parts.

The Edelbrock 1404 is a clone of the Carter AFB. The easy stuff to change or
adjust includes:

 idle stop
 idle mixture screws
 primary side jets
 primary side rods (controls primary cruise and primary power mode mixtures)
 secondary jets
 step-up spring (controls when the transition occurs between cruise and power
 mode mixtures)
 accelerator pump lever hole
 float drop

Instead of a power valve for enrichment like a Holley, the Carter/Edelbrock
AFB's use a rod and jet arrangement.  A 2 step tapered rod, controlled by a
vacuum-referenced, spring-loaded piston, moves up and down in the jet to
provide a two different area main metering orifices.  This allows the carb
to adjust the air-to-fuel ratio for differing loads, as sensed by the vacuum
level.  When the vehicle is cruising, the rod is on the lean step.  Under
high load, low vacuum, conditions the rod moves to the rich step.  The
metering rods and step-up springs can be changed without removing the top
cover but the cover has to come off to R&R the jets.  There's a series of
screws that hold the top cover on but you must first remove two small clips
that hold the accelerator pump and throttle lever linkages in place.  These
are very tiny and easy to drop, so keep a spare pair on hand and cover the
carb when removing them.  I used a very small pair of needle nose to remove
the accelerator pump clip but found a dental pick with curved hook worked
better on the throttle linkage.  You should also remove the rod and pistons
(uses torx headed screws to hold the covers on).  Unlike a Holley, there's
no need to drain the float bowls when changing jets.  Use a screwdriver with
a tip that is the same width as the jet, otherwise it's easy to scrape a
little aluminum of the carb body.

The goal of my tuning was to get it lean of stoichiometric at cruise for
best fuel economy while in the range for maximum power at wide open
throttle.  14.7:1 is the stoichiometric ratio.  It's the "chemically ideal"
where there is no excess fuel or oxygen left after combustion.  Leaner means
there's excess oxygen left after combustion.  Richer means there's excess fuel
left. Generally, you want to run rich of stoichiometric at wide open throttle
(WOT) and a bit lean at cruise.  There's no single ideal ratio that applies
to all engines.  Some engines make best power at 13:1, others closer to
12.5:1.  Note that the air fuel ratio is by weight.  13:1 means 13 pounds
of air are mixed with 1 pound of fuel.  The usual target values for normally
aspirated 4 stroke engines are about 12.5 to 13 for WOT, 14.0-15.5 at
part-throttle cruise and 13.5-14.0 for part throttle acceleration (or
climbing a long hill, pulling a load, etc.).  If you want to lean out the
mixture at cruise for best fuel economy, be aware that you'll also need to
adjust timing.  Combustion gets much slower under lean conditions and if
you don't adjust spark timing, the combustion occurs much later and exhaust
temperature climbs.  That's bad for the exhaust seats and valves.  However,
if you adjust for MBT spark at each A/F ratio, exhaust temperature will
actually decrease relative to stoichimetric (rich will still be somewhat
cooler).  For typical gasoline engines, the range or ratios is:

  A/F  Characteristics
 Ratio
    5  Rich burn limit. Combustion is weak and/or erratic.
  6-9  Extremely rich. Black smoke and low power.
10-11  Very rich. Some supercharged engines run in this range at full power as
       a means of controlling detonation.
12-13  Rich.  Best power A/F for normally aspirated WOT.
14-15  Chemically ideal.  At 14.6 the A/F is at the theoretical ideal ratio
       with no excess fuel or oxygen after combustion. Good A/F target for part
       throttle cruise and light to moderate acceleration.
16-17  Lean.  Best fuel economy A/F ratio. Borderline for part throttle
       drivability (worse than borderline if EGR is used).
18-19  Very lean.  Usual lean limit (Driveability).
20-25  Lean burn limit.  Varies with engine.

If your engine has a cam with a lot of overlap, your wideband may read rich
at idle, even though it isn't.  In cases like that, I usually set the idle
mixture use the RPM drop method.  Adjust mixture to yield maximum RPM, then
lean it so the idle drops 20 to 40 RPM.  Even with milder cams on engines
with carbs, I usually only get around 13:1 at hot idle if the idle speed is
low enough.  Raise the RPM a bit and the air fuel ratio will quickly go to
the cruise value.  You can fine tune this by lowering the float level, if
need be.  Also placement of the sensor (in collector versus near tailpipe)
can yield different results.  Dyno shops using tailpipe probes will often
shoot for a leaner WOT air-fuel ratio in the 13.0 to 13.5 range.  Measuring
at tailpipe tends to read leaner than if it was measured at the header.
13.2 at the tailpipe can correlate to 12.5 at the header.  The best thing
to do is to correlate air fuel ratio on a dyno with maximum power for your
set-up.  You can accomplish the same thing with an accelerometer based G-meter
or at the drag strip.

I started the tuning session by driving the car around to warm it up before
setting the initial idle speed and mixture.  Timing and spark strenght had
been set previously.  It drove well with no bogs or flat spots but the old
plugs and tail pipes suggested the mixture was rich.  The carb was installed
out of the box without any jetting changes by the previous owner.  The
baseline specs of the manual choke 1404 carb are:

 1423 jets (0.086" diameter) in primaries
 1460 rod (0.065" x 0.052", stamped 6552)
 orange step up springs (5" Hg)
 1426 jets (0.095" diameter) in secondaries
 center position accelerator pump
 0.0935" needle and seat
 11/32" float height (+/- 1/4" float drop)
 middle accelerator pump link hole
 0.028" accelerator pump nozzle diameter
 500 CFM rating

I set the initial hot idle speed to 750 RPM.  BTW, with the MSD, it starts
just fine without the choke and will idle down to 400 RPM.  I set the initial
idle mixture using the method suggested in the Edelbrock carb owners manual
(richen to maximum idle speed then lean to a 20 RPM drop).  I then installed
a wide band O2 sensor in a bung I had welded in previously.  It's located aft
of the passenger side header collector with the sensor cable routed through
the engine compartment (zip tied away from anything hot) out the rear of the
hood into the passenger side seat.

<end of part 1>

The wideband showed the the idle mixture at 13:1.  I was initially going to
try for stoichiometric (14.7:1) but settled on 13.5:1.  Then I noticed the
representative charts in the Edelbrock catalog show a 13:1 idle mixture.  Be
aware that cams with a bunch of overlap can fool the wideband at idle due to
the unsteady nature of the idle (misfires result in unburnt fuel passing over
the sensor).  This will go away quickly as the revs rise.  Since it's
dangerous to drive and read the gauge at the same time, I had a neighbor do
the driving while I read the meter.  It was immediately obvious that the base
calibration was way too rich (2 points across the board) for my engine.
Edelbrock provides nice little charts that correlates the rod and jet
combinations to changes in mixture:

 http://www.bacomatic.org/gallery2/v/hidden/dan/dan-cars/album25/1404_tuning_chart.jpg.html

Each step is approximately a 4% change.  With the mixture readings from
the meter, it's possible to calculate how many steps are required to get
a desired mixture.  Be aware that not all possible combinations appear on
the chart and only the highlighted ones are possible using the tuning kit.
You'll need to purchase additional tuning parts for the other combinations.
I wrote a little program to calculate all the possible rod/jet combinations
and it helped me find a couple of rod/jet combos that worked better in my
application.  I'm now cruising at 15.5:1 (lean of stoich for better fuel
economy) with transition mixtures (climbing a hill, pulling a load, part
throttle acceleration) in the 13.5 to 14:1 range and WOT in the 12.5 to
13:1 range.  Pretty much ideal.  When I started, it was rich across the
board.  I've picked up 3 MPG and power to boot.  Using the combos on the
chart, I couldn't get the primary power mode lean enough so I had to
compensate by leaning down the secondaries a bunch.  Getting the right
primary rod/jet combo fixed that.  Depending upon where your cruise mode
mixture is, you may want your primary power mode to go between 15 and 25%
richer.

The initial reading was so rich, I didn't even bother to write the air-fuel
ratios down.  Returned home and went three steps (12%) lean on both the
primary cruise and primary power modes:

 #10 on chart, #1422 jets 0.083", #1463 rods (0.067" x 0.055")
 Effective Cruise Jet Area (sq. in.) =  1.8849555921538760E-03
 Effective Power Jet Area (sq. in.)  =  3.0347785033677404E-03
 3.0348 / 1.8850 = 1.61 or 61% more area
 Flow is proportional to area squared
 0.61**2 = 0.372 or 37.2% power enrichment flow

That was too lean at cruise (had a mild surge) so I went to #11 on the chart
(2 stages lean on both cruise and power modes):

 #11 on chart: #1422 jets 0.083", #1460 rods (0.065" x 0.052")
 Effective Cruise Jet Area (sq. in.) =  2.0923007072908024E-03
 Effective Power Jet Area (sq. in.)  =  3.2868913138183213E-03
 3.287 / 2.0923 = 1.571
 0.571**2 = 0.326 0r 32.6% power enrichment

That put me near stoichiometric (14.7:1) at cruise but was rich in power
mode which also made WOT rich, even after leaning the secondaries from
the baseline 0.095" jets down to 0.080" (five 4% lean steps).  The power
mode was too rich, causing me to try compensate by leaning out the
secondaries.  The tuning chart doesn't show any other combinations that
would allow me to fine tune this set-up but using the program I wrote, I
came up with some additional combinations not on the chart that worked
better:

 not on chart: #1422 jets 0.083", #1461 rods (0.065" x 0.057")
 Effective Cruise Jet Area (sq. in.) =  2.0923007072908024E-03
 Effective Power Jet Area (sq. in.)  =  2.8588493147667119E-03
 2.860 / 2.092 = 1.367
 0.367**2 = 1.135 or 13.5% power enrichment

 not on chart: #1421 jets 0.080", #1441 rods (0.062" x 0.052")
 Effective Cruise Jet Area (sq. in.) =  2.0074777056438778E-03
 Effective Power Jet Area (sq. in.)  =  2.9028316119169689E-03
 2.9028 - 2.0075 = 0.8953
 0.8953 / 2.0075 = 0.446
 0.446**2 = 0.200 or 20% power enrichment

I purchased the additional rods and jets for these two combinations.
When I went to install them, I noticed the #1461 rods were machined
differently.  Both carry the same part number, though it's obvious
they are from different production batches as the stamped lettering
is different.  Also, the length of the machined step is different
and the tip of one of the rods was blackened which solvent failed to
remove.  I put the micrometer on the rods and they varied more than
I cared for.  I also have a Carter 400 AFB here and, as luck would
have it, it was equipped with rods with the power mode of the #1461
rods but a slightly leaner cruise mode.  Since the rods are so easy
to change out, I swapped the rods over and took it for a test drive.

 not on chart: #1422 jets 0.083", Carter 400 rods (0.066" x 0.057")
 Effective Cruise Jet Area (sq. in.) =  1.9894135478857367E-03
 Effective Power Jet Area (sq. in.)  =  2.8588493147667119E-03
 2.860 / 1.989 = 1.437
 0.437**2 = 0.191 or 19.1% power enrichment

Cold, without the choke, there was some mild lean hesitation but that
quickly went away as the engine warmed.  The leaner rods did the trick.
By leaning out the primary mode, WOT also leaned out without a further
seconday jet change.  With the Carter rods, I'm now cruising at 15.5:1
(lean of stoich for better fuel economy) with transition mixtures
(climbing a hill, pulling a load, part throttle acceleration) in the
13.5 to 14:1 range and WOT in the 12.5 to 13:1 range.  Pretty much ideal.
I got 19 MPG on the last tankful when I was cruising at 14.7:1, up a
couple of MPG from where I started even though I was mainly WOT tuning
the carb or cruising at 75+ MPH.  With the latest carb changes, I think
it'll do 20+ MPG around town.  Edelbrock doesn't have a 0.066" x 0.057"
rod listed.  The next size listed is #1436 rods (0.068" x 0.057").  That
would be too lean at:

 not on chart: #1422 jets 0.083", #1436 rods (0.068" x 0.057")
 Effective Cruise Jet Area (sq. in.) =  1.7789268400952206E-03
 Effective Power Jet Area (sq. in.)  =  2.8588493147667119E-03

With all Edelbrock parts, the best bet would likely be the other
combination listed above that's not on the charts (#1421 jets 0.080",
#1441 rods 0.062" x 0.052")

There are some other rod/jet combos that give similar but slightly different
effective jet areas.  If I was really trying to fine tune this combo, I'd
test each on the same day with an accelerometer to see which gives the best
performance.  Vizard has gone so far as to put O2 sensors in each header
primary so he could stagger jet to better equalize the mixture distribution.
IIRC, he said it was worth 20 HP on an engine in the 300 cubic inch range.

Some miscellaneous observations.  While it starts just fine with no choke,
you don't get the fast idle cam so it idles at 500 RPM cold versus 750 hot.
The Offenhauser Dual Port/AFB combo is very smooth and I could not tell when
the secondaries opened.  It will pull 3rd gear idling through the subdivision
and is, in fact, smoother than the fuel injected 5.0L in my 1987 Mustang.
In case you aren't familiar with it, the Dual Port design has the runners
split into top and bottom sections with the plenum split fore and aft (rather
than the usual side-to-side).  It's essentially two single plane intakes
stacked on top of each other with the 4 barrel carb primaries feeding the
longer path lower runners and the secondaries feeding the shorter path upper
runners.  It's an interesting concept but the packaging required to fit within
a carb intake envelope does compromise the design.  The Offy Dual Port has the
reputation of being a good low to mid range intake with excellent throttle
response and fuel economy (probably the best cruise fuel economy based upon
BSFC tests performed by David Vizard... I've got the dyno results here some
place).  It's not a high rpm intake but Offenhauser says you can trim the
dividers back at the head flange to pick up some top end power.  An Edelbrock
Performer Rover intake will make better power but is taller so hood clearance
may be an issue.  The Edelbrock foam air filter has got to go.  The air fuel
readings were not quite as stable as the readings I got on my buddy's 428CJ
powered Cobra replica. That might be due to the size of the motor relative to
the carb (750 CFM on a 428 versus 500 on a 215).  Also, the accelerator pump
shot and/or clogged fuel filter may have had something to do with it.  Make
sure to tighten the fuel line after you've re-jetted or you'll spray fuel on
the hot manifold.  Doh!  It's tough to read the numbers on the rods and jets
once they get stained with fuel.  I resorted to the calipers (for the rods)
and a tapered rod (for the jets) to verify the sizing.  I then bagged and
tagged the rods in individual zip-loc baggies (the really small snack size).
I noticed a little sediment in the bowls and cleaned it out.  I'll check
again later. I was shifting at 6000 RPM and managed to bump the 6400 RPM rev
limiter once while watching the meter.  I bought a dual outlet for cigarette
lighter so I could use the radar detector during tuning.  The wideband O2
uses the lighter socket to provide power for the sensor heater.  The primary
side venturis of the 600 CFM AFB's are the same size as the 500 CFM AFB's
though the boosters are different.  Need to get a G-meter to correlate
acceleration (power) with mixture.

As a follow up, the Offy Dual Port seems to need a richer mixture at WOT.
Even though wide band showed I was in the range I wanted to be at WOT, top
speed was down a bit so I suspect individual cylinders are leaner than the
average reported by the wideband.  A recent dyno test of an Offy Dual Port
on another engine showed it to respond to richer than usual mixtures, likely
due to poor mixture distribution.  The Dual Port is an interesting design
but has quite small passages inside and, if any engine does not need a dual
port, it's the small port Rover/Buick.

The wide-band I used is an Innovate LM1 and it's very nice.  It allows you
to datalog or just watch a hand held display.  On most of the cars I've tuned,
we've welded a bung into the header collector but on some we've had to go a
few feet aft of the collector.  I typically use zip ties to temporarily route
the cable into the car via the passenger side window.  When datalogging, be
aware noise from high power ignition systems can cause an erratic RPM converter
signal.  A 50k-ohm potentiometer from Radio Shack will usually fix that problem.

Dan Jones