What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

O-290-D2/MA-3SPA Dies when heat soaked

snopercod

Well Known Member
This is my first post here. I built and own a Lancair 235 and I hope you won't hold it against me. The Lancairtalk Forum is no help at all with engine problems; Most of the posters there seem to be second owners who didn't build their planes and don't know much about what's up front. So I came over here - hat in hand - where my kind of people hang out.

When I first started flying my Lancair 235 in 2014, it was suffering from high cylinder temperatures on climbout. Part of the problem turned out to be poor baffling ? which I have since fixed - but also the MA-3SPA carb wasn?t producing enough takeoff fuel flow. My engine is a Don George high compression O-290-D2 rated at 140 HP (145 during takeoff) and the carb had the stock main jet for the 135 HP engine. I had read that a rough rule-of-thumb for takeoff fuel flow is 1 GPH for each 10 HP, which would have been 14 GPH in my case. Since I was only seeing 8-9 GPH on takeoff, I drilled out the main metering jet (in stages) to #38 (.1015). I also added two #57 ?aeration holes? in the jet to better equalize the fuel distribution between the front and back cylinders. Both of those improvements worked; I now see 11-12 GPH on takeoff and, in cruise, my JPI EDM-700 shows all four EGTs at the same temperature after leaning and a little tweaking of the throttle. That?s about as good as you can get with a carbureted engine.

But now, in the heat of summer, I?m experiencing another problem. During ground operations on a hot day, if tower has me wait in place for ten or twenty minutes due to traffic, my engine overheats and wants to quit on me; I?ve had this happen several times now. The last time was 2 days ago at KAVL (field elev. 2111?, density altitude 3,500?). My EGTs were bumping up against 400 and my oil temp was bumping up against 200 F. My electric boost pump was ON the entire time, putting out 6.0 PSI. I didn?t see any fluctuations in the fuel pressure which would have indicated vapor lock (my gauge is connected to the outlet of the mechanical fuel pump, just upstream of the carburetor). I was able to restart by pulling the mixture halfway out, and kept the idle up to around 1500 RPM to keep it running. I left the mixture out on the takeoff roll, and eased it back in after takeoff. The engine ran fine when it got some cooling.

So the problem seems to occur when the temperature inside the cowling is above 170 degrees. (Yes, I have a temp sensor in there ? a digital meat thermometer LOL! ) I read where 100LL starts to boil at 140F so if the carb itself got to 170F, that would be a problem. I guess I need to instrument the carb bowl and see how hot it?s getting. All my fuel lines are insulated with firesleeve. The gascolator, mechanical fuel pump and electric boost pump have no shrouds, but I put a blast tube on the mechanical fuel pump; It hasn?t seemed to help. In fact, my problem doesn?t seem to be vapor lock in the lines, but an overly-rich mixture when hot at idle. I have a Garmin Virb video camera mounted under the tail tiedown, so I have video evidence of black smoke coming out of the stacks when the engine is quitting on me. To me that says ?too rich? rather than ?vapor lock?, but I?m just a new guy at this stuff. I?m wondering if I drilled out the main jet too much. Well, that shouldn?t be the problem because, in theory, the main jet doesn?t even come into play until around 1200 RPM.

If anybody was wondering, my idle is set to 730 RPM (warm), 650 RPM (hot), and the idle mixture is set for a 50 RPM rise on idle cutoff, which I believe is correct for the 2,000? field elevation where I am based. I?ve also experienced the problem in Denver (6,000? elevation, 8,800' density altitude.) and in Tucson (2,650? elevation) on a hot day. I?m wondering if I should lean out the idle mixture regardless of what ?the book? says.

The exhaust stack in my tightly-cowled Lancair is probably only 4? away from the carburetor. Is the fuel in the float bowl beginning to boil a little causing the float to sink and flood the engine? Does the float level control the mixture to the idle circuit as well as the main circuit? Maybe I need some insulation between the exhaust stack and carb, or wrapped around the carb bowl?

I?m leaning (no pun intended) toward boiling in the carb bowl, but am open to all suggestions. Thanks in advance.
 
I have a 0-320 and lean the daylight out of it on the ground, just remember to push it to rich for take-off.
 
Thanks, wilddog. Your O-320 is pretty much like my O-290-D2 but with a MA-4SPA carb. I lean during ground operations, too, and my plugs look nice and tan. But two days ago I got stuck for 20 minutes at the hold-short line waiting for landing traffic. The engine heated up and quit on me. Have you ever had that happen?
 
My bet is fuel boiling inside the carb bowl, causing the float to sink and allowing far too much fuel into the main jet.
 
Kurt-- That's my best guess at the moment as well. Thanks. I guess my next task is to stick a temperature probe on the outside of the bowl to confirm. What I can't figure out, though, is that the main jet isn't supposed to kick in until >1,200 RPM.
 
I have a 0-320 and lean the daylight out of it on the ground, just remember to push it to rich for take-off.

Thanks, wilddog. Your O-320 is pretty much like my O-290-D2 but with a MA-4SPA carb. I lean during ground operations, too, and my plugs look nice and tan. But two days ago I got stuck for 20 minutes at the hold-short line waiting for landing traffic. The engine heated up and quit on me. Have you ever had that happen?

My bet is fuel boiling inside the carb bowl, causing the float to sink and allowing far too much fuel into the main jet.

Kurt-- That's my best guess at the moment as well. Thanks. I guess my next task is to stick a temperature probe on the outside of the bowl to confirm. What I can't figure out, though, is that the main jet isn't supposed to kick in until >1,200 RPM.

One thing I would like to add, Lycoming recommends having the Boost Pump on for high temperature ground operation. I had HIGH temperature issues during Phase I flight testing in SoCAL back in September 1997. That is how long I have known about using the Boost Pump on the ground in high temperature conditions. After almost 19-years, I have the boost pump on for almost all ground operation even after leaving SoCAL.
 
If the exhaust is close to the carb, I would put a heat shield on it to block radiant heat from the fuel system. It doesn't have to be fancy, I tested by using cut offs from steel 2x4 studs left over from constructing my hangar. Any bits of sheet metal that you can hose clamp or wire in place for testing. If you like the result, spend some quality time and build the nice one.
The mechanical fuel pump transfers heat from the engine block and the oil into the fuel.
One year I flew to the local summer fly in (in my T-18) with the oil cooler winter cover still on, and winter grade mogas in the tank. After sitting in the broiling sun all day, I started to get vapor lock on the return flight at about 3000 msl. I diverted to the nearest airport. Draining out the mogas and filling with 100 LL fixed the problem, but a few days later my brain finally put it together. the oil temp was 210F or better on the gage, so the mechanical fuel pump was hotter than usual, and the air on the backside of the engine was much hotter. Coupled with old winter grade mogas was the perfect disaster.
Maybe a blast tube to the carb? Unfortunately, blast tubes reduce the cooling air pressure drop across the engine cylinders, so too many blast tubes = poor cooling of the cylinders and oil cooler. It's a balancing act.
 
First I would pull the float and make sure it is in good condition, I come from a Holly and Carter background, but I have seen brass floats get a hole and them then partially fill with gas. Shake it, weigh it. Confirm that 6 PSI is not too high and pushing the float down and flooding the bowl, confirm float height. Fuel flow should be .5 to .55 Pounds per HP per HR. :D
 
There's no mechanism to positively "shut off" fuel from the main jet during ground ops that I'm aware of. Even with a (mostly) closed throttle plate, if the float is sinking and fuel can be pushed by the fuel pump's pressure right out the main jet during idle, the engine will suck it in.

Leaning aggressively when this first starts to happen and during all ground ops should help.

Just thinking out loud... Has anyone ever mounted an electrically-operated blower/fan to force air blowing on the carb bowl during ground ops? Would it even help, since it would be using air from inside the lower cowl, whose elevated temperature caused the problem in the first place?
 
Last edited:
When the fuel boils, it can cause the float to sink a bit which allows the fuel level to rise. All of the mixture in a carb is based upon the fuel volume/height in the bowl. This can cause you to run rich, even with proper idle mixture settings.

The fix is a fuel return circuit. You can do some research here. When I had the carb setup on my 320, I ran a fuel return. I had a tee at the carb and added a -4 fitting with a restrictor (#40 drill size). This was fed to a valve in the cabin and then back to the tank. I never had issues with heat on the ground when I used this. It ensures that the fuel going into your carb is cool and helps to keep the boiling down or two a minimum. The boost pump is run to keep the flow up.

The restrictor insures that even if you leave it open on take-off it will not reduce your flow to the carb below the required flow.

Larry
 
When the fuel boils, it can cause the float to sink a bit which allows the fuel level to rise. All of the mixture in a carb is based upon the fuel volume/height in the bowl. This can cause you to run rich, even with proper idle mixture settings.

The fix is a fuel return circuit. You can do some research here. When I had the carb setup on my 320, I ran a fuel return. I had a tee at the carb and added a -4 fitting with a restrictor (#40 drill size). This was fed to a valve in the cabin and then back to the tank. I never had issues with heat on the ground when I used this. It ensures that the fuel going into your carb is cool and helps to keep the boiling down or two a minimum. The boost pump is run to keep the flow up.

The restrictor insures that even if you leave it open on take-off it will not reduce your flow to the carb below the required flow.

Larry

Im pretty sure until the rpms get above 1200 or so, the carb is feeding from the idle circuit. The mixture control has no effect when you are taxiiing around below 1200 rpm or so other than at cut off. That mixture is controlled by the idle mixture screw.

Cm
 
Lycoming recommends having the Boost Pump on for high temperature ground operation.
I do that, but it hasn't solved the problem. That leads me to believe that the problem is in the carb itself, rather than in the mechanical fuel pump or gascolator. My setup is (all inside the cowling):

Gascolator==>Facet Boost Pump==>FF sensor==>Mech Fuel Pump==> Carb.

My fuel pressure line taps off at the outlet of the Mech Fuel Pump so if...say...the Mech Fuel Pump was causing the problem, I would expect to see fuel pressure fluctuations on my gauge, which I don't.
 
Im pretty sure until the rpms get above 1200 or so, the carb is feeding from the idle circuit. The mixture control has no effect when you are taxiing around below 1200 rpm or so other than at cut off. That mixture is controlled by the idle mixture screw.
That's exactly what I thought as well, but couldn't figure out why, if the mixture control has no effect below 1,000 RPM, we're supposed to pull the mixture control out during ground operations to keep from fouling the plugs. Then I found this: Precision Airmotive's MSA MSA Float Carburetor Handbook. If I read this correctly, the mixture control DOES effect the idle mixture (as well as the idle adjust screw). Words just confuse me, but this link has lots of diagrams ;) I've highlighted the important stuff:
a. IDLE SYSTEM, (Ref. appropriate figure for a particular model). ? With the throttle fly slightly
open to permit idling, the suction or vacuum above the throttle on the manifold side is very high.
Very little air passes through the venturi at this time, and hence, with very low suction on the main
nozzle, it does not discharge fuel. This high suction beyond the throttle, however, causes the idle
system to function as the primary idle delivers into the high suction zone above the throttle. Fuel
from the fuel bowl passes through the mixture metering sleeve, fuel channel, power jet, and into
the main nozzle bore, where it passes through the idle supply opening in main nozzle, through the
idle fuel orifice in idle tube, where it is mixed with air which is allowed to enter idle tube through
the primary idle air vent and secondary idle air vent. The resultant rich emulsion of fuel and air
passes upward through the emulsion channel, where it is finally drawn into the throttle body
through the primary idle delivery opening, subject to regulation of the idle adjusting needle, where
a small amount of air passing the throttle fly mixes with it, forming a combustible mixture for
idling the engine.
The idle adjustment needle controls the quantity of rich emulsion supplied to the
throttle barrel, and therefore controls the quality of the idle mixture. Turning the needle counter-
clockwise away from its seat richens the idle mixture to the engine, and turning the needle
clockwise towards its seat leans the idle mixture. On idle, some air is drawn from the throttle barrel
below the throttle fly through the secondary idle delivery opening and blends with the idling
mixture to the engine as the throttle is opened, coming into play progressively and blending with
the primary idle delivery to prevent the mixture from beginning too lean as the throttle is opened
and before the main nozzle starts to feed. These Carburetors are provided with a third and, possibly
a fourth idle delivery in addition to the secondary idle delivery, depending on the application to
cover the broader idle range in these Carburetors.

b. METERING, (Ref. appropriate figure for a particular model). ? All fuel delivery on idle, and also
as steady propeller speeds up to approx. 1,000 rpm, is from the idle system.
At approx. 1,000 rpm
the suction from the increasing amount of air now passing through primary and secondary venturi
causes the main nozzle to start delivering, and the idle system delivery diminishes due to lowered
suction on the idle delivery openings as the throttle fly is opened for increasing propeller speeds,
until at approx. 1,400 rpm the idle delivery is practically nil, and most of the fuel delivery from
that point onto the highest speed is from the main nozzle.
However, the fuel feed of any full
throttle operation is entirely from the main nozzle. The idle system and main nozzle are connected
with each other by the idle supply opening. The amount of fuel delivered from either the idle
system or main nozzle is dependent on the whether the suction is greater on the idle system or
main nozzle, the suction being governed by throttle valve position and engine load. The main
nozzle feeds at any speed if the throttle is open sufficiently to place the engine under load, which
drops the manifold suction. Under such conditions of low manifold suction at the throttle fly, the
main nozzle feeds in preference to the idle system because the suction is multiplied on the main
nozzle by the restriction of the venturi.
 
If the exhaust is close to the carb, I would put a heat shield on it to block radiant heat from the fuel system.
Here's what it looks like. The carb heat muff seems like it would block most of the radiated heat, but maybe not. BTW, that redneck heat muff works GREAT! I get a 200 RPM drop during runup.

DnJFji.jpg
 
Has anyone ever mounted an electrically-operated blower/fan to force air blowing on the carb bowl during ground ops?
If it turns out that the carb is the problem, it seems like that would be an excellent idea. I'd use outside air, though. Maybe a cowl flap?
 
Fuel flow should be .5 to .55 Pounds per HP per HR.
That's about what I'm seeing now (12 GPH) and it seems about right. My CHTs now stay below 400 on climbout even at 90 KIAS on a hot day. Before, I was having to keep my airspeed up above 120 KIAS to keep the cylinders cool.

It's interesting to learn that drilling out the main jet like I did can also effect the idle mixture. (See that information I posted to Christopher).

I have the blue epoxy floats which are two years old. I guess I should check them and the float height anyway, though. More quality time laying on my back on the hanger floor :rolleyes:
 
neadle seat and float

This is just a guess, but what you are describing here sounds like a flooding problem after setting at ideal for a time. As has been mentioned above a weak float or needle and seat will let the fuel level in the bowl rise up to a level that will start over fueling at ideal if it dose not close off the fuel flow at its set point. Even the set point adjust angle of the float if set to high will do the same thing and allow to much fuel in the bowl. It is kind of like the float level in your potty, if you set it to high it will overflow into the stand pipe or if the shout off valve "needle and seat" dose not seal off well the level in the bowl will creep up to overflow also. If you turn the bust pump on and this is the case it will flood even more and make the black smoke even more. If you are not seeing any singes of vapor lock, I would start with a bowl gasket, needle and seat and a good new meatal float. And adjust the float to the correct level just as in the potty. Just an idea of what I would start looking at first from what you are saying here. Hope this helps, Yours as always, R.E.A. III #80888
 
Im pretty sure until the rpms get above 1200 or so, the carb is feeding from the idle circuit. The mixture control has no effect when you are taxiiing around below 1200 rpm or so other than at cut off. That mixture is controlled by the idle mixture screw.

Cm

I wasn't discussing the mixture control. I was referring to the fuel level, which is set by the linkage on the float AND influenced by the density of the fuel that this float sits on. This fuel level influences the fuel delivery at any given air flow state. The mixture control can only reduce the flow beyond what the jets have set as the maximum flow through a circuit (yes, the idle circuit has a jet too). They often flow less than what the jet allows, even with mixture at full rich.

The main, transition and idle circuits are all influenced by the fuel level in the bowl eventhough they are metered and adjusted separately. Fuel is pulled from the idle or main circuit via pressure differential. Systems are engineered to deliver different fuel flows based upon the differential pressure at any moment. In order for this to work, the fuel must have a constant pressure, which is set by the fuel level in the bowl and the atmospheric pressure.

Finally, the idle mixture adjustment is not absolute. If you look at your fuel flow when idling at 500 vs 1000, the fuel flows will be different (you are still on the idle circuit at 1000 rpm). It just influences the pressure differential that the idle circuit sees which is also influenced by the fuel level.

Larry
 
Last edited:
That's exactly what I thought as well, but couldn't figure out why, if the mixture control has no effect below 1,000 RPM, we're supposed to pull the mixture control out during ground operations to keep from fouling the plugs. Then I found this: Precision Airmotive's MSA MSA Float Carburetor Handbook. If I read this correctly, the mixture control DOES effect the idle mixture (as well as the idle adjust screw). Words just confuse me, but this link has lots of diagrams ;) I've highlighted the important stuff:

In simple terms, the mixture control is a valve that sits BEFORE all of the other fuel delivery circuits in the carb. All it can do is reduce total flow into the carbs various delivery circuits, allowing you to choke off flow and thereby over-ride the carbs metering function the the lean side.

You can use mixture to lean at idle. However, it is a very specific point in the travel and difficult to get to. If you lean properly with mixture control at 800, you will not be able to run at 1200 RPM, as you would have choked down the flow to a rate for 800, which would be too lean for 1200.

However, there really is no need to do this at idle, as this can be fully covered by the idle mixture setting. If the idle mixture is properly set, manual leaning can only make you idle TOO lean. If your idle is set too rich, manual leaning can fix it, but why do that? However, manually leaning will help during taxi as you will likely be getting on the transition circuit or even the main circuit and running rich without overriding it. The carbs tend to run on the rich side in that area to help with transition from the idle to main circuit.

Larry
 
Last edited:
Well this is embarrassing. While looking for a good place on the carb to attach a thermocouple, I looked up at the flange where the carb bolts onto the oil sump and said to myself, "What the heck is that hole doing up there?" It turns out that somewhere along the way, the "Idle Drill Plug Screw" had fallen out, leaving a .20" hole all the way into the barrel of the carb. I confirmed this by sticking the butt end of a #10 drill bit into the hole and observing it inside the throat of the carb, sticking out above the venturi and below the throttle fly. That can't be good, but I'm not smart enough to predict what effect it would have. You would think it would make the engine run lean, but it seems that adjusting the mixture control would compensate for the extra air. At idle, the throttle fly is mostly closed so it wouldn't make much difference; There's lots of air coming up from below anyway. All I know is that the replacement plug/screw is supposed to arrive tomorrow by FedEx. I'll know more when I install it (with Loctite!) and fire up the engine.

SUksbZ.jpg
 
Last edited:
Here's a photo of the replacement plug. $2.50 for the plug and $14.50 for shipping. I don't know if MA-4SPA carbs have one of these or not:

GtltyK.jpg
 
The part pictured above doesn't fit. It's threaded 1/4-20 and the hole in the carb is (apparently) 12-24. My carb is probably 60+ years old and somewhere along the way, Marvel Schebler changed the design. I'm going to have to build my own plug, apparently.

Side note: This hole is where a carb temperature probe goes, if you have one.
 
You can fill the hole with JB Weld. Just form a piece of thick tape on the inside of the carb before filling and make sure to form an oversized head on the outside and be sure to clean with acetone first for good adhesion.

Larry
 
Larry-- I'll use J-B Weld as a last resort, but I might want to install a carburetor temp gauge some day and that's the hole I would need to use. I found a 12-24 brass machine screw at Ace Hdwe. this afternoon and will try to plug the hole with that tomorrow.
 
Larry-- I'll use J-B Weld as a last resort, but I might want to install a carburetor temp gauge some day and that's the hole I would need to use. I found a 12-24 brass machine screw at Ace Hdwe. this afternoon and will try to plug the hole with that tomorrow.

I would put some red thread locker on it as well.

good luck.

Larry
 
I would put some red thread locker on it as well.
I plan to use the blue "medium, removable" since someday I will want to install a temperature probe in the hole. As an extra precaution, though, I'll drill a small hole in the 12-24 screw head and safety wire the screw. It ain't gonna' be pretty, but it should work.

Thanks to you (and everyone) for all the good advice. Hopefully today I can fire up the engine and see how it runs without the air leak :eek:
 
Yip! that will mess with the whole metering process, glad you found that! so did the problem just show up one day?
 
Yip! that will mess with the whole metering process, glad you found that! so did the problem just show up one day?
I don't know when that plug screw fell out, but I hadn't noticed any symptoms. A few days ago the engine was running just fine both at cruise and when taxiing. One would certainly think that hole in the barrel of the carb would mess things up, but maybe not so much. Today I plugged up the hole and ran up the engine. Amazingly, nothing much changed. The idle was smoother (natch!), and the idle mixture test gave me 100+ RPM rise at idle cutoff rather than the 50 RPM rise a few days ago. Those were the only differences on the ground. So I adjusted the idle mixture screw 1/4 turn in the lean direction, and re-performed the test. I got 50-60 RPM rise that time so I left it and put the cowling back on. I would have expected to have to enrichen the idle mixture to make up for the extra air coming in that hole, but that wasn't the case. Anyway, that problem is fixed, but I have no idea whether the heat soak problem is still there or not. This missing screw may have been just a detour on the road to finding the original problem. OTOH, maybe it has something to do with it. Time will tell. Here's my redneck fix. I used blue loctite on the threads and J-B Weld on the head in lieu of safety wire.

C05zLv.jpg
 
I made 3 trips around the pattern at KAVL this morning and didn't die (me or the engine)! Fuel flow on takeoff was slightly less (10.3 GPH) which, I guess, makes sense since there was less air going in through the hole. CHTs were around 375 and oil temps were in the middle of the green. OAT was 80F on the ground before takeoff. The intra-cowling temperature got up to 177 degrees after landing, with no tendency for the engine to die. EGTs were in the 1300-1400 range on climbout, full rich at 3,500'. It's still a mystery...
 
It was 87F today in Asheville so I flew once around the pattern to check things out. I was only held for 5 minutes at the hold short line and my oil temp was 185 F. Lower cowling temp got up to 192 F, CHTs were around 350 F. No problems with the engine running rough or wanting to die. This doesn't mean the problem has been solved, but it's going to be real embarrassing if that missing plug was the problem and it's been been missing for over a year and I never noticed. :eek:

Just before shutdown I performed an idle mixture test and got no rise at all with the engine hot. I think I'm going to leave it as-is for now.
 
Still haven't found the problem

The missing plug wasn't the problem because the engine still dies when hot on the ground, like waiting for 20 minutes in a que to take off. I instrumented the entire fuel system with thermocouples and the only area that's getting excessively hot is the carburetor itself. I added more insulation between the exhaust system and the carburetor, and that didn't help. So I'm wondering if the hot oil in the sump is making the carburetor bowl hot enough to boil the fuel inside. I've never heard of that being a problem in a Lycoming. Neither have I ever heard of anyone putting a blast tube on their carburetor, but maybe that's what I need to do. Actually, though, I'm not sure how well that would work on the ground. I'm really running out of ideas here. Anyone?
 
The missing plug wasn't the problem because the engine still dies when hot on the ground, like waiting for 20 minutes in a que to take off. I instrumented the entire fuel system with thermocouples and the only area that's getting excessively hot is the carburetor itself. I added more insulation between the exhaust system and the carburetor, and that didn't help. So I'm wondering if the hot oil in the sump is making the carburetor bowl hot enough to boil the fuel inside. I've never heard of that being a problem in a Lycoming. Neither have I ever heard of anyone putting a blast tube on their carburetor, but maybe that's what I need to do. Actually, though, I'm not sure how well that would work on the ground. I'm really running out of ideas here. Anyone?
Give me a call tomorrow afternoon, lets talk through this.
 
Thanks

Thanks, Bill, I'll give you a call later. And thanks to all the commenters on this thread - those are all good ideas. To recap:

1. The problem only occurs when idling on the ground when the lower cowling temperature gets above 180F. The probe is mounted just below the outlet of the oil cooler so the temp is a good proxy for oil temperature as well. (I don't have a calibrated oil temp gauge.)

2. All fuel lines are insulated with firesleeve.

3. Neither the gascolator, electric boost pump, or mechanical fuel pump are getting excessively hot. They're around 120F when the problem occurs, although the fuel is maybe 140F at the outlet of the mechanical fuel pump.

4. Turning on the boost pump doesn't help, which leads me to believe that bubbles are not forming anywhere upstream of the carburetor.

5. The initial boiling point of 100LL is around 210 F and it really starts to boil about 250F​

My current theory is that the boiling is occurring in the bowl of the carburetor and the source of the heat is probably a combination of the oil sump and exhaust stack. The only thing I could do to cool off the oil on the ground is add another SCAT duct from the baffles to the oil cooler plenum. I already have a provision for that so it would be no big deal, although I would need some kind of shutoff valve on the second line so the oil doesn't get too cool in flight.

I like krw5927's idea of a thermostatically-controlled fan blowing outside air on the carb bowl. I also lr172 's idea of a recirc line although it would mess up my fuel flow readings slightly. I actually already have the plumbing for that via an 1/8" aluminum line to the panel for my fuel pressure gauge. i already have a valve that would work.

Thanks again for all the good ideas here.
 
Thanks, Bill, I'll give you a call later. And thanks to all the commenters on this thread - those are all good ideas. To recap:

1. The problem only occurs when idling on the ground when the lower cowling temperature gets above 180F. The probe is mounted just below the outlet of the oil cooler so the temp is a good proxy for oil temperature as well. (I don't have a calibrated oil temp gauge.)

2. All fuel lines are insulated with firesleeve.

3. Neither the gascolator, electric boost pump, or mechanical fuel pump are getting excessively hot. They're around 120F when the problem occurs, although the fuel is maybe 140F at the outlet of the mechanical fuel pump.

4. Turning on the boost pump doesn't help, which leads me to believe that bubbles are not forming anywhere upstream of the carburetor.

5. The initial boiling point of 100LL is around 210 F and it really starts to boil about 250F​

My current theory is that the boiling is occurring in the bowl of the carburetor and the source of the heat is probably a combination of the oil sump and exhaust stack. The only thing I could do to cool off the oil on the ground is add another SCAT duct from the baffles to the oil cooler plenum. I already have a provision for that so it would be no big deal, although I would need some kind of shutoff valve on the second line so the oil doesn't get too cool in flight.

I like krw5927's idea of a thermostatically-controlled fan blowing outside air on the carb bowl. I also lr172 's idea of a recirc line although it would mess up my fuel flow readings slightly. I actually already have the plumbing for that via an 1/8" aluminum line to the panel for my fuel pressure gauge. i already have a valve that would work.

Thanks again for all the good ideas here.

I still struggle with a rough running idle when it is hot out with my Bendix FI. I believe that the unit is absorbing heat through the oil pan and without a solid flow of air to cool it off, it is getting hot enough to boil fuel. I can hear it boiling off down there right after shut down. The Oil temp is a bit misleading, as it measures the oil temp AFTER the oil cooler. I believe most of these coolers deliver a 25-30* delta T (i.e.temp drop) with air moving through them. Therefore a reading of 190 likely equates to a sump temp of around 220* Over time, without meaningful air flow, the carb or FI will quickly reach 220* They are made of aluminum and will therefore take the heat pretty quickly. The only insulator is the thin gasket, which I don't believe does much to slow it down.

Just my speculations here. I intend to install a 1/4" phenolic spacer this summer to see if I can slow down the heat transfer.

Larry
 
I believe that the unit is absorbing heat through the oil pan and without a solid flow of air to cool it off, it is getting hot enough to boil fuel.
That explanation fits the symptoms. Which do you think I should try first: Adding another SCAT duct to the oil cooler plenum to cool down the oil, or adding a SCAT duct blowing on the carburetor? I could try either one. Either would originate from the back of the baffles behind cylinders 2 & 4.
 
You want to be careful pressurizing the lower cowl. If you do that, you will have cooling issues in flight, which you currently do not have an issue with.

Your installation might be a good candidate for the AntiSplat cowl flap. That will help get the heat out on the ground and keep your CHT's happing in cruise.
 
Don?t touch that jet!

Hi Snopercod,

To start I'd like to tell you I really enjoy your flight trip write-ups.:) Keep up the good work!

Sorry about your engine woes...It's been really HOT about the country.

I just have a few things for you.

Excerpt from the carburetor link that follows...

""Don?t touch that jet!
A restriction at the base of the discharge nozzle serves as the main jet in MA series carburetors. In HA-6 (horizontal side-draft) series carburetors, the power jet in the base of the bowl performs the same function. Some well-meaning technical writers have instructed mechanics to hone these jets in an effort to cool cylinder head or exhaust gas temperatures. Apparently they believe this to be a ?cure-all? for lean running engines. A mechanic would be ill advised to follow such instructions since no process approval exists for opening these jets in the field. Furthermore, tampering with the jet size could mask other serious problems. An induction leak, a missing carburetor body sensor plug... O.K. I added that one.;), an incorrectly sized economizer jet, an incorrect float setting, or perhaps even the wrong choice of carburetor could all contribute to a ?lean? running engine. Another problem with this mode of attack is that the jets vary in design. Some jets are straight, while others are contoured or stepped. A mechanic who takes a drill bit, a ream, or sandpaper to the entrance of a stepped opening may soon regret that decision. Fuel flows could exponentially increase by merely breaking the stepped edge and thereby creating a venturi entrance. By contouring a stepped jet we?ve managed to decrease the pressure and increase the velocity of the fuel through the jet. When you start boring out nozzles, several things can happen. As you begin to see appreciable results ? the tendency is to take out more and more material. ?If a little bit is good, a lot may be better.? Wrong""...

Please read the rest in this informative article HERE.

Sounds like it touches on a few of the symptoms your experiencing...

Secondly, new jet not in your future?

Go HERE...And look at "Accelerator Pump Adjustment" pages 1-5 and 1-6.
Perhaps the high ambient temperature setting would help a little?

Couple more things... when you did your instrumentation. Did you insulate your carb body (outside of the bowl chamber) temp probe from the ambient air? Armorflex insulated tape works well...and what's the oil sump housing temp? Lastly, (whew) is your carb heat damper shutting off completely?


Respectfully submitted,
 
Don?t touch that jet!
Too late :) When anguishing over what to do to get my takeoff fuel flow up, I read all the information I could get my hands on (including the excellent link you provided). The usual suspects said "Don't touch that jet" and maybe they were right. But many others have had success with drilling out a jet. After all my research, I decided to use my best "It's experimental" judgement and drill out the jet. Doing so brought the fuel flow into the normal range, and adding the "aeration holes" (AKA "peppermill") solved the uneven fuel distribution problem. There's always the chance, though, that those mods introduced some other problems like I am experiencing.

Years ago I adjusted my accelerator pump linkage to the dreaded "hole 1" - the minimum setting. I did that based upon video evidence showing a puff of black smoke coming out my stacks when I pushed the throttle forward during the takeoff roll. Hole 1 seems to be right for me because I never experience any kind of hesitation when adding throttle.

No, I didn't insulate the carburetor bowl, and no, I didn't insulate the thermocouple from the ambient air. It was just a quick and dirty check to see what was getting hot.

What I did yesterday was to double the air flow to my oil cooler by adding a second 2" SCAT duct to the inlet plenum. I had it that way when I first started flying, but the oil was staying too cool. Well, I think it was too cool...I really need to calibrate my oil temp gauge to be sure; I have the test equipment to do that (Fluke 87V) so I'll put it on my to-do list.

I'm glad you are enjoying the trip writeups. It's fun flying around with my RV buds. We should have an RV-10 join us pretty soon when Jason gets out of Phase 1.
 
That explanation fits the symptoms. Which do you think I should try first: Adding another SCAT duct to the oil cooler plenum to cool down the oil, or adding a SCAT duct blowing on the carburetor? I could try either one. Either would originate from the back of the baffles behind cylinders 2 & 4.

Cooling the oil won't help the issue of boiling fuel in the bowl, if that is actually your problem (way too much heat to get rid off and too little air flow during ground ops). I would also be looking at vapor lock in the lines or pump. Fuel boiling in the bowl causes the float to drop a bit, allowing the bowl to overfill and create a rich condition. Fuel is metered by an interaction between the negative pressure in the venturi and the positive pressure imparted on the fuel due to it's weight/vertical volume (higher fuel levels in bowl create a higher pressure). However all fuel is routed past the mixture gate and can therefore be over-riden. If you can get the mixture knob adjusted for the exact fuel flow needed at your idle condition (varies based upon RPM, temp, etc.), it will be immune to an over-filled bowl and run fine (you have fixed the fuel flow manually and no longer requiring the pressure-based variable metering). Just be sure to give it a good run up before take off to stabilize the fuel in the oowl. Also, your engine won't run more than about 100 RPM greater than idle, as there won't be enough fuel flowing past the gate to support it.

Vapor lock in the lines, gascolator, pump, etc. will starve the carb and create a lean condition (carb can't replenish the bowl and level goes down, leaning your mixture). The only fix for this is to utilize a return system to keep the temps down. When I had a carb, I had a return line plumbed from a tee at the carb inlet, through and .040" orifice, back to a valve in the cabin and on to the tank. This keeps all of the fuel up to the carb cool when on the ground. I didn't think I would need this for 100LL, but knew I would need it if I ever used Mogas. The orifice keeps the flow low enough that there is still plenty of fuel flow available to give the engine 15 GPH if you forget to shut it off.

I would be doing some troubleshooting to determine if the engine stops are due to a rich or lean condition.

Larry
 
Last edited:
No, I didn't insulate the carburetor bowl, and no, I didn't insulate the thermocouple from the ambient air.

No, by all means DON'T insulate the Bowl!:eek: I was asking if you had just insulated the PROBE. To get an accurate reading of the bowl temperature. Inaccurate readings are worthless and waste time and effort. By all means get that oil temperature gauge calibrated!

Ahhh your way ahead of me. But I think we're on the same page. I'm just on the first sentence and you've read the whole thing...:)

And that being your picking up excessive heat from the oil sump. I'm with you as your trying to keep it at a lower temp. Had me goin there a second as you mentioned scat tube...What the...:confused: Thinking RV you know.

I suspect that you've eliminated the possibility of radiated heat from the exhaust piping being an issue. And the ambient temp in the cowling should actually be helping cool, same with the fuel. So what's left.

Is the carb heat run damper leaking hotter air into the intake?

And just a last thought.

Shouldn't that nifty evaporative cooler located in the middle of the float bowl keep temps down? Perhaps if the fuel was atomizing closer to the middle of the carb?



Regards,
 
Last edited:
Is the carb heat run damper leaking hotter air into the intake?
I'm pretty sure that's not the case. We Lancair guys had to build our own airboxes (Can you believe that???) and I modeled mine after one off a Piper Pacer. The ram air comes in the front and the heated air comes in the back with a "butterfly" damper in between to direct the chosen source up into the carburetor throat. There's really no way the heated air could "leak" into the carb unless the linkage is moved.

Here's a photo of the two SCAT tubes feeding the firewall-mounted oil cooler plenum:

ecfMgp.jpg
 
I would also be looking at vapor lock in the lines or pump.
I've eliminated that possibility in my (alleged) mind. First, turning on the boost pump doesn't help. Secondly, I see no fluctuation in fuel pressure when the engine quits. Thirdly, I measured the temperatures of the gascolator, boost pump, mechanical fuel pump, and carburetor bowl with the engine hot. Only the carburetor bowl was over 120 degrees - it was up around 180 F. At that time, the air exiting my oil cooler was also 180F, so it makes sense that the heat source here is the hot oil.

I'm sure it's a rich condition because I have video evidence of black smoke coming out of the stacks when this problem occurs. Also, I can restart the engine only by pulling the mixture control way out.

FWIW, I did recheck the idle mixture today with the engine just warm - not hot. I got a 30 RPM rise just before it quit. I think that's in the ballpark for a 2,000' airport. I'll perform the same check again next time the engine is hot.

I appreciate your thoughts.
 
I've eliminated that possibility in my (alleged) mind. First, turning on the boost pump doesn't help. Secondly, I see no fluctuation in fuel pressure when the engine quits. Thirdly, I measured the temperatures of the gascolator, boost pump, mechanical fuel pump, and carburetor bowl with the engine hot. Only the carburetor bowl was over 120 degrees - it was up around 180 F. At that time, the air exiting my oil cooler was also 180F, so it makes sense that the heat source here is the hot oil.

I'm sure it's a rich condition because I have video evidence of black smoke coming out of the stacks when this problem occurs. Also, I can restart the engine only by pulling the mixture control way out.

FWIW, I did recheck the idle mixture today with the engine just warm - not hot. I got a 30 RPM rise just before it quit. I think that's in the ballpark for a 2,000' airport. I'll perform the same check again next time the engine is hot.

I appreciate your thoughts.

If you're sure it's a rich failure, boiling in the bowl is most likely with your symptoms. Try manually leaning at idle next time it's hot and you're waiting for a while. lean until you get some movement of the RPM. That should keep it from dying. You can also wait at 1100-1200 RPM. That will move more fuel through the bowl and keep it cooler. The engine will also be more tolerant of poor mixture at that RPM.

I like to adjiust idle mixture via the highest vacuum/lowest MAP method. It's more accurate. Given that it is taking 10 minutes to die, idle mixture is not the core problem.

Larry
 
You can also wait at 1100-1200 RPM.
Yes, if I'm in a que, I lean aggressively and keep the RPMs up around 1400. It's a tradeoff, though, because the higher RPMs cause the CHTs to climb up faster. Lately, I've been thinking, "Why not just shut off the engine?"

Could you explain the manifold pressure method of checking the idle mixture, please. I've never heard of that.
 
Have you done a mag check immediately after the hot engine die-out and subsequent restart?

If so...and they're both good.

How 'bout a play by play description of the engine running conditions leading up to the quitting? Fighting to keep it running? Rough running, hesitations, stumbles, rpm fluctuations, etc.
 
Have you done a mag check immediately after the hot engine die-out and subsequent restart?
No, I didn't think of that. I'll do one next time.

As for the sequence of events, the engine just dies. I lean aggressively to restart and it runs rough until I get the RPM up above 1,500. I have video of the latest event, but thanks to Garmin updating my editing software so it no longer works, I can't post it. I do have this old video I took in Denver two years ago. This was before I learned to lean aggressively when restarting:

https://www.dropbox.com/s/ay7prqm4i68uvz2/Engine quit at Centennial.mp4?dl=0
 
Yes, if I'm in a que, I lean aggressively and keep the RPMs up around 1400. It's a tradeoff, though, because the higher RPMs cause the CHTs to climb up faster. Lately, I've been thinking, "Why not just shut off the engine?"
Not just lean agressively. Let's say your going to idle at 800 RPM. Set your throttle to 800 RPM. Once you get near the point on the mixture pull, slow down and lean until your RPM goes up by about 20-30 RPM and stop. At this point, your carb cannot deliver more than the current fuel flow, regardless of an overfilled bowl. Can't say what's going to happen when it's time to take off, but it will prevent you from going rich at 800 RPM if the bowl over-fills. You'll probably have to nurse it a bit to clear the extra fuel once your ready to advance the throttle.

Could you explain the manifold pressure method of checking the idle mixture, please. I've never heard of that.

An engine will pull the highest vacuum when it's mixture is at it's optimum level. It is a means of judging peak efficiency at idle. Highest vacuum is the same as saying lowest MAP. Warm your engine up and keep adjusting the mixture until you achieve the lowest MAP possible for a target RPM. The RPM will go up or down with each adjustment, so you'll need to adjust it back to your benchmark RPM. It is the lowest MAP for a specific RPM, not just lowest MAP overall, that you are shooting for. Remember that optimum idle mixture is temp dependent. If you plan to do this again in late fall go optimum. If you are setting a year round mixture level, set it a bit on the rich side so that you are not too lean in the winter when the air is denser and requires a higher fuel quantity to achieve the same mixture.

I use this method to set idle mixture. I do not not manually lean on the ground and have never fouled a plug in 350 hours. Manual leaning is completely unnecessary below about 1000-1100 RPM if your idle mixture is set properly. Below this level, you're running completely off the idle circuit in which you have complete control over the mixture via the idle mixture screw. Further, manual leaning via the red knob is simply not possible at anything other than at a static RPM. Once set properly, your idle mixture circuit will provide a constantly variable flow that meets your mixture setting from 500-1000 RPM with no manual intervention. Feel free to ignore this aspect, as most here consider me a heretic for saying it.

Larry
 
Last edited:
Back
Top