neuperg

I have spent an hour looking at fuel/FP threads but I don't seem to see an answer to my question.

Can an electrically failed or disconnected internal FP prevent the external FP from developing any pressure at all.

i.e. I am not seeing pressure even with the FP relay bridged and external pump rinning. .
.
Details below.

======================

Dual fuel pumps on my 1990 GT. One internal, one external.
Less than 500 miles after replacing both pumps, interconnecting hose, filter and flushing the tank.

Gas tank half full.

No start. No fuel pump sound. No fuel pressure.

I bridge the FP relay
External FP runs, internal FP makes no sound and still no start and no fuel pressure.
I see 12 V at the terminals of the internal FP but no sound.
I check continuity of internal FP and it is open.

HERE IS THE QUESTION

Can an electrically failed or disconnected internal FP prevent the external FP from developing any pressure at all.

i.e. I am not seeing pressure even with the FP relay bridged.

worf928

The internal pump is hard to hear with the main pump running.

A failed or non-running in-tank pump will not - likely - cause a running external pump to produce no pressure. The in-tank pump is there to prevent vapor lock. So, assuming you don't have vapor lock, ...

If you have fuel, the main pump is running and there is no blockage in the system (you have a new filter) then there really should be fuel flow.

How are you testing for fuel pressure? Do you have a gauge attach to the fuel rail?

neuperg

Yes gauge on fuel rail.

No continuity, open circuit on internal fuel pump suggests that it is open circuit and cannot run.

Mrmerlin

remove the internal pump replace with a fuel strainer ,
replace the external pump with 044 pump

worf928

Yup. Rather than buying a new $900 in-tank pump. This I did to my '91 when the in-tank pump failed a couple of years ago. It's been running fine with the strainer, OE main pump (and the supercharger.)

Not necessary IMO. The main fuel pump has identical specs to the pumps used without the in-tank pump despite the Bosch part number change. And you don't have to do the recommended wiring 'upgrade' for the 044 pump.

Ok. Good.

Yup.

I have pulled non-working in-tank pumps from 928s that ran on the main fuel pump only. So, even though the (new?) in-tank pump appears to be dead, I'm going to bet that it has ingested something that's blocking flow. Might be what caused its infant mortality.

You'll have to pull your in-tank pump again and see what's what. If it is indeed dead, then buy the strainer: 928 201 081 04

FredR

This is an interesting problem and I look forward to learning of the outcome as it unfolds.

Different folks have different experiences and my history [in a very hot climate] tells me that I need both pumps under certain conditions however my pumping system still worked but not 100%.

You say the in tank pump is not working so the first question I would ask is "how do you know?" This pump flows the same as the main pump but it only adds a small amount of head to help ensure the main pump does not cavitate. As you will know it is attached via a small section of hose and is submerged in fuel so whether one can actually expect to hear anything when it is running remains to be seen. If I were trying to understand such I would disconnect the main pump, put my stethoscope on the mounting nut and see if I can actually hear anything when the power is applied and even then there is no hard metal connection so even that may not be conclusive.

I have searched for as long as I can remember to get a data sheet for the in-tank pump and found nothing. All I can say is that it will likely only add in the region of a few psi- sufficient to prime the main pump and at this level of pressure you most likely will not see the pressure gauge you have move off the bottom stop such is the way a Bourdon gauge works. Even if the in tank pump has failed I would still expect something to get through to the main pump but the in-tank pump then acts as an additional flow restrictor and cavitation is then most likely in the main pump. The main pump should still generate some measurable pressure but could possibly struggle to see the 3 barg expected at the rail with the in tank pump dead.

Now, referring your input data- if the main pump is not running until you jumper the relay then that is the first problem you need to solve. Either the relay is goosed or possibly the LH computer- that or anything else anyone can think of.

Next point- when you jumpered the fuel pump relay and got the main pump running, did you then try to start the engine? One is left to assume you did and still it refused to start but...?

If the pump is running and you are not seeing any pressure the cause would likely have to be one of the following: fuel is not getting to the pump, the pump impeller has sheared from the motor shaft, there is a blockage between the pump and the pressure gauge, the pressure gauge is not working or the pressure controller has failed.

The thing to remember is that if the car was running correctly until this incident then a single failure is quite possible but the odds of two things going wrong drop dramatically unless one failure logically leads to something else going wrong, for example when the in-tank rubber pipe fails and fragmented bits consequently jam the main pump impeller.

Trust the above might help your analysis a little.

SeanR

Why, he has new parts in there. If you are replacing the intake pump with a strainer, use the '87-88 pump. No reason to toss in an 044 unless you are planning major power.

neuperg

This is an interesting problem and I look forward to learning of the outcome as it unfolds.

Different folks have different experiences and my history [in a very hot climate] tells me that I need both pumps under certain conditions however my pumping system still worked but not 100%.

I am in PHX AZ so plenty hot here. Is the only reason for the internal FP for the vapor lock due to heat? If so, why not use a strainer followed by an external peltier fuel cooler?

You say the in tank pump is not working so the first question I would ask is "how do you know?"

No continuity on internal FP. i.e. Infinite impedance when checked with multimenter
No sound when the internal FP alone is energized from a regulated 12V power supply with built in AMp meter. Amps showing 0A on meter. .

This pump flows the same as the main pump but it only adds a small amount of head to help ensure the main pump does not cavitate. As you will know it is attached via a small section of hose and is submerged in fuel so whether one can actually expect to hear anything when it is running remains to be seen.

Tank half full, listened with ear pressed to bottom of tank. .Then did the electrical continuity tests outlined above to be certain that internal FP was likely failed.

If I were trying to understand such I would disconnect the main pump, put my stethoscope on the mounting nut and see if I can actually hear anything when the power is applied and even then there is no hard metal connection so even that may not be conclusive.

See the above.

I have searched for as long as I can remember to get a data sheet for the in-tank pump and found nothing.

Will test with my 0-5PSI 6" gauge or manometer if it is ~ 1psig. ...assuming I do get another internal pump. Will post the test result.

All I can say is that it will likely only add in the region of a few psi- sufficient to prime the main pump and at this level of pressure you most likely will not see the pressure gauge you have move off the bottom stop such is the way a Bourdon gauge works. Even if the in tank pump has failed I would still expect something to get through to the main pump but the in-tank pump then acts as an additional flow restrictor and cavitation is then most likely in the main pump. The main pump should still generate some measurable pressure but could possibly struggle to see the 3 barg expected at the rail with the in tank pump dead.

Yeah surprised that the external pump is spinning and there is not sufficient leakage past the in tank FP that some pressure cannot be developed. FYI, my high pressure gauge will not show any indication till ~ 5 psi.

Now, referring your input data- if the main pump is not running until you jumper the relay then that is the first problem you need to solve. Either the relay is goosed or possibly the LH computer- that or anything else anyone can think of.

Was thinking solving FP first and then the LH.

Next point- when you jumpered the fuel pump relay and got the main pump running, did you then try to start the engine? One is left to assume you did and still it refused to start but...?

Yes, one of the first things I did. With jumered relay, ext FP runs but no pressure as measured a fuel rail and of course no start.

If the pump is running and you are not seeing any pressure the cause would likely have to be one of the following: fuel is not getting to the pump, the pump impeller has sheared from the motor shaft, there is a blockage between the pump and the pressure gauge, the pressure gauge is not working or the pressure controller has failed.

Checked pressure gauge by using my air compressor (cheap) regulator set to to 45 Psi and my Fuel pressure gauge indicates 47 Psi.


The thing to remember is that if the car was running correctly until this incident then a single failure is quite possible but the odds of two things going wrong drop dramatically unless one failure logically leads to something else going wrong, for example when the in-tank rubber pipe fails and fragmented bits consequently jam the main pump impeller.

Trust the above might help your analysis a little.

Yes, thanks for everyone's feedback.

Now just need to sort out if I replace a $440 internal FP or go with strainer and an inline Peltier fuel cooler to avoid the vapor lock. I am off to research cooler prices. But far simpler (and likely cheaper) replacing and external cooler vs the internal FP if this ever happens again. No moving parts in Peltier cooler. Not likely to get obstructed.

worf928

I don't think anyone knows FOR SURE. Pre-'89 there was no in-tank pump. (IIRC, the very early cars had two pumps, but don't quote me as my expertise drops off significantly for the pre-'87 928s.) IMO the purpose of the in-tank pump is to prevent vapor lock if the pump check valve and/or regulator check valve is leaking.

There is an external A/C-cooled fuel cooler in the engine bay at the return line.

Yeah. It's dead. Don't know why there's a question about it at this point.

I have no more questions about your metrology methods. You seem to be ensuring that your observations are backed by measurement instruments you've tested or you've got multiple ways to ensure the validity of the conclusions you draw from your observations.

If you can get a new in-tank pump for $440 that isn't a low-quality knock-off... last price I saw was >$900 for a new unit.

If vapor lock on '87-'88 (and the earlier single-pump cars) was a serious problem I think it would be a much-more common topic here.
Above, I wrote that I run the strainer with the original main pump on my supercharged '91. I don't drive around in 100-degree temps often. But, the heat soak from the Autorotor is pretty spectacular. I don't have vapor lock issues.

That said, I'm not trying to convince you to not put in either a new in-tank pump or another external cooler.

But, I would certainly want to thoroughly understand the current failure mode before deciding on 'upgrades.'

It's a very puzzling failure you have since you, apparently, did everything right: New pumps, clean tank, new filter...

worf928

One question: Have you tested for flow after the fuel filter? (Run a hose into a bucket?) Maybe you've got blockage between the filter and the fuel rail...

FredR

It sure seems strange that with new components you are having such issues. Clearly there is a problem on the power supply side that you need to resolve. Beyond that once you got the main pump running then logic suggests either there is a blockage on the suction side of the main pump or the pressure controller is not working. If there is no back pressure the pump will simply run down its curve pumping more flow and generate little to no head in the process.

In a hot climate you are trying to avoid cavitation in the main pump. The problem is that gasoline has light components in it and the consequence is a boiling point range that may start at or around 50C. The fuel heats up as it circulates and for hot weather Porsche tried to deal with this by having the fuel cooler in the a/c loop their logic being that if it is hot the a/c will be running. If the fuel reaches its initial boiling point then pumping the liquid becomes much more difficult as there is so little hydrostatic head in the tank and as I found, once the level drops to about 1/4 of tank I had fuel starvation issues. Had the in tank pump not been there the problem would simply be exacerbated.

neuperg

FYI summer AZ temps are often over 110 and have seen 117 summer before last. And this is in the shade. .

With the tank over hot asphalt in slow traffic or parking lot. I am sure temps are in the 125+ F under the car. Summer asphalt can hit 200 degF.

My AC has not working at the time of the last failure which was in late October. Temps 90 to 100. .

If either pump cavitates, can it damage itself if it is sustained?
I know cavitation erodes boat propellers and other fluid handling equipment.

Frank

worf928

To what do you refer above? As far as I know the closest thing to a 'pressure controller' in the system are the check valves, regulator and the dampeners. I'm pretty sure there's nothing in the pumps.

A long time ago I did a fuel pump 'autopsy.' The main pump is basically an electric motor that spins a centrifugal pump mechanism with - if memory serves - 4 moving parts besides the motor itself. There's not much else in there - a couple of seals. All the innards are cooled and lubricated by the fuel flow in which they are immersed. It doesn't generate a lot of torque: a tiny bit of the in-tank pump (or a bit of debris not much bigger than a grain of sand) can seize the pump's moving bits, thereby stopping the motor which then blows the fuel pump fuse. But, if you clean out the 'gook' from the pump mechanism, the pump will once more run (good luck getting it back together though as getting to this point destroys the housing.)

However, reversing the polarity on a 'seized' pump can often cause it to 'spit out' whatever's seized the pump's moving bits.

As far as cavitation damage is concerned, since the bits are fuel-cooled I would opine that heat-induced damage would occur before classical cavitation damage (erosion due to the high-energy of the 'bubbles popping.') This for the main pump. I've not dissected an in-tank pump.

FredR

Dave,

The "regulator" is a "self regulating pressure control valve" referenced to the pressure inside the fuel rail. The device drives tight shut until the pressure builds to that required and then the diaphragm force balance causes the thing to progressively open to maintain a constant set pressure in the rail. This pressure differential is needed to accurately meter the fuel flow that is calibrated in terms of injector pulse width duration and why there is a vac line to the inlet manifold to ensure the rail pressure is referenced to that in the inlet manifold where the fuel discharges to. If that controller fails [for instance the spring inside it fails], the diaphragm will open and the pump will flow well down its curve and in all probability will not deliver anywhere near the required head. The normal failure mode for this item is when the diaphragm fails but it is not beyond the realms of possibility for the spring to fail.

I suspect you well understand this just the nomenclature I used.

Thus if the pump is spinning and there is no pressure, then there are only three possibilities I can visualise: either there is a system blockage, the impeller has sheared from the shaft or the pressure controller has failed. The info presented suggests to me the most likely scenario is a failure of the pressure "regulator" but the other items need to be eliminated.

Cavitation is a pump killer- make no mistake about that. As to how often it occurs in a 928 I have no idea and neither will most folks as pumps more often than not are tossed away and replaced, usually due to motor failure. My late S4 died on me some 15 years ago, fortunately it died close to the main dealers premises which was also close to my residence. We recovered the vehicle back to their workshop, they diagnosed a failed fuel pump, I asked them to drop the pump and spin it up with reverse polarity and sure enough out popped a little piece of rubber. We pulled the in-tank and of course the rubber pipe had split. Fitted a new piece of fuel hose and the thing was good to go in little more than an hour.

FredR

Frank,
Cavitation, should it occur, will wreck any pump irrespective of the design and it will do so rather quickly.

Refiners configure the gasoline to minimise the chances of this happening but what they cannot control is the fuel temperature in your petrol tank. Not sure what the current legislative specs are but last I checked Reid Vapor Pressure specs were somewhere in the region of 9 psia at 38 degrees C [there or thereabouts] in the summer season. In the winter season this number is increased and this allows refiners to throw in more butane which is cheaper so gas prices drop during the winter season. It does not need a genius to understand that if the ambient temperature exceeds that of the RVP spec [38 C] as in your summer case, then the margin between the pump NPSHA and the NPSHR is reduced. A pump with a low NPSHR will typically be about 5 ft. The difference between atmospheric pressure and the rated Reid vapour pressure is 5.7 psi or 17 ft of head so in the case of the in tank pump there is a margin of around 12 ft before cavitation occurs. This can quickly reduce if/when the fuel temperature rises above 38C as it surely will in your case. If the in tank pump stops working then this creates additional losses. Thus in a hot climate the in tank pump is desirable but if it is not working then it is better to have a strainer and no in tank pump at all.

In fuel injection systems like ours the fuel is circulated and picks up heat in the pump [not much] and by conduction/convection in the engine bay. For a pump the NPSHA [net positive suction head available] is defined as Atmospheric pressure minus the vapour pressure plus the head above the pump suction centreline minus any losses in the piping system. If the NPSHR [nett positive suction head required] is greater than the NPSHA the pump will cavitate. Running at altitude will eat into these margins as will running winter gasoline during summer months.

To summarise, one would have to try pretty hard to get cavitation to occur but it is not beyond the realms of possibility if the in tank pump fails.

Your value for the black bulb temperature is a bit optimistic - we design for 82C [180F]. Thus eggs fry on the pavement and you need to run pretty quick when crossing the beach to get into the sea! When I first went to the Mid East I developed a skin rash- the company doctor told me my hands were burnt on the steel handrails and that I should have been wearing gloves!