7 march 2019

tuning the Weber carburetor for non-standard applications

specifically here, the Rambler 195.6 OHV (3200cc) inline six

reference: Top End Performance, highly recommended by me.

reference: Reline/Weber IDF adjustments

reference: factory manual

reference: tim's roadster, weber tuning

reference: 240260280 weber tuning

reference: Rambler 195.6 OHV inline six

nearly all of the Weber tuning information, including that from Redline/Weber, make assumptions about the motor it's on that simply do not apply to applications like mine -- a single carburetor on a manifolded, small bore long stroke multi-cylinder engine.


there are many web pages on tuning Weber carburetors; here's another. the differences here, mostly low-speed and idle tuning method, i try to back up with facts and measurement, not rigid rules. my experiences trying to use existing sports car guides forced me to actually understand what was going on underneath the rules given by knowledgeable sources such as Redline and racetep.com.

old inline motors are not sports car motors

 "sports car" old multi-cylinder inline
A small engine, high RPM larger engine, low RPM
B large bore, short stroke small bore, long stroke
C one venturi per cylinder single carb (2V) feeds all cylinders
D higher-overlap cam (low vacuum) low-overlap cam (high vacuum)
E no "vacuum leak" sources PCV and other "vacuum leaks"
F weak part-throttle signal strong part-throttle signal

A) this is the most basic problem: reference to RPM as a stand-in to talking about flow. there's no easy way to know or state how much air mass is moving through the carb at any point, and for a given engine, RPM is "good enough". i do it too, here. and it's true that many of the tuning guides out there (240 260 280 Z cars, for example, or Tim's Roadster) are in terms of a specific engine. but the general guides from Redline et al don't distinguish at all, giving the impression that RPM itself actually means something to a carburetor.

B short stroke big bore (ratio) motors make for short fast sharp intake pulses. skinny bore long stroke and intakes all manifolded smooth the pulses that the carb sees as signal. there is merely one of the components of A above.

C the number of cylinders per venturi matters deeply. most traditional american car lore generally discusses air flowing through a carb as a continuous "flow". it's not a flow at all, it's a series of discrete pulses, especially in traditional sports car one venturi per cylinder (or pair). in my engine all six cylinders pull from one carb feeding a single trough (manifold) and in single-manifold cars the pulses are somewhat smoothed together, into a "bumpy" flow. and this matters to carbs, adn specifically Webers, in the part-throttle response. it's discussed in the Weber factory manual, but not at all in the simplistic how-to-tune guides.

D "most sports cars" (there i go with assumptions) or at least many have relatively large cam lobe separations and weaker vacuum signal at low speeds. note that Weber docs rarely talk about "inches of vacuum" or tune low-speed in terms of vacuum. this is not an accident. american passenger cars (stock-ish) generally have excellent smooth vacuum due to low lobe overlap, and were designed to make torque at smooth low speeds. though even a "hot cam" in an inline six/single carb will have the advantage of overlapping pulses.

E vacuum leaks are bad, of course, but PCV and other good, useful things inject air into the cylinders, rendering much advice around idle-setting wrong. PCV is not a problem in any way with Weber carbs but it confounds the nearly universal instructions on setting idle and low speed. i cover this in detail below.

F multi-cylinder inline, single carb engines have a nice strong "signal" (combined air pulse flow) at so-called part throttle -- which is exactly where carburetors suck at sucking. idle: fine. WOT: fine. part-throttle, i confuse. this difference manifests when trying to tweak the so-called crossover or transition from low-speed to high-speed jets.

in summary -- when i make assumptions below i am assuming my engine or ones like it -- six or more cylinders, one carburetor on one manifold feeding all cylinders, a relatively mild cam, and a working PCV system. and vacuum windshield wipers (as i don't have enough tuning challenges).

carburetor circuits

briefly (it's covered everywhere) the circuits in the carb are:

IDLE SPEED SCREW: this is the mechanical throttle stop. this is as all Weber docs say: open no more than one turn of the screw against the stop. this is not for reasons surrounding air flow, but to set the position of the edge of the throttle plate over the tiny transition holes in the side of the bore.

the exact location of the thin edge of the throttle plate is critical and the source of off-idle flatspots. see the table of photographs far below for a detailed view of the relationship between idle, throttle plate transition and the transition holes.

IDLE MIXTURE SCREW: a tapered needle on a screw in a conical orfice that controls the amount of fuel fed to below the nearly-closed throttle plate. though the idle mix screw is fed by the low-speed jet the size of that jet doesn't matter (much) -- you can properly set idle mix regardless of the jet(s) installed. the idle mix screw does it all.

the most common misstatement in Weber documentation is the alleged rule of "...if the mixture screw is out more than X turns, the low speed jet is too lean..." THIS IS NOT CORRECT or at least not universally true.

the low-speed jet has only a minimal and incidental effect on idle mixture. the fuel flow through the screw and its orfice is small compared to the low-speed jet. you can always set idle mix correctly regardless of the low-speed jet size.

IDLE TRANSITION HOLES: these are well described elsewhere and worth the effort to understand them: with the throttle closed (idle position) the edge of the throttle plate must block the transition holes. off-idle flatspot/bog is caused by the throttle plate being slightly above or below the transition holes, instead of precisely over them.

"transition" refers to the response of the carb when the throttle plate is just lifted off the idle speed stop, but before it is open enough to fuel through the low-speed jet. the transition holes (should be) blocked by the edge of the throttle plate; as it begins to open, the holes are uncovered, drawing fuel through the holes, richening the mix as the throttle plate continues to open, finally flowing enough air to draw from the low-speed circuit (and eventually the main venturi).

this drawing from page 10 of the manual is illuminating:

i've not had a Weber carb have these circumstances correct "out of the box" even once. fixing this is done after you've nailed the low speed jetting. i have some very detailed photos of the transition further below.

LOW-SPEED JET: aka "idle jet". nearly all moderate driving is done on the low speed jets! these jets are in their own holders and changeable without taking the air cleaner off in all Weber models. low speed jets supply fuel to the engine from idle through highway light cruise speeds. low speed jets are the critical ones to get right, first. they feed the idle circuit fuel well, but that hardly matters -- no matter what jets you have in there you can get the actual idle mixture correct.

MAIN JETS: these are the obvious target of high-speed tuning, and my experience is that they're easiest to get right, even though it involves controlled high-speed highway driving and pull-over-swap-jets sessions to get right. main jets are fairly straightforward in operation: they determine air/fuel ratio when the throttle is "substantially" open. however the main jet, air bleed, and emulsion tube work together to create the foamy emulsion that's actually drawn into the carb throat.

AIR BLEEDS: also called air correctors, these are jets that meter the air mixed with fuel in the emulsion tube. the bottom of the emulsion tube is fed fuel metered with the main jet; the top of the tube has the air corrector open to the atmosphere. in the middle fuel mixes with air into an emulsion (foamy fuel+air) that is drawn through the booster venturi and down the throat of the carb.

since air is less viscous than fuel as the velocity of air down the throat increases it draws more fuel through the main jet, but proportionally more air through the air bleed; therefore the ratio of fuel to air varies with the air flow down the carb throat. for a given main jet, a smaller air bleed richens the mixture as flow (RPM) increases, larger air bleeds leans the mixture. though jet sizes have to be determined experimentally, the effects are straightforward and fairly easy to tweak.

SQUIRTERS: aka accellerator pump. stock has always worked for me luckily. a bit too much is better than not enough.

EMULSION TUBES: lucky me, never had to mess with. unlikely you will, either. with the DGV/DGEV series there's little point to it for most of us. for the IDF (and IDA, DCOE) this may be different.


as with many things, if you don't have a strategy, you'll get lost. here's mine.

above is a pretty great chart. where it says "RPM" think "relative air mass flow". as flow increases from idle to maximum, the contribution of fuel from each circuit -- idle mix screw, low speed jet, main jet -- changes. the crossover can be quite distinct when the mixture is lean in that area, when either the previous system (eg. idle) is lean and the next (eg. low-speed) isn't yet operating.

keep this graph in mind while tuning; hopefully it will remind you (and by you i mean me) to work on one circuit at a time, starting with idle and low speed.

nearly all driving is done on the low-speed circuits

once you having idling well enough to drive the car (below), low-speed tuning absolutely must be done first. it's common to bolt it on, get it to idle, go for a drive! then freak out when it won't run, farts, backfires, dies completely when you press the pedal, etc. STICK TO THE PLAN, WOE-MAN. the difference between impossible-to-drive and running-pretty-good (as a step on your way to running really great) is to pay attention to only low-speed stuff at first. only when low-speed is mastered will you be able to make sense of high-speed stuff.

...because "low speed" is a misnomer; the low-speed jet and even the idle screw dominate driving up through moderate highway light cruise, at which point it transitions -- crosses over -- to the high-speed (main) jet. in my car, 5 to 15 mph, flat level cruise is off the idle mixture screw; 10 to 50 mph or so, entirely on the low speed jet.

lather, rinse, repeat

as a strategy, get the low speed stuff "pretty good", then work on the main jets, after which you will probably have better ideas on low-speed etc. when flatspots occur, don't freak out and try to immediately eliminate them; they indicate lean-ness in transition from one circuit to the next-higher (idle screw to low speed; low speed to main jet) and reveal much. then you eliminate them next. divide and conquer, then consolidate. and keep thorough notes.

my tuning procedure (YMMV)

this procedure works for my engine, through experience. your mileage may literally vary. for all three Weber carburetors i was able to run the car with the box-stock carb and get the idle right and drive the car down the road enough to test and tune. if you can't, something's broken, it's not carb tune.

as mentioned i have a panel-mounted air/fuel ratio meter installed in the car. tuning without one is possible but extra-tedious. i guess i'd run too-rich, then jet down til it gets slow-ish, then back up. the $200 AFR meter was worth every cent.

obtaining jets

i end up paying more in shipping for jets than for the jets themselves. i've found it impossible to know which jets i'm gonna need (though after this last round, i think any next time i'll have a much better idea). i buy jets from Top End Performance. not eBay. i want real, matching, calibrated jets.

above is the 'kit' i drive around with to jet. mains, air bleeds, low speed jets left to right. on an IDF the jet stack comes out the top of the carb with the air cleaner cover off.

set float level first!

before even bolting the carb on, take the top off and set the float level. you won't want to do this now, but you should. correct float level is critical to every tuning adjustment and setting. symptoms and adjustments will be meaningless otherwise. and right out of the box it's not filled with gasoline that you have to dribble onto your bench and breathe in.

i have not yet had one Weber carburetor with the float level set even approximately correct out-of-the-box. my 44IDF was wildly off; it was supposed to be 10mm and it was more like 5mm. adjustment requires some care as the tabs you have to bend are small and stiff and attached to delicate plastic floats so you pretty much have to pull the pin, bend, reassemble, check, remove, repeat, ad nauseum. while you're in there you might as well verify what jets are actually installed.

carburetors are unlikely devices that rely on gravity, vaccuum from moving air masses, emulsions crawling up smooth metal surfaces, tiny orfices, etc. of course it's dominated by float level. ignore at your own peril and frustration.

i made a "tool" from a short piece of aluminum sheet; cut and filed to exactly 10mm high, used as a feeler gauge under the float. you can see it in the bag on the right.

idle speed stop screw, initial setting

bolt the carb onto the engine, and make rough settings to get it warmed up this first time: mixture screws out two or three turns, idle stop two turns after it touches the stop. allow the bowl to fill with gas, flap the gas pedal and do what it takes to get it to warm up and idle, even badly. the IDF has no choke so it takes a lot of initial gas pedal violence. don't worry about idle stop or mixture positions until it's warmed up enough to idle, even badly.

idle mixture screw: IGNORE THE LORE, DO THIS INSTEAD

the major problem with standard lore on Weber tuning is this flat-out statement from even authoritative Redline, that if your idle mixture screws are out more than 1 turn, your low speed jet (idle jet) is too lean. this is not correct. refer back to the old inline motors are not sports car motors table.

i could find no mention of any rule about the number of turns out for the idle mixture screw in the Weber factory manual. it seems clear that the final screw position for correct mixture will arbitrarily depend on a number of factors and is unimportant. (it should probably be out at least a half turn so that you have some adjustment range; and not so far out that it's loose.)

fact: no matter what low speed jet is installed, you can adjust the mixture screws for good idle.

fact: low speed jets critically determine engine operation from idle through highway cruise.

fact: low speed jets must be tuned with the throttle plate above the transition holes (1000 rpm-ish) up through where the flow operates the main circuits. only then is the final idle mixture screw position set.

finally, the usual lean-best-drop method works fine, which is suggested in most carburetor documentation. start rich, bring both screws leaner in increments until it runs worse (on the lean side); then back out until it sounds great, then out a bit more until it worsens (on the rich side). the difference between too-lean and too-rich should be fairly small, 1/4 turn or less. the correct position is in the middle, best-smoothest.

idle speed stop screw, final setting

the lore is correct on the IDLE STOP screw; it should be no more than one turn in once it touches the stop, otherwise you will have an off-idle flatspot. most likely, you will have an off-idle flat spot.


successful tuning requires being methodical. a repeatable flat level route, notebook, and a bag of jets. tuning takes me a few weeks, not because it's difficult, but because i have to order and wait for jets.

i have a convenient road to test on, Riverside Drive. flat, level, mostly straight, three stop lights in three miles. i can generally pick a lane and drive a steady 35 - 40 mph in 4th gear (low speed, low load) for long enough to tease the carb with my foot and get solid readings off the A/F meter. this is my most basic test bed.

tuning is impossible without good notes. i use Open Office, free spreadsheet software. paper works fine. i make (usually) one change at a time (two, if i know the result, eg. idle mix change after low speed jet change) and drive it. when changes were rough a short drive sufficed; as it got closer longer drives and closer (paper) notes later edited into the spreadsheet.


i drive my test route once to warm up the engine before making measurement passes.

it took some practice to feel out the three basic low speed carb states (referring to the graph). moving in gear with thottle closed you're driving on the idle mix screw, obviously at a crawl. slowly and smoothly open the throttle, to avoid the effects of the squirter, you will almost immediately feel the flatspot, if you have one. pressing further past the flatspot (a blip of the pedal for a squirt may be necessary) gets you past the flat spot where the carb will be operating on the low speed jet. here's that graph again:

for comparison, for my car 30 - 40 mph in 3rd or 4th, 1200 rpm or so, is solidly up above the transition. my engine remains on the low-speed jets up to nearly 2200 rpm in top gear, which is 60mph!


there are two "places" in the throttle you may experience flatspotting:
where why
off-idle, cracked-throttle flatspot idle screw mix to low-speed jet transition
mid-range, part-throttle flatspot low speed jet to main jet transition

both are fixable. but as frustrating as a flatspot is, at the outset they are useful "markers" for the two transitions that you'll be tuning around.

once i got a feel for how the low speed circuit operates, swapping the jets to get the right mix on the AFR meter was easy. light cruise down Riverside at 20, 30, 40 mph (1200 to 1800 rpm, in 3rd or 4th) the AFR was at least smooth and constant; the as-shipped 45 jet too lean, producing 16:1 AFR, but predictable.

with every low speed jet change you'll need to reset the idle mixture screws. they're extremely consistent: i keep a spreadsheet of settings and when i change to jet XX i go back to that mixture screw setting and it's correct.


tuning the main (and air) jets uses a similar process to the low speed jets; a repeatable test track (for me, the Los Angeles end of highway 2, Fletcher to Holly Drive) and repeatable speeds; this time, 60mph then 80mph, along with two convenient grades for generating load. 60 mph in my car is just past the transition to the main circuit, and 80 mph up one of the grades is a decent moderate load. repeatability is the key, plus copious notes.

one important gotcha on main jet selection is that while the main jet determines overall mixture, the air jet's effect is to lean out the main jet as flow ("rpm") increases. a small air jet "pulls" the main jet richer, and a large air bleed will lean out a rich (large) main.

the way most applications resolve this dilemma is to copy someone else's installation. i had to work this out from scratch.

my 44 IDF came with 135 main/175 air correction installed; this proved to be insanely lean. once flow reached the end of the low-speed circuit (around 55 mph) the effect was immediate shut-off: not enough fuel to run, like ignition cutoff. the throttle could not be "feathered" to get it past the brick wall.

drilling jets

drilling jets is generally bad practice, and for good reason, weber apparently flow-rates each jet individually. matched pairs matters more to me than precision. i don't want one side rich the other lean while you are trying to measure combined exhaust oxygen to tune. it's not an item i'm willing to cheap out.

that said, drilling jets is a quick way to get in the ballpark. since i had no idea what jets i needed, i used my American drill set, #1 through #60, and with a pin vise drilled the useless 135's upwards in drill-bit increments... .071, .073, .076, ... then wide open throttle went 16:1. a final drill size of .086", 2.18mm, gave me 13.5:1. with that information i bought suitable real jets (180, 200, 220) and continued tuning from there. i avoided buying a lot of jets i would never use.

with (real Weber) 220 main jets moderate highway cruise was in the 12.8:1 range; far too rich but quite drivable, and importantly, tuneable.

main jet selection by-the-book

only later did i discover the nomograph below on page 24 of the Weber factory manual, which indicates 200 (2mm) as starting jet size. i know now that is too rich, but it would be driveable/tuneable, a solid starting point.

air bleed jet size

what i know to be correct, pending my own further clarity on main/air interaction is that given a main jet that achieves the right mix comfortably above the cross over from low-speed, smaller air bleeds enrich the mixture (restrict air) as flow (rpm) increases; larger air bleeds (allows more air into the emulsion tube) leans out the mixture as rpm/flow increases.

with 220 main jets installed i substituted smaller air bleeds experimentally, not to be "correct" but to see what effect it had. further experimentation led to the conclusion that a smaller main jet combined with a smaller air bleed would accomplish the same thing with better control. i'm currently running 180 main, 150 air to great effect. i am doing research on this area now and i will update this page later.

[understanding (or is it confusion) begins to dawn: i think, need to verify, that there are multiple combinations of main/air jets that will accomplish the same A/F ratio; smaller air jets (and corresponding smaller main jets) will draw fuel further up the emulsion well, which will change the way the emulsion tube affects fuel dynamics. time to call Redline.]

my goal with this engine is to have my typical highway cruise speed of 75 mph to be at/barely over stoichiometric (14.7 to 15.2) but richen to 13.5 at any load above that; my particular engine is, to put it kindly, not a performance engine. it has been pushed far beyond factory recommendation. for many reasons i want a rich mix above 3000 rpm through the 4500 rpm maximum. the combustion chamber design isn't very good and it's detonation prone.


out, damned spot

the inevitable off-idle flatspot starts out a mere annoyance during initial setup but quickly becomes harmful. the off-idle transition, the idle itself, the flatspot, all depend on getting the new carb in basic tune and even on adjusting your driving to match, so i suggest doing it after low speed jet tuning, if for no other reason than you'll have more experience with it and a better feel for what's wrong.

the table below has images of one throttle plate on my 44IDF after idle tuning. my idle stop setting is only 3/4 turn open past touching the stop (not much air), and my idle mix screw is open 2.5 turns (a lot of gas). this is the result of my overly-large PCV valve; if i pinch the PCV hose it needs more stop screw and less idle mix. to fix this i have to find a more restrictive PCV valve.

(raw note) flatspot test/characterize: in my case, where PCV adds air and so idle stop is only 1/2 turn, the throttle plate edge is covering but below transition hole (then causes enleanment). to test: put throttle plate right at ideal spot (WHAT IS THAT -- 1.5 TURN? 1 TURN?) and adjust mix. speed will be too high, but that should eliminate flatspot. not solution, just characterization.

throttle plate and transition holes

if you look carefully at the high-resolution photos below (click the image) around 1 turn open the throttle plate is barely cracked open; it's not really visible, but it's open about .010" in the arc perpendicular to the shaft. at about 2 turns open the edge of the first transition hole becomes visible. the flatspot occurs because as the throttle is opened, the crack opens to leak more air but the transition hole isn't uncovered until later opening.

when i restrict the air from my "lean" PCV valve, i will have to screw the idle stop in (more air) and probably lean the idle mix (screw in) to get the idle back to 600 rpm/13:1 AFR. and the additional "stop" will mean that at idle, the throttle plate will be right over the first transition hole. not until i fix that will i mess with transition hole fixes.

the point of these paragraphs and the table of photos below to is to get across just how precise the idle stuff is. my flatspot is fairly bad; near stalling when cold, and an annoying hesitation when hot eg. at a stoplight.

the last photo ("many") shows all of the transition holes in the 44IDF, the throttle opened by hand.

in my particular installation the idle stop screw is in half a turn, no more; PCV is admitting air so the throttle plate is more-closed than it would be otherwise. (experimentation shows that plugging PCV requires another quarter turn of the stop screw in.) when i begin to open the throttle that initial tiny crack, air flow past the throttle increases but the transition hole remains covered -- flatspot. i know that the fix will be the recommended filing of a very small chamfer on the bottom edge of the thottle plate but i'm waiting untril i have everything else nailed down perfectly.

idle stop screw, turns in(arrow points to where transition holes appear)
.75 turn
1 turn
1.5 turns
2 turns
2.5 turns
3 turns
3.5 turns
4 turns

here's why you buy a real Weber, made in Spain. a lot of the cheap eBay clones look OK, but are often not very precise. and carburetors need to be extremely precise. this 44IDF from racetep.com shows how nicely the critical alignment of throttle plates vs. transition holes align. this means a precise relationship between: throttle shaft bore in the body, throttle shaft, the very precise position of the plates in the shaft, the bores of the tiny (half millimeter) transition holes, all working in concert.

idle stop screw 2.5 turns in after contact 
left bore right bore


hammer too small