Carb talk for Beginners

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

GreenTwin

Well-Known Member
Staff member
HMEM Supporting Member
Global Moderator
Joined
Jul 2, 2021
Messages
3,751
Reaction score
1,501
Location
MidSouth, USA
I have been following along with some discussions in another thread, and the more I read, the more confused I get (more confused than my normal confusion).

I have seen some proposed forumlas for calculating the venturi dimensions.

Discussions of how carburetors scale down from a full sized engine to scale model engine.

Carbs for multi-cylinder engines, and single cylinder engines.

Carbs for 2-stroke vs 4-stroke engines.

Manifold designs.

Many variables discussed.

.
 
I understand carburetors to some extent, because they are much like my foundry burner.

I ran a series of tests for my foundry siphon-nozzle oil burner, and my furnace, in an attempt to find out which air and fuel settings/ratios would produce the hottest temperature inside my furnace.

My furnace interior diameter is 13", with a height of 14".

After a great deal of experimenting, and some help from our member ArtB, who gave a great explanation of burner operation/dynamics, I finally figured out how to adjust my oil burner for maximum temperature inside of my furnace.

As it turns out, my furnace has a fixed interior surface area, and that surface area is capable of completely combusting X amount of fuel (gal/hr) with Y amount of air (cfm).
Any more fuel or air other than these X and Y numbers, and any other ratio of fuel-to-air, produced a colder furnace interior.

Things I learned:

1. There must be an exact ratio of fuel-to-air to maximize combustion efficiency, and maximize the temperature inside the furnace.

2. Increasing the fuel/air flow rate would speed up the melt process, up to a point, and after that point, increasing the fuel/air actually created a cooler furnace interior, and slowed down melt times.

3. Using a pressurized fuel tank (diesel for fuel), with 10 psi on the tank, and a 30 psi safety valve, eliminated the need to ever adjust the burner. I discovered that pressurizing the fuel tank on my RC aircraft engine also greatly improved the stability of the engine operation, and allowed for a very slow idle regardless of fuel tank position, level of fuel in tank, or relative position of the tank to the carburetor (ie: the relative to whether the plane was climbing or diving).

.
 
I have toyed with various small engine carburetors for many years, and am especially familiar with single-cylinder motorycle carburetors, for both 2-stroke and 4-stroke engines, such as the one shown below (not my images).

I have read about adjusting these carburetors by raising or lowering the needle, and also by changing the jet.
This seems like a pretty simple design.

I always assumed that the intake side flaired out so that it would be an appropriate size to connect to the air filter box, but it would seem things are more complex than that.

I find these carburetors to be easy to adjust for multiple speeds, and easy to adjust for a good idle.



MjAxNDQ4NQ-d1f8a7f3-1728299687.png
motorcycle-carburetor-1496472631.jpg
 
I have also flown model airplane engines for many years, and am familiar with the simple carburetors used on 2-stroke model airplane engines.
(not my images).

carburetor_10-3816223850.jpg
carburetor_15-802093950.jpg
 
And I use to hang out at my buddy's house, and he built racecar engines, only Chevrolet, and generally 427 L88's and 454 LS6's.

He used Holley 1150 double-pumpers, and ran some radical high-lift/high-duration cams, and one engine ran on alcohol.

(not my image)

I am sure he was generating over 500 hp on many of his engines, and his 1/4 mile drag strip times were in the 11's.


.

s-l500-565372122.jpg
 
Back in the day (seems like yesterday), when the baddest bike that could be had was the big Norton Commandos (not my image), it was very common for the motorcyle magazine folks to take a stock motorcycle to a customizing shop, and have it ported/polished/adjusted, etc. to see just how much power a stock motorcycle engine could be made to produce.

One fascinating story revolved around the Yamaha SR500, which is a single-cylinder motorcycle that I am keenly interested in because I happen to own one.

The magazine guys would work on the engine/carburetor for the SR, and do all of their black magic tricks.
After adjusting the engine, they would put it on the dyno, only to find out that the engine produced less power.
They went through many iterations and adjustments, and no matter what they did, the engine produced less power than the stock engine.

Finally, in desperation, the magazine folks tracked down the Japanese designer of the SR engine, and asked him what was going on.
His response was that the stock engine was optimized, and any variation to anything on the engine would produce less power.
Even very tiny changes like the angle of the valve seat, or any modifications to the carburetor dimensions caused a power loss.

I found a printout of the SR500 torque band, and it is completely flat across the entire motor rpm range.
The flat torque curve makes the SR an absolute joy to ride, and eliminates most gear shifting, since the engine produces so much torque across the rpm range that you can just roll open the throttle and to, almost ignoring what gear you happen to be in.

.

vad_machineofyear.jpg
vad_historical_33.jpg
vad_historical_23.jpg
 
Last edited:
A brief review of "carburetor design" on the internet seems to reveal that like a foundy furnace, there is a maximum size carburetor that can be used with a stock IC engine, and any carburetor larger than that will not produce any additional power, and like I use to see in the old hot rod days, strapping on a large carburetor on a stock auto engine would often cause it to run slower.

Once an appropriate cfm carburetor is selected, then there is still much tuning to be done, and it seems that one has to determine what you want to achieve, ie: low speed torque, high-end horsepower, mid-range power, etc.

Auto engines are tested on a dyno, and the dyno graphs tell the story of the engine operation combined with a particular carburetor and those specific carburetor settings.

.
 
In the model engine carburetor discussions I have seen yesterday, there seems to be a lot of emphasis on the exact dimensions of the venturi for a given engine displacement.

Obviously most model engineers don't have a dyno with graphical printout that could be used to test various carburetor designs.

In the model engine world, we use to judge the performace of the engine by how large a prop with a give pitch the engine could turn, and how high an rpm could be achieved with that propeller.

For model engines that operate without load, it seems that the carburetor performance is judged by how high an rpm the engine can reach.

The discussions beg the question "How does one design a carburetor for a given engine type/size ?", and more importantly "How does one test the carburetor to verify that it performs as it was intended by design ? ".

And how does one test various venturi/inlet/outlet passages to see which dynamic range it best functions within ?

The more I read, the more confused I get, since I don't see any rigid criteria against which a given carburetor design can be tested.

And I don't see any discussion of how small carburetors such as used on a 2-stroke weed eater engine are designed.

.
 
Some of the discussions mention specific design criteria that have been proven for specific engines, but other design approaches seem somewhat random and arbitrary, especially given no concise method to test and compare different designs, other than perhaps reading rpm.

It would seem that a good starting point would be the smallest weed eater carburetor, and reverse engineering from a known design with known performance across a wide rpm range.

Why not capitalize on what must be many years of research by others with small engine carburetors ?

.
 
I'm the opposite and quite like a slow reving engine rather than one that buzzes it's nuts off. Though I have been known to make the odd model aero engine with their usually high revs.

Slower reving engine of the same displacement as a high revving one will not want the same venturi size as the volume of air being drawn through will be considerably different. maybe 10 -20 times less. This is why something like a Hoglet with it's 1.7cu in capacity can run on a carb that would not be out of place on a 0.15-0.20 cu in glow engine.
 
Last edited:
With foundry burners, there is typically a range of operation that a good oil burner will function, such as a 4:1 ratio (minimum to maximum output), while still getting efficient operation.

I suspect with carburetors, it may be similar, and thus a range of carburetor sizes could work for any given engine.

One carburetor size my be optimum, and may be the most efficient, but I would think a gas engine could still operate fairly well with a sub-optimum carburetor design, especially when achieving peak/maximum power is not critical.

I still say much depends on whether the engine will be loaded or not.
With an unloaded engine, I would think there is much more room for error in a carb design.

.
 
It would seem that a good starting point would be the smallest weed eater carburetor, and reverse engineering from a known design with known performance across a wide rpm range.

Why not capitalize on what must be many years of research by others with small engine carburetors ?

.

The problem with even the smallest weed eater carb is they are quite large and complex to recreate particularly if scaled down

The rev range is usually due to more than one needle valve and jet one for idle and another for top end running.
Various diaphrams that regulate flow.
Jets and needles are finer than many would be able to make with basic lathe and workshop tools
Look wrong for anyone building a scale IC engine

In the case of a hit & miss engine, you also need a carb that will work with the interrupted air flow as the engine only sucks on a hit and needs to deliver a small shot of fuel for just that stroke so the engine hits and does not stutter. This is the opposite to the carb you may find on a small diesel aero engine which is flat out from starting until the tank is empty.

Throw propane firing into the mix instead of petrol and you also need to consider shutting off the propane when the engine is not on the intake stroke. You may not even need a carb on propane if the valve is done right as their is no liquid to atomise
 
Initially considering a 4 stroke configuration, one would consider the operational RPM range, and a cam design that did not seek to generate higher percentage of cylinder filling from valve overlap impulse tuning, a single cylinder's swept volume would need to be filled with mixture thru a given orifice (venturi) size determined by a targeted atmospheric pressure drop, measured at the maximum intended RPM range. In a single cylinder, that would register as a single, sharp impulse being communicated to the venturi once every 720 degrees, again, not considering any impulse reflections or inertial, ram effect tuning.

In a multi-cylinder engine, if still fed by a single venturi, the impulse signals would be less sharp, but more of them, separated in time by the rotational degrees by the number of cylinders drawing on the venturi in 720 rotational degrees of the 4 cycle process. The more cylinders, the more constant but less sharp the communicated signal to the single venturi would be, and the smaller it can remain as it begins functioning as a device exposed to lower but more consistent signaling.

Mixture is best smoothly and constantly accelerated from its initial velocity as it leaves the venturi to the valve throat where it is delivered into the cylinder. This is done by reducing the runner angularly along its length.
 
I have several of the smallest 'weed eater' carbs. Several fellows from the BAEM club used them on their engines. I was in my learning phases so I thought I would give one a try. Before I get into particulars let me say that the smallest WE carb is larger than most engines would require. Now the fine points. WE carbs operate on crankcase impulses to draw fuel to the carb. A 2 cycle engine has a large crankcase impulse signal because of the closed crankcase and the way the fuel is moved inside the engine. A 4 cycle engine generally has a vented crankcase either to the atmosphere or back through the inlet tract. Excess crankcase pressure would cause oil to be pushed back past the rings. A WE carb regulates the liquid fuel by means of a vacuum signal through the venturi. This pulls the diaphragm which in turn allows a small needle valve to open and allow fuel to flow through the internal passages to the control needles for idle and high speed. Now the needles. Way too coarse for small model engine work, both in taper and thread pitch.
Can they be made to work the answer is yes but why when a simple air bleed carb works fine. Miniature engines are built to replicate the internal combustion process in 99% of the cases. Rarely are they put to work. Probably the ones that work the most are the type that would power an airplane.
When I go to a show and display and run my engines people are thrilled and amazed that something that small could be made to operate.
 
I have several of the smallest 'weed eater' carbs. Several fellows from the BAEM club used them on their engines. I was in my learning phases so I thought I would give one a try. Before I get into particulars let me say that the smallest WE carb is larger than most engines would require. Now the fine points. WE carbs operate on crankcase impulses to draw fuel to the carb. A 2 cycle engine has a large crankcase impulse signal because of the closed crankcase and the way the fuel is moved inside the engine. A 4 cycle engine generally has a vented crankcase either to the atmosphere or back through the inlet tract. Excess crankcase pressure would cause oil to be pushed back past the rings. A WE carb regulates the liquid fuel by means of a vacuum signal through the venturi. This pulls the diaphragm which in turn allows a small needle valve to open and allow fuel to flow through the internal passages to the control needles for idle and high speed. Now the needles. Way too coarse for small model engine work, both in taper and thread pitch.
Can they be made to work the answer is yes but why when a simple air bleed carb works fine. Miniature engines are built to replicate the internal combustion process in 99% of the cases. Rarely are they put to work. Probably the ones that work the most are the type that would power an airplane.
When I go to a show and display and run my engines people are thrilled and amazed that something that small could be made to operate.

Thanks for the info.
I think a design for a model airplane engine would have to be much more exact and correct than a display model engine.

I guess the engines I intend to build sort of straddle the full size engine / scale model engine world, and are really more like small functional engines, such as a weed eater motor, and producing measurable fractional horsepower.

I would like to power something with my engines, such as a scooter, a bicycle, or whatever I can hook it to, and thus I feel like I need more than just a simple carb.

From what everyone has said though, if a simple air-bleed carb will do the job, then there is no need to build a complex carburetor, and so I will keep it simple for now.

Less is more as they say.

Thanks for all the input/feedback from those who have experimented with carbs.
Much appreciated.

.
 
Lots of good stuff on this thread.
Last year I tried to "tune" my 1979 Guzzi V50 as it was built to open the throttle quickly and Go! Reliably, durably, and fast enough for regular police in Rome..... But not super quick!
So in modern traffic, tootling along in crowded cities with traffic lights, grid-lock, 20 mph limits, and no space to open-it-up, the Guzzi was a real pain to go 3 or 4 miles through the city. It didn't want to know below 2000rpm. It was so rich just off idle I was weekly cleaning black off the spark plugs. Or as soon as it started to mis-fire with over fuelling, as per original cure for the flat spot when vacuum dissapeared as the throttle is opened.
So I made smaller idle jets than available from the manufacturer, machined the slides to reduce over fuelling just off idle (when it changes from idle jet fuel to needle jet fuel), and still run on the lowest ethanol level of fuel I can get.
I bought carbs from China, "ready to fit", but my fuelling is changed to almost OK.
Now the black plugs are not too bad. Idle is best I have ever had. Slow riding is better (Honda's are better, but it is more than the "barely rideable" of the original).
Lower octane, lead-free and ethanol loaded modern fuel is so far from original design the carbs were not giving a suitable mixture for modern use. But now they are, and the new Chinese carbs are still in the box.
So changing use and fuel can be a major problem for carbs on low settings. No problem at mid-to high throttle.
K2
 
Lots of good stuff on this thread.
Last year I tried to "tune" my 1979 Guzzi V50 as it was built to open the throttle quickly and Go! Reliably, durably, and fast enough for regular police in Rome..... But not super quick!
So in modern traffic, tootling along in crowded cities with traffic lights, grid-lock, 20 mph limits, and no space to open-it-up, the Guzzi was a real pain to go 3 or 4 miles through the city. It didn't want to know below 2000rpm. It was so rich just off idle I was weekly cleaning black off the spark plugs. Or as soon as it started to mis-fire with over fuelling, as per original cure for the flat spot when vacuum dissapeared as the throttle is opened.
So I made smaller idle jets than available from the manufacturer, machined the slides to reduce over fuelling just off idle (when it changes from idle jet fuel to needle jet fuel), and still run on the lowest ethanol level of fuel I can get.
I bought carbs from China, "ready to fit", but my fuelling is changed to almost OK.
Now the black plugs are not too bad. Idle is best I have ever had. Slow riding is better (Honda's are better, but it is more than the "barely rideable" of the original).
Lower octane, lead-free and ethanol loaded modern fuel is so far from original design the carbs were not giving a suitable mixture for modern use. But now they are, and the new Chinese carbs are still in the box.
So changing use and fuel can be a major problem for carbs on low settings. No problem at mid-to high throttle.
K2

That is a good point, carburetor design is no doubt fuel-centric.
I recall my buddy changing his racing car to alchohol, and having to change the jet size significantly.

That Guzzi is a classic design, and somewhat unique as far as I can remember.
Are there any other V-twin motorcycle designs out there in that format with the cylinders to the sides ?

.
 
Steamchick,
There had to be something else going on with your carbs. I have worked on motorcycles since my late teens. (now 79) Everything from Whizzers, Cushmans, Triumphs, Ducatis, Gileras, and all the Japanese 2 and 4 stroke engines. I can honestly say I have never run across a carb that wouldn't run properly at low speed. Low speed, not idle, is controlled by the needle coming out of the needle jet. The needle works from off idle to about 3/4 throttle when the main jet takes over. Idle circuits on older machines have either a fuel or air adjustment on them. I'm not sure what Guzzis have probably a Delorto. With that vintage there should be an adjustment screw for the idle mixture. Off idle as the engine is being throttled is also controlled by the cutaway on the slide but if the bike ran fine originally then there's something else going on. Alcohol blended fuels used in engines require more fuel than old non alcohol fuels so if anything the bike should be running lean. Straight alcohol versus gasoline requires about 40% more.
I know the fuels these days is crappy but I've never had to change idle jets.
 

Latest posts

Back
Top