Fuel Pump Mystery

[Roger Long]

I accidentally ran my main tank dry, fortunately while charging batteries instead of underway. After transferring some fuel and bleeding the system, I discovered that my polishing pump (Walbro FRA-1 without pressure cutoff since it is in an open circuit) was dead. Verified 12 nominal volts to the pump. There are two independent supplies, one from the ignition harness for automatic operation when the engine is on, and one switched. About 500 ohm resistance across the pump leads. Dead when connected to my 12 volt emergency battery. Totally clean and unobstructed when disassembled.

The pump is rated for 4 hours dry running but knocks heavily when dry. I figured the physical knocking running dry while the engine ran on the fuel in the filter bowl broke some connection weakened by nearly 5000 miles of operation. That doesn’t explain the electrical continuity. There is no failure prone pressure switch in this pump. The pump coil appears to be solidly potted in epoxy so I can’t see any electrical connections to fail. No sign of excessive heat.

The pump is my only means for getting fuel into the cabin heater tank without jury rigging the other pump I carry for transferring from jerry cans so I went off to Napa and bought 60 buck plastic Facet pump. It said it was for all carburated automotive and marine engines but didn’t say anything specific about diesel. Neither the NAPA guy nor myself could figure out any reason it wouldn’t pump diesel though. It has a 7 psi pressure cut off switch.

After half a day of operation running the polishing circuit, the pump is dead. 12 nominal volts at the cut pump leads, about 300 oms resistance across the pump leads but dead as the proverbial doornail. I checked occasionally during the day for excessive case heat since it was a small, cheap pump. It was cool every time I put my hand on it.

A new Walbro will be meeting me at the Charleston Marina.

Can you think of any common cause that might be killing these pumps? I put a new filter in the polishing system and have a gauge so I know it isn’t a clogged filter. The Walbro is rated for 100 volt surges and anything like that coming from the ignition harness would probably have fried something else. I’m going to have 150 bucks plus two marina nights (in order to get the FedEx) invested in this new pump so I would hate to have it die the same way.

 

[joker460]

Their website says:

The FRA series fuel pump is designed to supply fuel for internal combustion engine applications only. It is not intended or designed for continuous operation applications such as heater or stove systems.

Not sure if that is part of the problem.But running all day would be considered continuous.

Good luck.

 

[Roger Long]

Their website says: Their FRA series fuel pump is designed to supply fuel for internal combustion engine applications only. It is not intended or designed for continuous operation applications such as heater or stove systems.

Interesting. When I ordered the pump, it had just been listed as a new item being introduced and the site had very little about it other than the basic specs.

I suspect though that this is just a CYA specification because of regulatory issues. This pump only runs on my boat when the engine is operating or for 5-10 minutes once a day when I top off the stove tank during long periods at anchor or when under sail. It just pumps fuel into a gravity feed tank which supplies the stove. The tank has an overflow pipe to the main fuel tank so I don't have to worry about level or overfilling. I just run it long enough to be sure it is filled.

Total hours per day and per year are easily less than it would run in a commercial truck.

Here's the schematic of my system:

 

[Don12364]

Seems like you might just need one of those old fashioned (and so unreliable??) pumps with mechanical contacts that click on and off (and which might need cleaning every five years).
The antique car restorers open theirs up, take the nice solid contacts out and fit a printed circuit board - and I worry.

A 100 volt surge is nothing compared to what comes from an engine starter solenoid or starter motor the instant it is switched off and we all rely on this spike never getting through to our electronics and being absorbed by the boat's batteries.
If you do suspect it is being caused by some form of spike(s) on the line then I suggest you look to use a 'transient surge suppressor'. Get the 18v one and just wire it across the pump's terminals. Less than $1 each.

This item or its equivalent available world wide from electronics suppliers.
e.g. Littelfuse sold by Mouser Electronics at $0.41 each.
See P6KE22(C)A in the table:-
http://docs-europe.electrocomponents.com/webdocs/074c/0900766b8074cd60.pdf
P6KE22(C)A 18.8 20.9 23.1 1 30.6 20.0 5
And
http://www2.mouser.com/Search/Refine.aspx?Keyword=P6KE22(C)A
And good luck and a Merry Christmas. Thanks for all the contribs of your voyage.

 

[Roger Long]

A 100 volt surge is nothing compared to what comes from an engine starter solenoid or starter motor the instant it is switched off and we all rely on this spike never getting through to our electronics and being absorbed by the boat's batteries.

It was religion when I was flying to always have the avionics master switch off whenever the engine was being started or shut down. The avionics stack cost as much as many cruising sailboats.

I did that for a long time but have gotten pretty casual about it. Even fairly cheap stuff I have charging on the system never shows any glitches or resets when the engine is started or stopped. I just watched my battery monitor while starting and stopping the engine and saw nothing except about a 1 volt drop while it was cranking.

If my cell phone on its car charger doesn't die when on line during start and stop, it's hard to imagine a big coil like the ones in these in-line pumps would have a problem.

 

[joker460]

Facet Pumps seem to be a hit and miss proposition.

Here is a thread from some car site where the guys talk about them.

http://www.thesamba.com/vw//forum/viewtopic.php?t=471764

 

[Maine Sail]

Roger,

"Spikes" depend a lot on where you tapped into the pos or neg system wiring. If this was done from the panel then the batts will act as an excellent filter if properly wired. If however you tapped into one of the wires before the battery, eg: the starter circuit wire, you can still see spikes.

 

[Stu Jackson]

Roger, we did some research on fuel pumps, two pages, second page has good info here:

http://c34.org/bbs/index.php/topic,2515.0.html

Our Facet has been running continuously for our 2480 engine hours. We're on our second one and I have the plastic one as a backup.

Good luck, Happy Holidays and thanks again for your stories.

 

[alexco38]

isolated starting and house systems

Blueseas has a dual circuit master switch that effectively isolate the two batteries. I have it for 5 years and never have any starter surge.
in reference to fuel polish pumps ask for diesel fuel lift pump for old pickup trucks. E.g GMC
as far as fuel polish we have a windup timer to shut off in an hour. You can get them in big box store for bathroom fan.

 

[Maine Sail]

I don't think Rogers problem is with voltage spikes but there is always potential that it could be. As for voltage spikes across a battery I have not been able to measure them even with an o-scope. In a properly wired system this should be a non issue.

 

[Benny17441]

Nothing to indicate or link that the failure of the Walbro pump is related to the failure of the Facet pump. Perhaps the Facet pump is not up to the task of running continuously for 6-12 hours. No pump can be reliably set to pump dry for any extended length of time no matter for the conditions it was rated at so I would not fault the Walbro. I'm a stickler for always calculating how much fuel I have on board but I figure that you will not allow for that to happen a second time. You indicate the tank ran dry when charging batteries. I understand is a common practice among cruisers and I"m guilty of it in running the main engine at anchor just to recharge batteries. Nowadays with the advent of the portable Honda gas powered generators it is no longer necessary to do so and I have not probably had to in the past five years. Yes they may be noisy and gasoline fuel requires careful handling and storage but they are relatively cheap and easy to replace when compared to the costs and labor of rebuilding the engine or repowering the boat. I would think you will have no problems with the new Walbro pump. I want to thank you as I'm enjoying immensely the following of your trip. Great narrations and photos. Thanks again.

 

[Roger Long]

Perhaps the Facet pump is not up to the task of running continuously for 6-12 hours.

I think that is the most likely explanation. I'm thinking back to the phrase on the package, "For all carbureted (italics mine) automotive and marine engines." Since the carburetor float float keeps stopping the flow which shuts off the external fuel pump via the pressure switch, the pump is only pumping as much fuel as the engine burns. Functioning in an open circuit, it's pumping many times that.

The Walbro is the same basic design but higher quality. These pumps are basically a solenoid, a coil of wire that sucks in a magnetic core. These are what lock and unlock your car doors. In the pump, the metal core is hollow with a check valve in one end. There is another check valve in the bottom of the pump. The coil pulls the core down against a spring and the fuel is pushed along when the spring rebounds. The weak point is probably the mechanism that shuts off the current to the coil. It could either be a timer or another small coil that senses the position of the plunger coil. In either case, it is a small switching circuit.


The Walbro model I'm using does not have the common pressure sensor to shut the pump off, as when the carburetor float chamber is full or the pump gets ahead of the engine driven pump draw. It's therefore intended for continuous operation against an unrestricted flow.


Both the pressure sensors and the coil current shut off would be a lot more delicate and trouble prone than the very simple and robust pumping mechanism. Failure of either could turn the pump into a paperweight.


I don't anticipate any problems with the new Walbro.


The good news is that I figured out that I can turn the emergency cross over valve in my system (see my schematic in post above) and the main fuel system pump will fill the heating day tank so I am nice and toasty while I type this without having to jury rig the electric pump I carry for pumping fuel out of Jerry jugs. This feature was always in the fuel system I designed but I didn't realize it.

 

[Bill Roosa]

This type pump does not need a pressure senors due to the fact that the carburetor has a float valve. When the valve is closed the pump continues to cycle until the pressure in the pump head is sufficient to not allow the plunger to return and start a new cycle.
I'd say you use an intermittent pump in a continuous fashion and it (predictably) failed.
For the record you can use this type of pump in a diesel also. Just not one with a polishing system that basically never lets the pressure build up.

 

[Roger Long]

I'd say you use an intermittent pump in a continuous fashion and it (predictably) failed.

That probably explains the nearly immediate failure of the cheap pump I bought to get me through the week or so it would take to get a new Walbro. Still, I would have expected more than a day from it.

The original Walbro, which I have replaced this week, is continuously rated. These pumps are used for injector cooling pumps which would be nearly identical service conditions to my polishing system except for the fuel being hotter.

From the Walbor web site:

The early failure of the Walbro is clearly not mechanical. The inside is perfect. Whatever died was inside the potted coil and circuitry. I bought mine just as they came on the market so there may have been some early production run bugs. I'm going to carry a spare though now.

 

[Bill Roosa]

That makes sense since it was cycling at full speed (never designed to do that except when the car is at WOT, a clearly intermittent usage for most of us) the whole time.

 

[woodster]

dont know if your pump is designed to shut down when it reaches certian presures or not...but if it is....you may want to consider putting a regulator on it so it does not run wide open all the time....extending the pump life.....just a though and may not be relateted to this application.......

regards

woody

 

[Roger Long]

you may want to consider putting a regulator on it so it does not run wide open all the time....extending the pump life

Just for grins, and while waiting for the dog walkers to return, I took the Walbro apart again. The plunger is absolutely free moving and with no sign of wear. I could make it suck my finger by pushing it against the spring with a screw driver. The check valve in the plunger passes the blow test. There is no pressure regulator on this pump to go bad.

I put it back together and it is still dead. 12.62 volts on my portable battery and 12.09 when the meter is in series with the pump so it is not a broken connection inside.

There must be either a timer circuit or a plunger position sensor to turn off the coil intermittently. I suspect the former because the pulse time doesn't seem to vary much when it is priming and the load comes on. Whichever it is, it seems to think the plunger is in the down position and the power shouldn't be turned on.

The plunger control circuit is potted in the epoxy. It's hard to see how its life could be effected by the difference in mechanical function between the pump running at light load and under pressure. This pump was designed to operate continuously at about the head pressures in my system anyway.

I'm sure now that the cheap pump did self destruct due to running free and don't expect any problems with the new Walbro.

 

[Bill Roosa]

The plunger moves away from the pump head sucking in fuel. Upon reaching the bottom the cycle switch "flips" and the circuit turns off. The spring forces the plunger up. But wait, the fuel is incompressible and the float valve is closed so the plunger can't go up until the float valve opens and releases some fuel into the carb. When enough fuel has been released by the float valve the plunger will have moved up to the top of its stroke and "flips” the cycle switch activating the plunger to move down again, drawing more fuel. flip and the spring .......... repeat till burn out.
So on a car/truck the plunger winding is not actually causing the fuel to be pressurized it is the spring. Course if the cycle switch does not compress the spring it never even gets started.
And if you operate it on your polishing circuit it will just cycle as fast as fuel can move through the polishing circuit.

 

[Roger Long]

Course if the cycle switch does not compress the spring it never even gets started.

Ah, there is a possible failure mode. The position sensor would probably be at the top of the stroke. The circuitry would then give just a pulse of voltage to compress the plunger. It will go quickly because it is only working against the resistance necessary to squirt fuel through the check valve. As you say, the spring does the work.

If the spring got compressed through too many cycles so that the plunger didn't quite reach the top of the stroke, the pump wouldn't start. That wouldn't have been apparent in my inspection.

It's a neat theory because the spring would have a finite number of cycles before fatiguing and taking just enough of a set to not quite bring the plunger to the sensor.

Later:

Nice theory but, no ceegar. I took the pump apart again and over stretched the spring so that the plunger sat well above the robust rubber bumper until compressed back in place by tightening the cover screws. Still about the same response as connecting 12 volt leads to a stone.

This is an electrical, not a mechanical problem. As such, I have a hard time seeing how it is related to the service conditions and duty cycle I have subjected this pump to.

 

[Don12364]

Wrong Type of Pump Altogether??

Despite the maker's claims and for all the reasons listed in the previous responses, in the fuel polisher application one would not make a reciprocating or diaphragm pump as the first choice.
IMHO it should be a Jabsco or a peristaltic type, either in front of the filter, or after it depending upon the pump's ability to pump sludge should this be present.
Looking at Jabsco's web site they show the very thing with a similar throughput (3 litres/min) as part of their 'Flat tank oil changer Kit' but this pump does not appear elsewhere in their listings.
However the whole kit, including the tank, is somewhat (30%) more expensive than the Walbro - but if the pump were available separately it might be competitive.
Regarding its ability to run dry - why should this ever occur in your application unless you run right out of fuel?
Please see:-
http://www.jabscoshop.com/marine/pu...umps/17860-0012-flat-tank-oil-changer-12v.htm
and:-
http://www.amazon.com/Jabsco-17860-0012-Changer-3-5-Gallon-12-Volt/dp/B000O8B0M0