IO-360-A1A engine driven fuel pump options

[Vance Harral]

Finished a flight yesterday and observed fuel dripping from one of the hoses that exits the left side cowl flap (not from aux pump drain, that's further back).  As the "streaming" blue stain on the nose gear door in the attached photo shows, there's evidence it dribbled in flight and streamed back along the gear door (there's a little oil there too, but that's normal for us).  Haven't uncowled the engine yet, but I'm 99% sure from memory this is the hose that goes to the engine-driven fuel pump's overboard drain.  As I understand it, this means the internal seals and/or diaphragm of the pump are failing, and it's time for a new or overhauled unit.

My question to the hive mind is about part numbers.  The application guide at https://www.aircraftspruce.com/catalog/pdf/TEMPEST-FP-APP-LYC.pdf  has three options for the fuel pump in an IO-360-A1A, part numbers  41234, 40296, and LW15473.  Researching these part numbers suggests variations involving "channel arm", "laminated arm" and "dual diaphragm".

Anyone care to render opinions on why we should choose one part number over another?  All three flavors appear to be readily available in new and overhauled forms.  Some are more expensive than others.  The cost difference isn't an issue for us, but I want to understand the pros and cons before engaging with our mechanic.

 

[takair]

Vance

There is an older thread on the topic. One of the available pumps tends to put out over red line pressure. Lycoming says it’s ok but it is outside Money spec.   I’ll see if I can find the thread. 

[takair]

Maybe this thread will be helpful. 
 

 

[Vance Harral]

Thanks very much for the pointer to that thread, @takair.  It's regrettably disappointing in two respects.  One, it doesn't seem to answer the questions of whether an LW15473 that produces over 30psi is actually mechanically problematic, vs. only legally problematic, and how/if one can ensure getting a unit that doesn't exceed the legal spec.  Two, it documents that the cost of the pumps has nearly doubled in the last 3 years. 

Aside from pressure regulation, I'd still like to understand what's supposedly better about the LW15473 that justifies a higher price, given that the other flavors are readily available. 

[MikeOH]

FWIW, and you probably are already aware, but only the LW15473 is shown in the Lycoming manual:

 

[Ragsf15e]

I had to replace mine about 4 years ago (on my F) due to fuel leaking out that drain. Unfortunately I don’t remember which exact part number we used, but it did put out slightly too high pressure for the first ~50 hours of its life.  Maybe 32psi or so?  It settled around 28-29 after a while and worked great.  The lycoming spec is something like 60psi max for the engine, it’s just Mooney that limited it to 30.  We have the documentation for that around here somewhere…

[takair]

1 hour ago, Vance Harral said:

Thanks very much for the pointer to that thread, @takair.  It's regrettably disappointing in two respects.  One, it doesn't seem to answer the questions of whether an LW15473 that produces over 30psi is actually mechanically problematic, vs. only legally problematic, and how/if one can ensure getting a unit that doesn't exceed the legal spec.  Two, it documents that the cost of the pumps has nearly doubled in the last 3 years. 

Aside from pressure regulation, I'd still like to understand what's supposedly better about the LW15473 that justifies a higher price, given that the other flavors are readily available. 

I can’t answer why 15473 might be mechanically better.

From an operational perspective, I first noticed the issue in starting, I seem to recall my mixture being too rich, but I am not sure that is technically possible given the fuel injection system…but that’s what I recall.

If you have an engine monitor with fuel pressure, you will set off the alarm all the time.

From a TC perspective, you will be outside the Mooney limits.  I had notified Mooney at the time, they didn’t say much.  Maybe I’m anal about limits, but it seems that going over the airframe limits is pretty clear cut, even if Lycoming says it is ok,.  They don’t set the airframe limits…. If Mooney were still building airframes they might have some leverage with Lycoming or they would possibly do the analysis or testing to broaden the limits..  Vance…I know you know all of the legalities and you know the Mooney factory better than me…..I’m just kind of venting now….lol…old wounds…

[PT20J]

That thread is 10 years old. I purchased a rebuilt IO-360-A3B6 from Lycoming in 2018 and it came with a LW15473 pump and it puts out about 27 psi. (BTW, the RSA manual states that the fuel injector works fine with up to 10 psi greater than whatever the aircraft manufacturer specifies for a maximum).

[N201MKTurbo]

The limiting factor for fuel pressure is the servo. At the maintenance symposium a few years ago I asked the engineer from Precision Airmotive how much pressure the servo could take and still meter the fuel properly. He said they don’t test that, but he estimated it would be around 75 PSI.

[N201MKTurbo]

I believe all the fuel pumps are dual diaphragms. With a single diaphragm, any leak will go into the oil and there is no drain line to indicate a leaking diaphragm. 

[Vance Harral]

5 hours ago, MikeOH said:

only the LW15473 is shown in the Lycoming manual

That's certainly a definitive reference, but the Tempest guide says thusly:

Another Tempest document has this data, in which the 41234, 40269, and LW15473 are seem to be listed as alternates for each other, albeit with "Channel Arm" vs. "Laminated Arm", whatever that means.

[Vance Harral]

I think one difference may be that if you want new, the LW15473 seems to be the only option.  The 40296 and 41234 are available only as overhauled.

[Vance Harral]

Out of curiosity, does anyone have a diagram of the internal guts of this particular pump?  This video is pretty good at showing the internals of a pump from a Piper Warrior that's obviously very close to ours.  But it appears to only have a single diaphragm, and it's not exactly clear from the diagram or the actual pump how fuel winds up dribbling out the overboard drain.

[N201MKTurbo]

44 minutes ago, Vance Harral said:

Out of curiosity, does anyone have a diagram of the internal guts of this particular pump?  This video is pretty good at showing the internals of a pump from a Piper Warrior that's obviously very close to ours.  But it appears to only have a single diaphragm, and it's not exactly clear from the diagram or the actual pump how fuel winds up dribbling out the overboard drain.

The two diaphragms are hooked together and move together. They are about 1/4 inch apart. The drain line hooks to the space between the two diaphragms. If the main diaphragm gets a leak, the fuel will leak out the drain line, instead of into the crankcase. This provides a few advantages, it gives a  visual indication of a leaking diaphragm at the first indication of a leak, without the dual diaphragms, you could have a leaking diaphragm for a long time before the fuel pressure dropped to where it became a problem. All the leaking fuel would go into the crankcase and dilute the oil.

The pump chamber is on the bottom of the pump. The bottom diaphragm is the main diaphragm and the top diaphragm is the backup.

[Vance Harral]

Thanks for the explanation, I think I understand it.

As with all "redundant" things, there are some interesting failure modes.  It sounds like there's really no way to tell if the backup diaphragm fails before the main diaphragm, you just realize it later, when the main fails, and the pump fails altogether instead of just dribbling fuel out the overflow drain.  That's no worse than a single diaphragm design, but any reason the main is more likely to fail first, or is it 50/50?

[N201MKTurbo]

5 hours ago, Vance Harral said:

Thanks for the explanation, I think I understand it.

As with all "redundant" things, there are some interesting failure modes.  It sounds like there's really no way to tell if the backup diaphragm fails before the main diaphragm, you just realize it later, when the main fails, and the pump fails altogether instead of just dribbling fuel out the overflow drain.  That's no worse than a single diaphragm design, but any reason the main is more likely to fail first, or is it 50/50?

The backup diaphragm has no pressure on it. It is doing a lot less work than the primary. It is not constantly exposed to fuel. It is unlikely to fail first. Besides, if the backup diaphragm fails first oil will come out the drain line.

[Fly Boomer]

12 hours ago, Vance Harral said:

That's certainly a definitive reference, but the Tempest guide says thusly:

What is an M20-3?

[N201MKTurbo]

The pump in the video has the laminated arm, the U channel arm is pressed from a single piece of sheet steel.

[Vance Harral]

6 hours ago, N201MKTurbo said:

Besides, if the backup diaphragm fails first oil will come out the drain line.

Oh, that's a good observation.  Previously I had always assumed the only source of oil dripping out the lines near the cowl flap would be from the breather line, but there's another possibility.

[Vance Harral]

13 minutes ago, Fly Boomer said:

What is an M20-3?

Haha, that's a good catch.  One presumes that's supposed to say M20-E.

[Fly Boomer]

1 minute ago, Vance Harral said:

Haha, that's a good catch.  One presumes that's supposed to say M20-E.

Must have been written by a 20-something. 

[KSMooniac]

I installed the LW15473 on my overhaul almost 10 years ago now and am satisfied.  It did run a bit over 30 psi for the first few hours but settled in nicely.  EDM warnings were a nuisance during that period, though.  I'm a big fan of Tempest products and would recommend that option without hesitation.

[Ragsf15e]

3 hours ago, Vance Harral said:

Oh, that's a good observation.  Previously I had always assumed the only source of oil dripping out the lines near the cowl flap would be from the breather line, but there's another possibility.

Yes, that’s how mine started to go bad.  The dripping out that line can end up on the nosewheel like everything else, and mine was tied off to the sniffle drain as well, but once we identified the source of the drip, a new pump fixed it.

[Vance Harral]

Thanks again, all.  I've ordered an overhauled LW15473 from Aircraft Spruce, hopefully in the mechanic's hands for installation by the end of this week.

[cliffy]

Back in the OLD days one could order a pump kit and rebuild them on the bench.

Nothing near rocket science to do it but torque the screws correctly and check for wear.

All the over haulers have the parts but no one can release them to the public now.