Vapor lock with our modern fuel

[Dave Young]

After spending the past few months chasing a periodic "running lean" condition with my 1928 Chandler, I finally surmised that she was experiencing vapor lock, due to the temperature of the cast iron carb.  Apparently, the new blends of fuel have a much lower boiling point than even just a few years ago.  This stuff begins to bubble at 160 degrees.  My problem arose from my fitting a different carb to the manifold, a big iron Zenith, in place of the dreadful Schebler air-valve monster that was on there.  The Zenith, being much wider, sat fairly close to the exhaust manifold drop leg.  I had fit a heat shield at the onset, but eventually, after multiple roadside "sit-and-cools-offs" , made a much larger shield and now she runs along fine.  

 

I'm wondering if anyone else out there is having trouble with carburetor, ignition and valve type symptoms and at a loss as to the cause.  My car ran great any time it was less than 60 outside.  T-shirt days she would run along for 20 minutes and then I'd be limping home as she back-puffed thru the intake and ran like crap.

[mrspeedyt]

part of the curse of modern gas with ethanol. try to find some gas with no ethanol.

[caddyshack]

So much has been written about the ethanol problem out there and Mrspeedyt is right on. Buy ethanol free gas. It is available, more expensive, but it is available. I buy it all the time for boat motors, Corvette, mowers, Model A and T, and buy ethanol gas for the modern car and truck. 

I won't go into how I messed up the Corvette with Ethanol gas. 

If you made a metal heat shield, that may be a problem. I have talked with a few guys that have used the old style muffler wrap on their carbs. It looks like 1-inch wide asbestos tape from years ago. Works great for keeping the heat from transferring in tight spots. 

Hope you find a solution. 

[Dave Young]

I feel that I have solved the issue by adding the heat shield.  She runs along fine now.  The shield is a two-layer metal affair with a small air gap between the layers, as would typically be done on a wood stove to reduce clearances to combustibles. I would love to find some non-ethanol fuel. Any tips as to hunting it down in New Jersey?

[41 Su8]

Quote

Dave, when I drove my Packard in the summer months, I had the same problem. Solved it by using middle grade gas with a few oz. of Marvel Mystery Oil. I don't know if it was due to the higher octane or different volatility, but the difference was noticeable.  John 

 

[Larry Schramm]

Vapor lock is not new.  It has been around from the beginning of cars.  More so on some cars than others and more or less depending on altitude you live. 

 

We are just more sensitive to it because with pressurized fuel systems with the pump in the fuel tank vapor lock can not happen on modern cars and we do not see it any more..

[vermontboy]

Th old cure was to mount a can on the firewall and put a few coils in the fuel line and place it inside the can. Then you put ice in the can, cooling the fuel before it enters the carburetor.  There were also more elegant "cans" available on the aftermarket....

[caddyshack]

2 hours ago, wilbur said:

I feel that I have solved the issue by adding the heat shield.  She runs along fine now.  The shield is a two-layer metal affair with a small air gap between the layers, as would typically be done on a wood stove to reduce clearances to combustibles. I would love to find some non-ethanol fuel. Any tips as to hunting it down in New Jersey?

It looks like New Jersey does not have many sites, but "pure-gas.com" shows 8 locations. The price/gallon is outrageous. Maybe a little more research would help. Any New Jersey members care to comment?

Best of luck in your search.

[Dave Young]

Hey John, I need to see a picture of your Packard!  I have a wedding drive this Friday evening with the Chandler and it's going to be chilly out. All should be fine, but I do want to hunt down some straight gas somewhere.

[13CADDY]

been getting mine at the local boat marina--some small private airports will sell it too--just get yourself a couple of 5 gal gas cans--that's my biggest problem , you can't get close to the pump,without walking---good luck --Tom

[41 Su8]

Dave,  this is the 160 I used to own.  For me it was just too new. I am having more fun with the Special Coupe than I ever did with the Packard.  John

[emjay]

It also depends on when you buy the fuel.  They still used additives in the warmer months to raise the vaporization pressure and also the price.  With fuel injection becoming the norm, they may stop doing it or at least reduce the amount of additives.

[Guest]

We are having a huge problem here in California with the ethanol gas and carburated cars. Palm Springs hits the triple digits pretty often and I was having a similar vapor lock type issue. Especially after a 10 or 15 minute shut off and heat soak. I decided to wrap the steel fuel line from the fuel pump to the carb with Thermo-Tec, also covered the top portion of the fuel pump with it. I think I might be OK now, just have to wait for those triple digits again to be certain.

[Rusty_OToole]

There was much discussion on this topic earlier this year. One cure is to add from 10% to 25% kerosene or stove oil to the gas. The percentage depends on the compression ratio. Your Chandler should take the full 25%. Everyone knows not to use gas of too low octane but it is a mistake to use too high octane. Your car was made when gas was low in octane and contained a lot of heavy ends or oily elements like kerosene. One Australian cured a similar problem by adding diesel fuel. This is heavier than kerosene or stove oil but it worked, the car did not smoke, but there was a slight diesel smell at idle from the exhaust.

 

Owners of 20s cars report that their car runs better, smoother, cooler, and develops more power on the mixed fuel.

[DavidMc]

Adding Kerosene is certainly beneficial but its a common misconception that the increased volatility of modern fuels is related to octane rating, refer to the attached article by BP on the subject (Petrol =Gas)  

 

modern-petrol-vintage-engines.pdf

[SC38dls]

I would add the muffler wrap to the heat shield. But only put the wrap on the carb side of the shield. With the air gap and the wrap you are providing a double protection barrier. 

[PFitz]

Correct, the octane rating has nothing to do with vapor lock. It's what's added to the gas, such as ethanol, which has a much lower boiling point than gasoline.

 

Around here many stations have switched to e-free 91 octane . It matters not that it's high test, what matters is that there's no ethanol in the high test and that is what benefits cars prone to vapor lock.

 

One other common cause of vapor lock is lugging the motor, and not downshifting sooner to get the rpms up.  Lugging the motor makes it generate more heat while at the same time making the cooling system go slower.

 

Paul

[Marty Roth]

I've been very lucky to have good access to No-Ethanol gas locally, but my PureGas App does help when travelling. Here we generally pay $2.29 for pure 87 octane when normal gas is around $2/gal. Driving the '88 'Vette to central Florida last month for a Region Tour, we paid as much as $4/gal for No-Ethanol in the Florida Panhandle, and $3/gal in the Orlando area at Wa-Wa Stations.

 

When I cannot get No-Ethanol gas during a Tour, I use the LOWEST OCTANE AVAILABLE, and ADD 10%-15% DIESEL - readily available !

This will cause your engine to smoke, and to give the appearance of "burning oil", because it really is the case, but don't believe your friends who tell you that your rings are worn out - it is only the Diesel mixture,.

 

We did this during the 2015 Glidden Tour in Oklahoma city with temps in the 100 - 115 range while driving the 1941 Cadillac which is extremely prone to vapor lock, and never experienced the issue at all (while many others succumbed to the dreaded vapor lock).

 

Other things we did to the Caddy were to fashion a heat shield to protect the carburetor from the exhaust crossover pipe immediately behind it, and to install a COOL-CARB gasket which is two aluminum plates with a sandwiched poly-core to minimize heat transfer from the intake manifold.

[Guest]

Marty,

              I'm surprised to hear you had any issues with the ethanol gas in your '88 Corvette? Our gas here in California is terrible on all my carburated vehicles but I don't experience any trouble with my injected cars.

Greg

[John_S_in_Penna]

27 minutes ago, GregLaR said:

... Our gas here in California is terrible on all my carburated vehicles...

 

I've heard that gas must be different out West, 

so maybe other urban areas have similarly bad compositions

that contribute to the problem--more than just ethanol.

 

A friend in Washington state, said his gas would go bad

in only 6 weeks.

 

Another person, from California, agreed.  He told us in a

newsletter interview in 2013, "Modern gas, especially in California,

is so bad, that in some of my steam cars, if I don't go out

or the car hasn't been moved in a month or so, I've got to

drain the gas out because it literally won't fire.  You put a 

match to it and it will burn like oil."  He said he was using

a Sta-bil product, filling the tank right to the top, and 

typing to get the best-sealing gas-cap possible to try to

keep the fuel from going bad.  I assume he solved the

problem more or less, because he had had one steam

car up to 75 m.p.h. on the open road.

 

So car fans in Kansas or Pennsylvania or other states

might not have all the answers needed--but let's hope 

poster Wilbur has his Chandler's fuel problems solved now!

 

[Marty Roth]

1 hour ago, GregLaR said:

Marty,

              I'm surprised to hear you had any issues with the ethanol gas in your '88 Corvette? Our gas here in California is terrible on all my carburated vehicles but I don't experience any trouble with my injected cars.

Greg

 

Greg, the Corvette didn't show any problems, but I prefer not to use Ethanol, knowing the damage and rust issues which will ensue if left in the system. Our 1988 BMW 528e had Ethanol left in the system for a moderately extended time, destroying the tank, pump, steel lines and all six injectors - and you don't want to know the cost of parts and labor to replace all of that stuff, especially at BMW prices. Ethanol fuel allows water vapor to condense into the fuel and to precipitate the rust issues - think $,$$$.$$

(enjoyed your Kaiser last spring)

[mikewest]

On the 1929 Franklin I drive the most I add 2-3 gals of diesel to a full tank of gas and it corrects the problem, It lowers the flash point of the fuel. The car performs just fine and no more vapor lock ... Try it. Ive also used a inline electric pump that I use only for starting after the car has sat a long time (eliminates the need to crank a long time to get gas to the carb.) and when the car hiccups from vapor lock , I flip the pump on for a few seconds and it clears out the line. Im sure the heat shields help and its a good idea.

[DavidMc]

It is another common misconception that vapor lock is caused only by the ethanol added to the fuel.  There is no doubt that ethanol increases the problem,  here in Australia  we have the option of fuels with 10% ethanol and we avoid it it in our old cars. however in hot weather, vapor lock can occur even when running with non ethanol fuels .

 

Vapor lock has been an issue from the days of early motoring , consider this opening paragraph from a letter from the Packard Motor Company dated July 31 1931 addressed to its dealers and distributors  Quote:

 

'"Technical Letter 1895 dated March 17 1930 reviewed the gasoline situation and outlined the steps which had been taken to prevent the development of vapor lock in the fuel lines and the carburetor.  ...................."  The letter goes on to describe further changes recommended to avoid the vapor lock.   Incidentally those changes may well have worked in 1931 but with increasingly volatile fuels  they are not entirely effective now. 

 

The solution is to lower the temperature of the gas in the carburetor however that can be easier said than done.

 

.

 

 

 

 

 

 

 

 

 

 

 

 

 

[Grimy]

A few years ago, Skinned Knuckles magazine stated that the boiling point of 10% ethanol at 3-5 psi was 182*F underhood temperature at or near sea level.  Today's gasoline is formulated for EFI engines with fuel pressure of more than 35 psi, sometimes substantially more.

 

For those of us with updraft carbs especially those fed by gravity from vacuum tanks, it's even worse:  They had problems with gasoline being not volatile enough in winter, hence exhaust downpipes at the front of the engine where the fan would blow heated exhaust air at the carbs, and the variety of hot-spot manifoldings.  No summer/winter blends then.

 

In 1943 or so, Cadillac changed the design of their fuel pumps for 1937-forward V8s due to vapor lock issues.  Those engines' manifolds are on top of the heads and today must usually be shielded.

As mentioned, we need to reduce the temperature of the fuel in the engine compartment by shielding fuel lines and/or carbs from sources of heat.

[DavidMc]

Shielding is good for reducing radiant heat transmission  however  it does nothing to prevent the convective heat transfer from the hot engine bay and this might be the greater cause of the problem.  I have experimented with a heat shield that  fully encapsulated the vacuum tank and it seemed to make the problem worse and I suspect this was because the shield was  actually retaining heat.  There are no easy solutions, In hot weather I carry a water bottle and rag to wet the vacuum tank and carburetor which gives instant cure but not for very long.  On a very hot day here it can be a PITA.

 

[Grimy]

DavidMc, I agree with the convective concept in theory but have had excellent luck with a quilted-style heat shield around the vacuum tank on a 1925 Pierce 80, but the shield covered the bottom of vac tank and the sediment bowl as well as the vertical circumference.  I used velcro and a single large-diameter gear-type clamp (for HVAC ducts) to permit easy removal for show and not-quite-so-easy reinstallation for the ride home.  I also insulated the downpipe from the vac tank to the carb with a long piece of 1/2-inch heater hose split lengthwise.  Depending on the particulars of the individual car, your mileage may vary...  :-)

[DavidMc]

Sounds like your Pierce and my Packard have the same issues.  The last heat shield I made was from mirror finish stainless steel for maximum radiant heat reflectivety.  However I have been looking at the heat shield material as used on modern cars so based on your experience I think I will give that a try.  My shield covers the sediment bowl and it is interesting that  Packard recommended the removal of the glass bowl  filter to avoid vapor lock, in their 1931 letter, that i referred to in post #23 above.

I have thought about insulating the fuel line from the vacuum tank to the carburetor and could not decide whether it would  hold heat or reduce transmission of heat.   It is heated from radiant heat from the engine but it is also heated by the hot air.  The  insulation will not prevent 100% of convective heat  transfer so it could make it worse.  Its all a bit trial and error and I agree that the downdraft carburetor fed by the vacuum tank is good for warming the fuel in  very cold weather but  the opposite in hot weather.

Another solution is an electric pump bypassing the vacuum tank but that is not a perfect solution.

One right out of left field:  what about fully shrouding the vacuum tank then with a small  (concealed) CO2 bottle inject a small gas flow into the cavity between the tank and the shield to seriously cool the tank.  Works with beer coolers!  Might work but  I can think or more difficulties than solutions!  One of the above posts refers to another method of cooling and I am told this is widely used in racing and drag cars to cool the fuel.   Keep thinking!!

[Rusty_OToole]

Volatility is related to vapor lock and is also related to octane. Octane is one of the chemicals that make up gasoline and was chosen for testing purposes because it is very resistant to knock. But when I say octane I am referring to its common meaning, the knock resistant qualities of the fuel.

 

So briefly, those elements of the fuel that are light and ethereal  are more resistant to knocking but contain less energy than the heavier, oilier elements. They also boil at low temperatures and vaporize easily. This is what I meant. I am quite aware that "octane" can also be increased by adding certain chemicals that have nothing to do with petroleum, like tetraethyl lead.

 

But, my answer still stands. Today's gas is formulated for cars with fuel injection. They do not have a vapor lock problem because the fuel is always under pressure. Quite the opposite, they must vaporize the fuel in a split second between the time it leaves the injector, and travels a few inches into the combustion chamber.

 

Old cars especially those of the twenties were built with 'hot spots' in the intake to vaporize the oilier fuels of the day. They were deliberately built with a long stroke and small bore to better utilize the slow burning fuel. Adding kerosene gives you something closer to what they were made to run on.

 

Back in the day there were conversions to make low compression, long stroke engines run on straight kerosene. They had to be warmed up on gasoline, and special provisions made to vaporize the kerosene but many tractors trucks and industrial engines ran on straight kerosene, as they were made to run on low octane fuel (40 - 50 octane) that was not much better than lamp oil to begin with.

Edited November 9, 2017 by Rusty_OToole (see edit history)

[Grimy]

DavidMc, I neglected to mention that a key element of my modifications was to apply exhaust wrap to the exhaust downpipe (at front of engine) from its flange at the manifold to where it meets the splash pan. This minimizes the hot air being blown at the carb bowl.  However, it is not practical to remove the exhaust wrap for show.

[fraso]

If the vapour lock is occuring in the carburetor, I think much of the heat gained by the carburetor comes through its base from the intake manifold.  I think it would be more effective to add an insulating spacer between the carburetor and intake manifold.  This often done with modern hot rods with layers of gaskets (like Edelbrock PN 3710) or a phenolic spacer (like Holley PN 108-37).  I think many people have used wooden spacers effectively too (like Edelbrock PN 8723).  It is important to maintain the intake manifold hot spot to vapourize any atomized fuel that falls out of the airstream and this is necessary even in hot summer weather.

 

Vapour lock can also originate at the fuel pump.  In this case, an electric pump mounted in a cool area away from the engine can push liquid gasoline to the carburetor while the engine is cranked.

[PFitz]

If it is vapor locking at the carb, often the updraft carbs  of the vintage  of Wilbur's car - that had an atomizing hot-spot in the manifold near the carb mount - had a heat block gasket between the carb throttle body mounting flange and the intake manifold flange.  Often those original type mounting gaskets are missing and replaced with a thin gasket made out of modern gasket paper.

 

The original carb mounting gaskets had a thick layer of asbestos sandwiched between thin copper facings, but the copper did not extending into the bore of the gasket like the copper/asbestos/copper exhaust gaskets.   They really do help shield the throttle body from conduction heat from an intake manifold hot-spot.  Just one of the many little things that can contribute to vapor lock.

 

Otherwise, with a low boiling point gas that some states have, any part of the fuel system that is too hot to leave your hand on it is very likely contributing to vapor lock.   Fast way to double check is when the car starts to run rough when it's hot, wrap the suspected part of the fuel system in wet paper towels and if the engine starts to run smoother, there's your too-hot part.

 

When following hot weather car tours I always bring a quart bottle of water and a roll of paper towels, and some old kitchen towels. Anytime a car on tour dies from vapor lock I wrap the fuel pump and carb in wet towels.  It's a fast way to get that car back on the road.  For reoccurring case, I have them leave the rear hood latches undone and put a rolled up kitchen towel between the rear edge of the hood and the cowl to create a gap on the vacuum tank side of the engine bay. That lets air move through that area at the top of the firewall where hot air tends to get trapped  around the vacuum tank. You only need about a one inch gap to let a lot of trapped hot air escape from there. It does make a difference on hot days.  

 

Paul  

[emjay]

The compounding issue here is the use of one term, vapor lock, to all cars with different fuel designs and different symptoms of the vapor lock, yet think one solution will work for all.  Any discussion must clearly define the fuel delivery design and when the vapor lock occurs.  For example some owners only experience at start up with a warm engine bay, some only experience it while idling in traffic, and some simply experience on a hot day.  Each case requires different approaches.  Very important issues are where is the pump (if any) and the routing of the lines.  The tools available are temperature and pressure.  The suction side of the pump is very susceptible to vapor lock and the most debilitating since the pull of the pump lowers the pressure of the fuel and if the line passes hot surfaces, there is a double whammy and there will be little to no fuel flow.  Any place else on a vehicle with a fuel pump should be able to push the fuel through with the vapor venting out at the fuel bowl or carb., but you should be on your way soon.  Of course there just may not be enough air flow to cool the air around the fuel line enough until you are up to speed which you can't do because there is not enough fuel.  Sorry I'm rambling, but the point is understand where, when, and what your vaporization is and define your system when discussing your problem and your solutions.

 

And most importantly consider when you filled up.  As mention older vehicles will probably benefit better with some of the home blends mentioned  than later cars but still with carbs.

http://www.popularmechanics.com/cars/a3180/summer-blend-vs-winter-blend-gasoline-whats-the-difference-13747431/   

[60FlatTop]

On 11/7/2017 at 10:09 PM, wilbur said:

I had fit a heat shield at the onset, but eventually, after multiple roadside "sit-and-cools-offs" , made a much larger shield and now she runs along fine.  

 

Put the first heat shield back on. It will probably run fine. Whatever you jiggled into its "happy" spot inside the carb is probably OK now.

 

I am not much of a believer in vapor lock. Each time I see it mentioned I remember being about 13 or 14 years old and seeing a flathead '50 Ford with a bunch of clothespins on the fuel line. At the time I didn't know the three adult words of exclamation, but the image is in my mind and I, conceptually knew what I utter freely now.

 

Today I can take a piece of paper and pencil, write down the temperature/pressure relationships and calculate "it ain't gonna happen here" in technical terms.

 

Of, the '50 Ford with vapor lock problems, after about 10 minutes of cranking and trying to start it the oil filler cap blew off and hit Mike in the forehead. That was when the fumes in the crankcase ignited.

 

A vacuum tank needs to be under a vacuum to draw fuel and at atmosphere to feed the carb. How are you proving that transition takes place correctly? That would be my first test.

 

Bernie

[DavidMc]

Grimy,  My exhaust pipe from the manifold is wrapped and that was standard.  

Referring to other comments there is no doubt that vapor lock issues differ widely between different types of cars, all of my experience and comments are based on a vacuum tank and updraft carburetor which is possibly the worst arrangement for vapor lock to occur.

As for not believing in vapor lock because some one thought they could over come the problem with a few clothes pegs (which might have had some minor beneficial insulating effect), consider the case of a car that starts and runs well most of the time then in hot weather or during a long steep climb it starts to miss and splutter,  wetting the vacuum tank and carburetor instantly cures the problem for a short while until it warms again. If that is not vapor lock (or percolation etc whatever you want to call it)  what is it?   Furthermore I can hold my ear to the vacuum tank and hear the fuel boiling, so it acts as a "pre-heater" with further heat gain in the fuel line and carburetor. Once its boiling in the carburetor float bowl the fuel flowing to the venturi in the carburetor will be a mixture of vapor and liquid which causes the erratic running prior to it stopping.  A wet rag  temporarily condenses the vapor and the car runs normally for a short while   

To fully understand what is happening it would be necessary to measure the temperature and pressure through out the whole fuel delivery system under various conditions and it would also help to have all fuel delivery components transparent so we can see what is happening..  This is all obviously not practical so we persist with trial and error .    

Shielding and insulation is helpful but what is really needed is cooling, and to that end I have run my car with the hood sides partially raised so that air flows into the engine bay bypassing the radiator - that helped.

 

[Grimy]

DavidMc, thus far shielding has worked for me.  I've had similar symptoms.  On a few occasions, I've placed a new roll of paper towels, still in its cellophane wrap, on the hood side's sill, then bungeed or otherwise lashed down that side of the hood--but that was in 95+*F weather at 5,000 ft elevation.  That opened an air passage at the front and rear of the hood.  I've also, on occasion, removed the splash pan on the manifold side of the engine to help hot air escape the engine compartment.

 

I agree that scientific measurements are impractical, and that trial and error is the method to follow.  If you find a cure for your issues, please share it.

[maok]

Why not just use an electric fuel pump? Problem solved.

 

Biggest issue for me with engine compartment temperatures is the power loss from inducing hot air. Its a substantial difference between summer (even winter) day and night power levels.

[DavidMc]

Grimy,  My car is fine under almost all conditions now, I just keep looking for further improvements.

maok. I agree an electric pump is a good solution, but it should only be used when needed to overcome vapor lock, the lowest pressure pumps available are  a bit high for the float valves in carburetors that were only intended to be fed by gravity from a tank little more than 12" above the carburetor  I have tried that solution on a car that I have since sold .  I set it up so that with the aid of small one way valves it could be switched on to feed the carburetor without back feeding either the vacuum tank or main tank.  The one way valves were not very effective because of the negligible differential pressure across them so the present owner removed them and used manually operated valves.  That works OK but means stopping to set the valves when needed  which is certainly better than stopping every few minutes to cool the fuel system with a wet rag.  It was a good solution and its essential that the pump bypass the vacuum tank.  It would also be a useful back up if the vacuum tank failed although I have never experienced that. A further  improvement would have been to install a return line to the fuel tank so that the fuel was circulating giving it less time exposed to heat.   

I plan to fit an electric pump to the current car.  

[Guest]

3 hours ago, 60FlatTop said:

 

 

 

Of, the '50 Ford with vapor lock problems, after about 10 minutes of cranking and trying to start it the oil filler cap blew off and hit Mike in the forehead. That was when the fumes in the crankcase ignited.

 

Boy, does that statement take me back! Experienced the same with a small block Chev in the '70's. I was in standing Mike's position while Jim cranked the engine. , when it finally blew, the oil cap missed my forehead by about a quarter inch and sailed across the street, both valve covers blew off as well. A crank case full of fuel makes an exceptional little bomb.

 

Edited November 10, 2017 by Guest (see edit history)

[Grimy]

DavidMc, that's an interesting idea in your post #37 about a supplemental electric pump bypassing the vac tank.  Personally, I don't want to run even a regulated-down (to maybe 1 psi) electric pump *IF* that pump feeds the vac tank itself.  And I stay away from the small hockey puck regulators entirely.  On non-vacuum-tank cars, I use Holley regulator no. 12-804 (1-4 psi) to avoid overpressurizing the needle valve and seat.  Could you provide photos or a diagram?

[maok]

A fuel pressure regulator fixes that problem very effectively, as Grimy has noted. I run an electric fuel pump (Facet) with regulator (Speco) here in sub-tropical Brisbane all year round with no issue of vapor lock in my 28 Chrysler with a updraft Carter BB1 carb. Which would have been fitted with a vacuum tank originally to a Stromberg carb.

 

Its an easy install and relatively discrete, even though its noisy, I have some re-assurance that there is fuel being feed to the carb with the rattle sound that can be heard from the back.