What's new
Van's Air Force

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

Dying Engine Driven Fuel Pump

nigelspeedy

Well Known Member
When my IO-360 with mechanical fuel injection was new the engine driven pump would put out ~27 psi from idle to take off power with nary a fluctuation. Now three years and 600 hours later the output pressure has steadily dropped to an average of 20 psi and the fluctuations have increased to as much as a 5 psi drops in the cruise. The drops down to 15 psi could be felt as slight misses or a loss of power. Putting the electrical pump would return the fuel pressure to 28 psi. The interesting thing is that it has happened very slowly so there was no real obvious change in pump performance. The fuel filter has been cleaned every 100 hours but has only ever had the slightest amount of debris in it. The fuel pressure sender was replaced at around 400 hours, along with the oil pressure sensor when they both became a bit erratic. They were VDO types supplied with early Dynon Skyview systems. I've made no changes to the fuel system since new so I was pretty sure it was the mechanical pump. One thing I have done since new is upload my flights to Savvy Analysis which makes it really easy to compare a recent flight with a very early one and use data to make comparisons rather than having to rely on my frail memory. So $422 later I have a new pump and it puts out a steady 28 psi and all is good in the world of RV again.

So a couple of questions:
1. Are the pumps rebuild-able, I'm thinking a spare would be nice.
2. Are there other failure modes, as opposed to the very slow degradation that I experienced.

Cheers

Nige
 
Just replaced mine too, Lycoming pump, similar age. Seemed to have a failing oil diaphragm. I'm planning an autopsy. Not inside yet, but past posts suggest that if replacement parts are available, the fuel and oil diaphragms would be an integrated assembly on the lift shaft, with the springs.
 
They are technically rebuildable; see:
http://www.aircraftspruce.com/catalog/eppages/TEMPESTfuelpumpLYC.php?clickkey=17874

However, I've talked directly with Tempest about it, and they say it takes 'special jigs' to properly align the diaphragms, etc in the pump body. They won't sell rebuild kits directly to us. You might be able to find an online seller that will sell to you, but I wasn't able to, several years ago when I was researching it.

Now, for you own 'education and recreation', you could disassemble your old one. Google 'AC Delco fuel pump rebuild kit', 'Carter fuel pump rebuild kit', 'Chevy 283 fuel pump rebuild kit', 'Ford 302 fuel pump rebuild kit', etc. Then compare images with what you've removed from your pump. Now I'd never advise you to actually fly with dangerous automotive parts attached to your engine, even if you found some that exactly matched the ones you removed from your old pump....

There's a recent thread here that details the pump's possible failure modes & results of each failure.

Charlie
 
Years ago, I had a friend's high pressure fuel injection mechanical pump just quit. In other words, it worked fine on one flight but the next flight did not work at all.

He purchased a new pump in a Lycoming box for his fuel injected engine. The part number on the box was the correct number for a high pressure fuel injected Lycoming engine but once installed, his engine would not run without the boost pump being on. Examination of the pump found a low pressure pump part number stamped on the pump.

All I am saying is if your fuel pump has the correct number on the box, verify that the number on the pump is the same correct number.
 
A few posts back I promised an autopsy. This is a Lycoming, probably manufactured in the 2006-2007 time frame. It first saw fuel in the fall of 2007. I'm going to assume readers know how it works, and go straight to failures.

First up...there is a shaft seal in the upper spring seat. Pretty sure it is intended to slide on the shaft, but here it was found with the center ripped out. The seal separates accessory case volume from the pump's internal volume, around the spring and above the oil diaphragm. That volume is vented back to the accessory case, so loss of this seal mostly just keeps the top of the oil diaphragm more wet than usual.

Broken%20Slide%20Seal.jpg


Th oil diaphragm appears to be a sandwich, two plies of rubber with a fabric core. It was deteriorated on its upper surface, enough to allow oil through into the core fabric, but the lower surface was not yet compromised.

Damaged%20Oil%20Diaphragm.jpg


There was a lot of oil between the oil (right) and fuel (left, with tab) diaphragms. Liquid in this space (fuel or oil) is vented through a restrictor and shows up in the tell-tale vent hose. As the oil diaphragm was not yet compromised on its lower surface, this oil seems to have found its way in via seepage at the perimeter screw holes.

Oil%20between%20Diaphragms.jpg


Reminder; the pump vent hose is not to be terminated above the exhaust pipe, as some might do with a crankcase vent. You want the fuel pump vent where it will make a mess obvious during preflight.

Vent restrictor:

Weep%20Hole%202.jpg


Next we have the fuel diaphragm. This is the side facing the pumping chamber. Yes, the rubber is deteriorating here too, but again, the cracking had not yet made it through to the other side. If it had gone that far, it would have spit fuel out the vent, rather than merely showing signs of oil in the vent.

Damaged%20Fuel%20Diaphragm2.jpg


Frankly, I was surprised at the poor condition of the rubber parts. Age is roughly 11 years from manufacture and a dyno run, and about 8 years/750 hours in flight, all on 100LL.

It is possible to inspect these diaphragms, but arguably not worth the trouble. A a practical matter it would require removing the pump from the accessory case. Then, on the bench, removing the top and bottom perimeter screws would allow viewing both diaphragms. With the pump off the engine, may as well just install a new one.
 
Last edited:
Y guess

I would look at the spring tension or the spring perch to see if it has shifted. I recall these pumps are really a spring force driven pump; all the plunger does is return the spring to the bottom of its stroke to push again. JMHO.
 
Thanks Dan

Hi Dan,

Thanks for the fuel pump dissection and failure analysis. I contacted the pump manufacture to ask if they sold rebuild kits. Turns out they do not and they want you to send your old one in as a core for swap. Pretty expensive though and only about $80 cheaper than a new one.
Cheers
Nige
 
While polishing the belly of my 8, I found blue fuel stains on the exit ramp. Upon investigation, I found that fuel was leaking out the drain during operation of the engine / pump. What kind of surprised me is that if I hadn't noticed the blue fuel stain on the exit ramp, I probably wouldn't noticed the leak since the fuel was draining out near the exhaust pipe and evaporating so quickly, and only when the engine was running. It didn't leak when the used the boost pump to pressurize the fuel system. Once shut down, any fuel I smelled was attributed to the typical smells of engine shut down. My fuel pressure was ~ 22 psi, but occasionally it would dip to 18 psi during a post landing hot engine taxi, which I attributed to boiling of the fuel in the fuel line up to the pressure sensor. Ordered a new Lycoming pump for $422...and looking forward to possibly higher system pressure and no leaks.

Two lessons I learned from Dan H...

Use clear or translucent tubing for the pump drain line so you can see the fuel stain better (I suppose this is in the Vans plans somewhere and I missed it)

Don't locate the pump drain line next to the engine breather vent...near the exhaust pipe, but someplace where it is readily visible....

But I am sure everyone else already knew this!
 
Just adding my experience. Noticed fuel leaking from vent after shutdown after a 2.5 hour cross country IO 360. Engine had been running fine. Calculate that I had about 1.5 gallons less fuel than expected (via totalizer). Started back up, pressure 25 psi, engine ran normal, but fuel was leaking out of the vent at approximately 4 oz in 30 seconds at 1400 rpm.!!!

Not sure when the diaphragm let go or how long it had been leaking (prominate blue stain on belly back to spar area), or what the in flight leak rate was, but at that ground leak rate I would not have made it far.

Stranded at a very podunk airport in Allendale SC Saturday 4th of July weekend - no motels, no rental cars, no nothing. Local airport rat drove me 60 miles to Augusta GA where hopefully I can rent a car Monday. What a gracious guy!

With rental car can drive to Atlanta, get a pump at spruce, figure out how to get back to Allendale and install pump on ramp.

Pump was 12 years and 550 hrs old. Suggest that you change it as routine at 10 years/500 hours to maybe avoid a catastrophe.

Up until this point I had always believed that a diaphragm failure in these pumps was a relatively benign event indicated by a “non serious” fuel leak at the vent. Not so - this failure had the potential to be a serious flight hazard!
 
Last edited:
I had my fuel pressure drop. Thought it was my pump, but fortunately before pulling it I was told to check all the fuel lines and connections to insure there were no loose fittings etc. Discovered the filter B nut was backed off just slightly. No fuel was leaking but it was just enough to allow air to get sucked in. . Just that little bit caused a loss of fuel pressure from 23-25 to drop to 10-12 on an IO 540. It caused cavitation inside the pump. Afterwards pressure was back to normal and no other issues.
 
pressure fluctuation

I have just this week noticed that the fuel pressure will fluctuate between a fairly normal 25 psi and slowly drop as low as 21 psi for an instant, then back up to 25. The pressure fluctuation is like a sine wave variation over a 6--7 second period.

Based on what Nigel shared from two years ago, I think it is time for a new fuel pump?
 
New pump OK?

So, I ordered a new Tempest AF-15473 fuel pump. When it arrived, I was kind of surprised that I can not cycle the pump manually by pushing on the lever.

This is a high pressure pump for fuel-injected engines.

Are the springs just so stiff on the high-pressure pump that I should not be able to move the lever?
My recollection on old auto fuel pumps is that it was easy to move the lever and cycle the pump. But those are all low pressure pumps.

So I don't know if I should be worried that this one is so stiff or not.???
 
There is very little movement of the arm. Surprised me as well when I replaced my low-pressure pump last year.

Somewhere here on VAF, there is a post describing the limited sweep range of the arm.
 
So, I ordered a new Tempest AF-15473 fuel pump. When it arrived, I was kind of surprised that I can not cycle the pump manually by pushing on the lever.

Has a big 'ole spring, but...didja remove the plastic plug from the inlet?

Just spittballin'
 
So, I ordered a new Tempest AF-15473 fuel pump. When it arrived, I was kind of surprised that I can not cycle the pump manually by pushing on the lever.

This is a high pressure pump for fuel-injected engines.

Are the springs just so stiff on the high-pressure pump that I should not be able to move the lever?
My recollection on old auto fuel pumps is that it was easy to move the lever and cycle the pump. But those are all low pressure pumps.

So I don't know if I should be worried that this one is so stiff or not.???

When you see how small of an eccentricity you have in the drive cam, you’ll realize just how little motion the pump actually has - maybe 3/8” at the tip (from memory). With a stiff spring, it is hard to hold the pump body still enough to see motion at the tip of the arm.
 
Thanks guys.
Yes, I removed the plugs. Mine has a bit over 3/16" of what feels like free play motion, and then it is extremely stiff. Must be that big ole spring. If I push it against a bench, it does indeed cycle. I was just being a little too gentle. I'm not generally a 'use a bigger hammer' kind of guy.

I also note that the mounting holes are tapered (conical) as if the pump body fit over tapered pins. Didn't expect that.

I also decided that while i am at it, I will install a cooling shroud. I have a blast tube aimed at the pump currently, but no shroud. The ACS shroud doesn't fit as shipped. Think of it as a kit to make a shroud from and you will be happy. Think of it as a shroud ready to install out of the box and you will be unhappy.
 
Last edited:
Back
Top