98 Escort, fuel pump not running..

98 Escort wagon SE, 2.0 L, SOHC
I have Chilton's manual

I had the tank refurbished about a month ago, there were several pinholes in the tank near the filler hose connection. Yesterday morning the car would not start. I noticed immediately that the fuel pump was not coming on when the key was turned to the run position.

So i performed a battery (pun intended) of electrical tests on the fuel pump circuit. Here is what i know:

The battery voltage is 12.6 V when everything is off
With the ignition int he run position the voltage drops to 11.8

There is continunity in the inertia switch (ie. functioning properly).

When i traced the wires to the fuel pump (the harness going into the pump has 6 wires (2 for the guage, 2 for the warning light and 2 for the pump...i assume).

With the ignition switch in the run position, the voltage across this harness (while disconnected from the fuel pump) reads 11.8 and it stays on. However when i connect the fuel pump and try again, the voltage climbs to 7.8 and then drops back to zero. Presumably the 11.8 is 'stray voltage' that disappears once a load is placed on the line. In addition, the 7.8 is the voltage across the pump which is probably too low to run the pump.

Why is it too low? Well based on the wiring diagram, the only 2 faulty parts are the fuel pump itself and the fuel pump driver module (which i assume is somewhere under the dash).

For the next step i measured the resistance across the fuel pump lines which was 1.1 Ohms. This suggests to me that the pump is pooched, (ie. has a short). The resistance across the leads of my multimeter alone is about 0.6, i would have expected the resistance across the pump to be 5 to 10 Ohms.

I am still not positive thet it is the pump though. So my next 2 steps, before pulling the pump out of the tank were going to be 1. measuring the resistance across a new pump and 2. connecting my current pump directly to a 12 V power supply (i was going to use my battery charger set on the 10 Amp setting, with a 5 or 10 Amp fuse in line on the power lead). My thinking is if the pump runs on a direct power source, its probably fine.

I have searched this forum for similar porblems and came up dry. I am wondering if anybody has any more helpful advice, or if there is a flaw in my logic (ie. am i missing something)?

please help,
dllewell

My Escort is diesel-engined so I can't offer any direct help, but perhaps this page in the Autozone repair guide may be of some assistance if you haven't already seen it:

Escort fuel pump fault-finding (http://www.autozone.com/servlet/UiBroker?ForwardPage=/az/cds/en_us/0900823d/80/13/c3/a8/0900823d8013c3a8.jsp)

Edit: also, be sure to check out that hidden inertia fuel cut-off switch which they mention. That turned out to be the problem for somebody in a recent thread on here although in that particular case the car was refusing to start following a low-speed impact in a parking lot - the inertia switch had tripped. Not likely to be the fault in your case but still worth a look.

thanks for the response Selectron, I mentioned the Inertia Switch in my first post, it is fine. I also checked out the link that you attached. It looks like it was taken directly, word-for word, out of the Chilton manual that i have. Note the last 3 sentences:

"Using the scan tool, enter output test mode and turn on the fuel pump circuit.

Using a Digital Volt Ohmmeter (DVOM), check for voltage (approximately 10.5 volts) at the fuel pump electrical connector.

If the pump is getting a good voltage supply, the ground connection is good and the fuel pressure is not within specification, then replace the pump. "

I do not have the scan tool, but it seems to me that turning the ignition switch to run activates the fuel pump circuit (at least for a few seconds). I have done this and, as i mentioned, the voltage was only 7.8 (which is significantly lower than the 10.5 that they mention). The manual does NOT mention what to do if you dont have 'good' voltage.....

I am still suspect of the pump itself due to what i percieve is a low resistance across the pump leads....this implies (to me) that there might be a short in the pump.

However, since i do not have 'good voltage' at the pump, there might be something else wrong in the fuel pump driver module (which is the only thing between the inertia switch and the fuel pump in the wiring diagram). I have no idea where this module is or what it looks like (but i am guessing that it is somewhere under the steering wheel.

any insight is much appreciated,
dllewell

Ah yes, you did mention the inertia switch - sorry, I missed it twice when I read your initial post.

Speaking as an electronics engineer, I'd say you've made a good series of tests and I would say you're on the right track. Provided that you're sure that you have positively identified the motor DC feed terminals then I think that powering it directly would be a good idea. I've never measured the current into a fuel pump but I'd be surprised if it were to be any more than one amp or so, so I'd use maybe a three or five amp inline fuse for protection during the test. I wouldn't have the charger on the 10 amp setting - I'd select the lowest current setting that you have - I'd guess that a couple of amps would be plenty.

If I knew what the current consumption of the fuel pump was then I'd be keen to connect a dummy load in its place and see how the feed voltage then held up under load. I'd select a suitable low wattage 12V lamp to serve as a dummy load, and illumination of the lamp accompanied by a healthy voltage reading would indicate a faulty pump. Conversely, if the lamp failed to illuminate and the voltage collapsed then it would point to a fault in the driver module. That's a bit academic at this point though because without any knowledge of the fuel pump current consumption it's a bit of a gamble to select a lamp wattage at random. If the current is being fed via relay contacts then it would be safe enough but if the current is fed via a series pass transistor or similar then you could easily exceed the current capabilities of the device and damage it. As I say, it's a bit academic but I'll leave this paragraph in as food for thought in case it triggers any further ideas.

A resistance reading across a known good pump would certainly help but that may not be so easy to obtain eh.

I think then, all things considered, that I'd go for feeding the pump directly from a 12V supply as the next step. If there's any fuel in the vicinity then make the 12Vdc connections to the pump first before applying power to the power supply - that avoids any sparking at the pump terminals. And then when disconnecting power after the test, pull the mains connection first and the low-voltage 12Vdc connections last, again to avoid sparking.

Ok...now we are getting somewhere :)

I believe that your assumption about current draw is wrong. For example, the fuel pump circuit is protected with a 20 A fuse. Furthermore, in various places on the internet there is mention of the current draw required for today's high pressure fuel systems (55 to 85 psi according to my manual). These numbers vary but are all in the range of 4 to 9 Amps, with higher draws being associated with higher pressure systems as well as systems that are not operating properly (eg. clogged fuel filter).

I am not sure how measuring the current draw will help me in this case since i dont know what it is supposed to be like.

Thank you muchly about the advice re. hooking up an alternative power supply. Though there is no gas around the pump wiring harness, this is very good advice to follow whenever such a situation arises.

I will start with the charger on the 2 amp setting and will a small fuse in-line. However if this fails, how far do i go (how big a fuse) before i call it quits?

I have found a arts supplier with a new pump in stock, so i think i will purchase it at least to see its resistance. I could also connect it to the fuel circuit (but not install it) and see if turning the ignition to run activates this pump (just for a few seconds).

As an interesting side note, premature pump burn- out is often caused by clogged filter or a clogged or improperly installed in-tank strainer (so i have my suspicions about this as well since my tank was recently refurbished...perhaps the pump/strainer assembly was not properly installed).
Furthermore, when i bought the car, the person i bought it from just happened to have a spare OEM fuel filter that he gave me. Perhaps just a coincidence.... I have put off replacing this filter because it looks like a pain to replace (attached to the firewall behind the engine). But since my fuel system is likely depressurized, i guess now is as good a time as any to do that too! (sorry, off topic a bit and now i am rambling0

dllewell

Ah, I see - I've been driving only diesels for many years now and I'd no idea that modern petrol (gasoline) engines were fed at such high pressures.

There are two scenarios for feeding the fuel pump from an external supply:

1. If you externally feed your present pump, then it will still be in-situ and still attempting to feed fuel through the system, and thus working against a physical resistance, as normal. Therefore it will run at its regular current rating, so probably you would need to have your charger set to 10 amps.

2. If however you obtain another pump and run it but without actually installing it, then it will not be developing any pressure - i.e. not working against any physical resistance and therefore I would expect the current consumption to be a lot less. In that case I would set the charger to a 2 or 5 amp setting.

What you have said about premature burn-out due to clogging is interesting because that ties in with what I was just saying - the greater the physical resistance that it has to work against then the more strain that will be placed on the pump itself and the more current it will consume and therefore the greater the wattage which it will be dissipating - none of which are particularly good for the pump or the driver circuit.

If you're sure about the current consumption of the pump being probably in the range of 4 to 9 amps then if it were me I would be connecting a lamp as a dummy load, in order to observe what happens to the feed voltage. To achieve a current of, say, 2 amps, just for test purposes, then a lamp wattage of around 24 watts would be required (24 watts divided by 12.6 volts = 1.9 amps). But that of course would only be if you wanted to pursue that route.

since i dont have any low voltage bulbs laying around (like the MR11/MR16), i am not yet ready to pursue that route...but good information nonetheless.

I am quite interested in measuring the current draw however, unfortunately i do not have a benchmark (ie. i dont know what it should be). Having said that, i think i will measure this current on whatever pump i do manage to get running both before and after i replace the fuel filter...a fun little test!

Wish me luck, going back at it tonite fully armed with a 20A digital multimeter and a fuseable link (With 10 A fuses).

I will post my results tomorrow

dllewell

One thing which does occur to me is that a bad ground connection for the pump could also cause the symptoms described. The pump will not be grounded physically via its casing, for safety reasons, but instead the ground connection will be achieved via one of the leads coming out of the six-way connector - I'd follow that lead back to its grounding point if possible and unbolt it and make sure it's clean and free from either rust or oxide. Then prior to reassembly I'd apply a coating of what we (in the UK) call petroleum jelly (Vaseline) or similar - whatever you use over there to protect your battery terminals.

The reason for that is: with the pump out-of-circuit, and measuring the voltage on the feed end of the connector, you are reading a healthy 11.8 volts. When you put the connector back in place and take the same measurement with the pump in circuit, the voltage sits at 7.8V and then falls to zero, and that could be explained by a high-resistance connection to ground.

I'll pick that up again in a moment but first I'll need to digress. In the Autozone notes it says:

"Turn the ignition key from the OFF position to the RUN position several times (do not start the engine) and verify that the pump runs briefly each time, (you will here a low humming sound from the fuel tank)."

That suggests to me that it's on a timer of some sort and therefore is designed to switch off and the feed voltage intentionally falls to zero after a few seconds.

So then, let's return to the voltage reading of 7.8 volts. You start out with 11.8 volts (open circuit) which falls to 7.8 volts when the pump is back in circuit, so I'm wondering where the missing 4 volts went to. If you have a near short-circuit within the pump motor then that could be causing it but equally possible is that the 4 volts are being dropped across a high-resistance ground connection at the chassis. The current enters the pump, goes through the motor windings and then makes its way out again via the multi-connector and heads for ground. Let's assume an operating current of around 6 amps. Now then, since voltage = current x resistance, resistance must equal voltage divided by current, so what resistance would it need at that connection to drop 4 volts?

Answer:

resistance = voltage divided by current

resistance = 4 volts divided by 6 amps = 0.67 ohms

So, it would take an oxidised/rusted ground connection of less than 1 ohm resistance to drop 4 volts at the ground connection, and as you said earlier - the remaining 7.8 volts may simply not be enough to power the pump. Therefore I would want to check out the ground connection. It does specifically mention checking that connection in the Autozone notes, and that will be the reason why. Sorry that the explanation is a bit lengthy but you obviously know your way around voltage, current and multimeters so I'm sure you will see what I mean. If I haven't made it clear though, just ask.

Good luck with your next tests, and if anybody following this has any direct knowledge of the vehicle and its fuel pump then please do jump in and comment - I know nothing about the system but my electronics knowledge may be of some help which is why I've joined the thread. Thanks.

if i were you i would just run a couple of wires directly from the battery(ground and power) to the pump(for testing only) it will only draw what ever amperage it needs(which since it hasn't blown the fuse its under specs). As long as its a 12 to 14 v source it should be fine.

A good idea is to jump the pump relay terminals at the fuse box to keep power going to the pump without having to cycle the key on and off. After doing that connect the harness back to the pump and with an alternate power source backprobe the pump ground you should have less than .5v anything greater would indicate excessive resistance on the ground side. If thats within specs backprobe the power side and you should read power source which should be battey voltage(check the battery to make sure it has good voltage)going into the pump anything less would indicate excessive resistance on the power side of the circuit.

Thanks again for your suggestions.

Last night i got the fuel pump to work again. Firstly, i took the battery out and put it on a trickel charge (2A) while i replaced the fuel filter. This was a no-brainer since the batter was a little low on voltage after all fo the cranking that i have been doing. As well, the only possible way to access the fuel filter is by removing the battery shelf. The filter was pretty darn hard to get to but once i got my hands on it, the replacement went smoothly. BTW the gas that came out of the pump side of the filter was an opaque brown dirty mess, but performing the blow test (just blowing on the inlet) informed me that it was not clogged....just doing its job i guess!

The battery terminal were quite corroded so i cleaned thes up as well on installation (hoping that a clean, charged battery might be enough to convince the pump into action...not so).

Anyway, then i started checking the voltage at the inertia switch and at the pump wiring harness and got nothing (where there was voltage before). This was a puzzler for a while till i checked the fuel pump fuse located under the engine and found that it had blown...this was odd, it either happened at the end of my previous diagnostics or after i reinstalled the 'refreshed' battery.

Anwyay after replacing the fuse, i performed the same diagnostics as before and wound up witht he same results.

BTW, Selectron, the when in the run position, the fuel pump comes on for a few seconds and then shuts off again until some sensor tells it that the engine is actually spinning (thats why the voltage dropped to zero after hitting 7.8).

My next step was to connect the pump directly to my battery charger (heres where things interesting):

First time through, i had it ont he 2 A setting and the pump activated (could hear the hum) but was definitely not turning. SOunded like a stuck motor (like when an electric drill gets overloaded and cannot spin the drill bit). I activated/ deactivated the charger several times (3 s each time). Then the in-line fuse popped (10 A). Hmm.... not working yet, but getting there.

Then i replaced the in-line fuse, and tried again, this time with the charger on the 10A setting. And what do you know? Off the pump went! Cool!

Then i reconnected the pump to the car, turned the ingntion and off she went. Initially, there was some stuttering but i think this was probably while the filter was filling up etc. I tried it a couple of times and it worked like clockwork...turn the key to run and the pump comes on and then shuts off in seconds, and then she started right up.

I dont think i am out of the woods quite yet though. WHen i tested the voltage at the pump (when in run) now it went up to about 10.5 before dropping to zero. Then i wired the multimeter in series to measure current and found that during that initial few second burst, the pump draws 4.1 A.

It was getting late so i quit at this time, though i think i should have gone 1 step farther and measured the 'running' voltage and current draw (ie. while the engine was actually running).

So now what? Sure the pump runs, but for how long? What caused it to stop working? Was it something mechanical (like a blockage or contaminant inside the pump), or is there something wrong electrically (like is there a 'dead spot' in the pump motor...and will it get stuck again?).

Should i be replacing the pump, or should i take my chances in getting stuck in a parking lot? I was very fortunate to have this failure with the car parked in my semi-warm garage (its -10 C here in Canada), what are my chances of the next failure happening while parked int he garage.

Will the pump work for an hour, day, week or year?

I know that there is probably definitive answer to this but i would like to hear your best guesses based on the synptoms that i have described.

BTW, Selectron, I tested the resiatance of the new, aftermarket pump thaht i bought and it was the same as the current pump 1.1 Ohms.

Thanks so much for your advice, I hope you can take the time to read this last post and make a suggestion on where i should go from here.

dllewell

I would be suspicious of your battery, it seems to be dropping quite a bit of voltage for such a low current draw. Low voltage will prevent relays from switching.

I would be suspicous of your battery, it seems to be dropping quite a bit of voltage for such a low current draw. Low voltage will prevent relays from switching.

I am suspicious of my battery, there was quite a bit of corrosion at the negative terminal. Having said that, turning the key to run was enegising the fuel ump circuit, and once i got the pump to run 'remotely' it now runs properly when i turn the key to run.

That's good news, dllewell - pleased to hear that your efforts were rewarded with success.

Regarding the motor not running when the charger was on the 2A setting - that's obviously because not enough current was available, with the motor then running okay when 10A was available. I would be tempted to say that your problem all along has been the corroded battery terminal - I assume the missing 4V was being dropped across that resistance - however, that doesn't explain why the pump still refused to run, after cleaning the battery terminals. That's a bit of a head-scratcher and it would point towards a dead spot in the pump motor, the spindle of which perhaps rotated slightly (away from the dead spot) when it was physically disturbed during the removal process.

The fact that you now have 10.5V at the pump, where before it was 7.8V would reinforce the belief that the missing voltage was being dropped at the corroded battery terminal.

Regarding the measured current of 4.1A - I assume that was measured with a digital multimeter and I've found that they always read low for current measurements - if I want accuracy then I always use my old moving-coil analogue meter. Therefore I'd say even though you've measured it at 4.1A I'd guess that in reality it would be closer to 5A or 6A (the series resistance which the meter introduces to the circuit accounts for the lower measured current - this series resistance is always lower in an analogue meter for some reason).

If the pump had operated correctly first time, after cleaning the battery terminals, then I'd say you had solved the problem and could leave it at that but unfortunately you are still left with the possibility of a dead spot in the motor. It might take weeks or months before it came to rest on the dead spot once again, and there is the possibility that it could leave you stranded anywhere, anytime, following a long period of apparently reliable operation. Therefore if you have actually purchased the replacement pump which you mentioned than I would fit it for sure, and clean the old one and store it in a dry location as a possible spare for emergency use only.

12Ounce makes a fair point - I'd keep an eye on the state of your battery, although if it's beginning to fail then I would have expected you to be having problems running the starter motor before you had problems running the fuel pump. Do keep an eye on it though.

thanks yet again Selectron,

Everything you have said has been uber helpful. And i concur fully that if the battery was indeed the problem then the starter motor should be the first thing affected.

your comment about the charger at 2A vs 10A....if there wasnt enough current to run the motor at 2A, why then did i blow a 10A fuse at 2A?

cheers

I shouldn't pay too much attention to the fuse because I'm sure it would prove to be a red herring and would only cloud matters. I'll ignore it for now and add a footnote by way of what I think is the likely explanation.

Whilst I was cooking dinner, I was thinking about your recent fuel tank repair, and about your fuel filter and pump. Which side of the pump (in terms of direction of fuel flow) does the filter sit on? Is it:

fuel tank -> to filter -> to pump -> to engine

or is it

fuel tank -> to pump -> to filter -> to engine?

The reason why I ask is that if it's the latter arrangement and fuel enters the pump unfiltered, then there exists the possibility that a piece of swarf/debris left in the tank following the repair could have physically jammed the pump. Given the direction of fuel flow and the physical design of the pump inlet port, would that be a realistic possibility?

That's the only further thought that I've had so I'll just finish off with what I think probably happened to your fuse.

I would say the fuse failed because of fatigue (very common), and specifically it was probably a voltage spike which killed it. Whenever you make or break the circuit to an inductive load (in this case the pump windings would possess significant inductance) you generate something called back-emf, where emf is electromotive force, more commonly known as voltage. This voltage can be very high indeed (hundreds or thousands of volts, even when being fed from, say, a 12V source) and can be calculated from the formula e = L x di / dt, where e is the generated voltage, L is the inductance of the device (pump windings), di is the change in current and dt is the time taken for the current to rise (if switching on) or fall (if switching off). Now let's put some figures into that:

L = 1.5 H (I'm taking a guess that the pump has an inductance of 1.5 Henrys - the Henry being the unit of inductance)
di = 2A (that's the change in current - rising from 0A to 2A at switch-on or falling from 2A to 0A at switch off)
dt = 0.01S (that is 0.01 seconds - just a guess at the time it took for the current to rise from zero to 2A - or fall if switching off)

so:

e = 1.5 x 2 / 0.01

e = 3 / 0.01

e = 300V

The figures I've used are just guesswork but a voltage spike of 300V is by no means unreasonable, and probably such a spike just proved to be the last straw for your fuse - even though it was well within its 10A rating, its voltage rating has probably been exceeded once too often by a voltage spike, or back-emf.

That's what causes the bright blue sparking which you will often see when connecting a voltage (even a low voltage) to a load - it's the high-voltage back-emf thereby induced which is capable of jumping the gap between the terminals which are in the process of being connected. You hear it around the house too - a little 'pop' from the speakers of a stereo system perhaps as the fridge or the washing machine or the electric light is switched on or off - the voltage spike travels around the house wiring and enters the stereo system and is able to pass through the elementary mains filtering and into the amplifier audio section, generating the characteristic popping noise from the speakers which you've probably heard before. Well that turned into a long footnote but I would round off by saying that I really would ignore the fuse failure.

Wow... some very, very thorough stuff here.
My two cents...

You mentioned the appearance of the fuel on the pump side when you replaced the filter... after having the tank removed and who knows what done to it... I wouldn't be surprised if things got sludged up and the pump was simply clogged. In the process of removing it and all the tinkering perhaps it became unclogged and decided to start working again :)

I would be very concerned about my fuel supply if it ever came out looking like you described. If it happens again then you might want to take a better look at the condition of the fuel in the tank and find out what kind of gunk it might be sucking up out of there.

For the last 2 days, the car has been running fine...though i am still cautious ans now i carry all the necessary equipment to connect the pump directly to the battery in case it needs another 'jump'.

Both of the last 2 posts mention the recent work i have had done to the tank as well as the obvious sludging at the filter.

The line follows: tank-strainer (mesh)-pump(in tank)-filter (bolted to firewall)-engine. As such, there is a high likelyhood than any swarf (i love that word!) couold get into the pump...in fact i am leaning towards this possibility.

Now the tank had pinholes in it and when it was repaired, the rotten area was cut out and fresh metal was welded in. Then the tank was coated inside and out with some sort of epoxy.

In order to do any welding on a tank, they have to wash it out pretty good... so i imagine it was pretty clean but who knows what was still in there when it was reinstalled. Furthermore, perhaps the mechanic installed the pump improperly (like without the screen, or there is a rear in the screen.) I wont know unless i pull er out.

My brother-in-law, who is a mechanic has also suggested that letting a pump sit dry for a while can bugger it up as well (whenever he removes a pump he always seals it in a container with some gas). Chances are that my pump sat out on a workbench for at least 24 hrs.

He has recommended that i replace the pump, but without clearly knowing whether it is a mechanical blockage or an electrical issue i am hesitant to spend the dough (aftermarket the pump is about 250$ Canadian).

Thanks for the lesson on back EMF Selectron :)

D.

I'm pleased to read that it's still running well. I've been thinking about the age of the car - 1998 - if the fuel pump had a dead spot then it's had plenty of time to show up before now so I've moved away from thinking of that as being a likely cause.

I think it's too much of a coincidence that this has happened just a short period after the tank repair so I go along with what mightymoose_22 said, and my best guess would be that either a piece of debris or just a build-up of sludge has brought the pump to a standstill (not helped by the corroded connection at the battery terminal).

Hopefully, the worst of the sludge will by now have made its way through the pump and been caught at the (now discarded) fuel filter, and hopefully from here on in the fuel system should be cleaner and more reliable. However, given that this problem took one month to show after the tank repair, I'd be tempted to remove the fuel filter again one month from now and take a look at the input side and see how it compares in terms of cleanliness to the one which you just replaced - hopefully it will be a big improvement but if not then you will know that you have problems back at the tank and you may be able to do something about it before the system fails again. I know that you said the filter is awkward to access but I'd say that it would be worth the effort just so you can get an idea of the state of the fuel system.

Last night i did some electrical testing:

Battery:
1. sitting 12.8 V
2. key in run position 12.4 V
3. cranking to start 10.5 V
4. car running 14.2 V

Fuel pump:
1. key in run position 8.5 V, 4.0 A (drops to 0 V, 0 A)
2. cranking to start 7.6 V, 4.0 A
3. car running 7.6 V, 4.0 A

When i open the throttle (step on the gas) the running voltages for battery and fuel pump increases, however the Fuel pump Amps remains relatively constant.

Remember that i did see upwards of 10 V to the fuel pump at least once, but now it seems to be quite constant...and its roughly but not quite half of the running voltage of the battery.

This surprised me. Perhaps the FP is not meant to run on 12 V. Or perhaps there is another reason for the Voltage drop (other componentry, poor wiring, bad ground?). I wish i could measure someone else's car!

D.

My diagram (for the '97) indicates that the fuel pump is to get a full 12 volts. You have something impeding current flow. You mentioned a blown fuse ... maybe it blew from heat from a dirty base ... not from current flow. Perhaps a dirty relay contact or base.

I would start out at the fuse #15, measure both sides ... and then step back through the wiring all the way to the pump. As soon as the voltage drops away from the battery voltage, you know you are close to the problem.

Fuse #15. .... to
BL/Y wire to ter #11 on CCRM. ...to
Fuel Pump Relay. ....to
Ter #5 on CCRM to INERTIA FUEL SHUT OFF SWITCH via BK/PK ...to
1)INERTIA FUEL SHUT OFF SWITCH to ter #5 and ter #6 Fuel Pump (module) via BL. ...and to
2) FUEL PUMP DIODE via BL.
.
Both the diode and the pump are grounded thru BK. The diode, in electrical "parallel" with the pump, seems to be harness mounted near the pump.
.
Good luck.

Excellent input there from 12Ounce, which I'll add a few comments to, all based on the assumption that the pump should indeed be fed at the full supply voltage, which with the engine running you have measured at 14.2V

The reading taken directly across the pump connector is 7.6V so you have lost 6.6V somewhere in the circuit, due to either a fault condition or a high-resistance (corroded) connection. My first step would be to determine which side of the circuit the loss is happening on - the incoming feed from the battery positive, or the return path back to the battery negative, because it could be on either (this is what carlos80 was referring to when he talked about 'backprobing' in both directions).

It would make things easier if you had a length of wire maybe 12' in length, with which to extend the physical reach of one of your multimeter leads (just one - not both). If you have that, then I'd advise the following (all readings taken with the engine running):

Measure voltage at fuel pump connector and verify that it's still 7.6V

Now place the meter negative lead onto the 12V positive feed terminal at the connector, and extend the meter positive lead back to the battery positive terminal and, probing directly onto the terminal post, take a note of the voltage.

Next, remove the extension lead and connect the meter positive probe directly to the fuel pump negative connector, and this time extend the meter negative lead back to the battery negative terminal and, probing directly onto the terminal post, take a note of the voltage.

As carlos80 said, you should expect to find little voltage on either side - say 0.5V maximum. However, we know that you will find 6.6V on one side or the other, or perhaps the missing voltage will be split across both sides.

If you find 6.6V when measuring back to the battery positive terminal then your problem is in the supply feed side of the circuit and you should start probing back through the feed path - that path being battery positive to relay to inertia switch to 6-way connector to pump. Somewhere along the path will be a fault or a high resistance connection of the order of 1.65 ohms, and we know that because:

resistance = voltage / current
R = 6.6V / 4A
R = 1.65 ohms

If, on the other hand, you find 6.6V when measuring back to the battery negative terminal then your problem is in the negative return side of the circuit and you should start probing back through the ground return path - that path being from the pump 6-way connector ground terminal, to the bodywork, to engine block, to engine grounding strap, to battery negative terminal. Somewhere along the path will be a fault or a high resistance connection of the order of 1.65 ohms (same calculation as above).

If there is a problem in both the feed path and the return path then you may find, for example, 4V on one side and 2.6V on the other side (but totalling 6.6V), in which case you will have to probe back through both the supply feed path and the negative return path in order to locate both of the high resistance connections, because there will be one on each side.

One thing has just occurred to me, which is that if you have dirty relay contacts then they could be causing this. I'm not talking about the externally accessible terminals, but the internal conductors which actually make and break the circuit - if they are dirty or pitted then they could be causing the voltage drop. This, however, would be revealed by probing the feed into, and the feed out of, the relay - the output side would be 6.6V lower than the input side.

I hope you follow that and I hope it will be of some assistance. Any questions, just ask.

Edit: Just to clarify - when I talk about probing back looking for the high resistance connection, I'm not talking about taking resistance readings on the resistance range of your meter, because you cannot take resistance readings on a live circuit. Rather, I'm talking about taking voltage readings, and if you found, for example, a certain voltage going into the relay but a lower voltage exiting the relay, then that would, by implication, reveal a high-resistance contact - the readings would all be voltage readings though.

Edit #2: I just remembered that your battery is in the trunk (mine is under the hood). This makes things easier because you shouldn't need to extend the meter lead by any great length. Also, your negative return path will presumably be from the pump 6-way connector ground terminal, to the bodywork, to battery grounding strap, to battery negative terminal in trunk.

WOW, again excellent suggestions, and very well written, i understand completely. My grandma died yesterday and i am off halfway across the country to the funeral, back Tuesday. I will try this tonite but if i dont get to it then please check back here mid-next week and see how things went.

3. points: I still dont know where the fuel control relay is located, i was hoping someone would catch that in earlier posts and offer some helpful advice.

the battery is indeed under the hood (front of the car) but i can find an adequately long wire for this testing.

One of the wires at the fuel pump actually is grounded to the body of the car right there. However, this is not apparently the gorund for the fuel pump circuit. According to the circuit diagram, the return wire goes back to the fuel pump control module...and then to ground i guess

thanks, D

Sorry to hear about your grandmother. I hope you will have a safe and uneventful journey.

I'm wondering now about that fuel pump control module. I'm reluctant to introduce a hypothesis in case it muddies the waters but now that I've thought of it I'd better mention it in case it proves to be relevant. Consider this: with the engine idling, fuel consumption and therefore fuel flow rate will be minimal, so not too much effort will be required from the pump to keep the fuel system up to normal operating pressure. However, when driving at speed along the highway, fuel consumption and flow rate would be considerably increased, and I wonder therefore if this is a smart module with a feedback sensor, which can adjust the fuel flow rate according to the current driving conditions, by varying the voltage applied to the pump - reduced voltage at idle, increasing towards full supply voltage at greater speeds.

If that were the case then it could explain why you have measured the voltage at the pump (whilst operational) at both 7.6V and at 10.5V - perhaps the reading at 10.5V was taken after you had disturbed the fuel system so the pump would be working harder in order to refill the system and regain normal operating pressure with, perhaps, the supply voltage falling back to its idling voltage of 7.6V after the system had been re-pressurised.

That's all hypothesis, and I may be forcing the recorded voltage readings to fit in with a theory but it is a possible explanation of why the pump voltage might legitimately fluctuate intentionally, by design. Unfortunately, I can't prove or disprove it for you because my Escort is diesel-engined.

I would recommend that you measure the voltage, as outlined earlier, from the 6-way connector positive terminal back to the battery positive terminal, and then from the 6-way connector negative terminal back to the battery negative terminal. You will then know on which side of the system the missing 6.6V is being lost and you can then set about identifying which section of the system it is being lost in.

Yep, I'll check back for sure because I'm keen to know what the final conclusion will turn out to be.

Meantime, if anybody reading this has any knowledge on that fuel control module/fuel control relay - its operation or even its location then that would help a lot. Thanks.

Double-posting here so read the post above this one first if you haven't already seen it.


You know what? I think my hypothesis is the answer - I think the fuel pump is designed to run from a variable supply voltage - varying according to current driving conditions. There is a clue in your earlier post (post #19):

"When i open the throttle (step on the gas) the running voltages for battery and fuel pump increases, however the Fuel pump Amps remains relatively constant."

The battery voltage increase will just be an increase in voltage output from the alternator - that is to be expected so don't pay any attention to it. The significant point is that when you open the throttle, the fuel pump driver control module responds as we might expect - it increases the voltage to the fuel pump to keep pace with the demand for more fuel.

I did a search on Google and I found the following. It isn't from a technical article - it's just something which somebody quoted on a forum elsewhere, referring to a '98 Escort with a fuel supply problem:


Fuel Pump Driver Module
The fuel pump driver module (FPDM) receives a duty cycle signal from the PCM and controls the fuel pump operation in relation to this duty cycle. This results in variable speed fuel pump operation. The FPDM sends diagnostic information to the PCM on the fuel pump monitor circuit. For additional information, refer to PCM Outputs, Fuel Pump and PCM Inputs, Fuel Pump Monitor in this section.


In light of that, I wouldn't even bother taking any more voltage readings, because I suspect you're chasing a fault which doesn't exist - I think the voltage varies by design and that's all there is to it. Provided that the vehicle performs well at higher speeds (high fuel flow rates) and provided that the engine is not hesitating and shows no signs of being short of fuel, then I don't think you need worry any further, beyond of course bearing in mind that the contents of the fuel tank remain in question, because it was almost certainly debris or sludge from the tank which caused the initial malfunction in the fuel pump. Beyond that though, I don't think you have anything further to worry about.

Thanks yet again Selectron,
It would be truly helpful to have another one of these cars to compare with!

I had, in fact, considered the possible of variable voltage (not seriously though). I believe that some DC motors are much more capable of handling variable voltage than AC motors (perhaps a gross generalization).

I cannot hear the 'speed' of the fuel pump while running at speed, but theres nothing stopping me from hooking the multimeter up across the pump leads and observing how increased loads (fuel requirements) affect the voltage. (i will have to extend the leads of the multimeter a bit to make this a safe venture.)

This, too will have to wait till next week.

cheers, D.

I'm sorry to hear about your loss. I never knew any of my grandparents, I was born too late.

Don't get caught up in Ford semantics. The "fuel pump module" on your model is that thing in the tank. It supplies fuel until an internal pressure level is satisfied ... that is, as long as it gets a 12 volt source. I think there is a primary pressure regulator right there at the pump, and of course, the secondary vacuum-assisted one on the fuel rail.

In "Ford Speak" there are "door control modules" ... use to be "door regulators", etc, etc. ... all kinds of "modules".

On later Ford models, there is no return fuel line from the fuel rail to the tank ... then control items get a little trickery. More sensors. Different power forms. Does your fuel rail have the regulator and return fuel line on it?

12Ounce:

there is only one fuel line attached to the pump itself, however there may be a return line that goes directly to the tank, at least this has been suggested to be the case.

so you recommend that i dont go poking around under the dash for this mysterious control module?

furthermore, this would be contrary to Selectron's theory that the fuel pump is controlled by variable voltage. care to comment?

I dont know what to look for under the hood in terms of a regulator/ return fuel line on the fuel rail.

I still plan to hook the multimeter up and go for a drive and see how load (fuel requirement) affects the voltage across the fuel pump.

I am also very tempted to try to find another 98 Escort at a used car lot and go for a 'test drive'...ie. back to my garage for the multimeter test.

d.

I don't have a diagram for '98, but I do have one for '97 (somewhere, can't lay hands on it at the moment). The '97 is very straight forward ... 12v is shown at the in-tank module. I would be very surprised if the '98 changed this feature that much...but would not bet on it!

The "secondary" (my term) regulator is mounted right on the fuel rail that feeds the injectors. At the belt end of the engine. It has a small vacuum hose attached. Do you see this gadget?

Your idea of comparing another '98's voltages is right on target!

OK, finally did the fuel demand vs. voltage test.

Hookekd the multimeter to the 2 leads feeding the fuel tank and went for a drive. The voltage definitely varies. It varies from 6.9 to 9.9. At idle, the voltage is pretty much stable at 7.4 to 7.7. On the road, the voltage increases during acceleration or while trying to maintain the same speed while going up hill. If i let off on the gas, push in the clutch or even gear down (to slow down) the voltage generally drops, and it drops right away.

It is now clear to me that voltage is tied to fuel demand (engine load). It is not specifically tied to engine speed or road speed. You can gun the engine under no load (neutral) or low load (first gear) and the voltage doesnt go up that much. However, i did accelerate up to about 100 kph in 3rd (uphill). I dont have a tach but based on engine noise and vibration i'm guessing that i was up between 5500 and 6000 rpm. This is when it hit 9.9V. (i suppose i could calculate the engine speed for a given gear and road speed...knowing the gear ratio and tire diameter).

So what do we know now thats new? Voltage varies (proportional?) with fuel demand. I was however disappointed that it only got up to 10 V, even at 140 kph. Not disappointed though, that there is no apparent hesitation at this speed, suggesting that the fuel pump is capable of keeping up with fuel demand.

as for 12Ounce's info on the fuel control module. The 97 and 98 are completely different according to the wiring diagrams in Chilton. I am still hoping that this variation in voltage is by design. And if so, I assume that the fuel control module is to blame...hence it is located somewhere upstream of the wiring harness that connects to the top of the fuel pump (and not inside the tank with the pump).

so far unable to get my hands on another 98',

still considering backtracing the circuit as well...but i need to know where this darn fuel control moduel is!

D.

Welcome back, dllewell. I've been following the thread but I didn't really have anything to add. Well, reluctant though I am to disagree with 12Ounce (and I'll be happy to be proven wrong on this, just so long as we solve the mystery!) I have to say that that does look like a variable-voltage pump. You've taken a comprehensive, meaningful set of readings, and all of them point towards the voltage varying by design. I say that because:
Voltage increases under acceleration (as one would predict)
Voltage increases under increased load (hill climb - again, as one would predict)
Voltage decreases under deceleration (as one would predict)
No engine hesitation has been observed
At engine idling speed, the battery voltage may fluctuate somewhat, probably at thirteen-point-something volts, but shortly above idling speed, the voltage regulator will take over and clamp the voltage at 14.2V (as recorded earlier in this thread). Therefore, with the voltage firmly clamped at 14.2 volts, there is no reason for the voltage at the fuel pump to be fluctuating, unless by design. Even if there were a fault condition which caused it to fluctuate (e.g. a connection which was both corroded and also loose), it's beyond the bounds of probability that the voltage fluctuation would mimic precisely the behaviour which we would expect to see from the feed to a variable voltage pump.

As for being disappointed that the voltage only reached 10V, and not the full supply voltage, I would make two points.

Firstly, if it's a variable-voltage pump then it would not necessarily follow that the upper voltage limit would be set at the same level as the full supply voltage - the upper voltage may have been determined by electronics circuit design considerations within the control module.

Secondly, although you are only seeing an increase in voltage of around 44%, from 6.9V to 9.9V, that is only half the story. Bear in mind that along with the voltage increase, the current flow will also increase and therefore the power dissipated at the pump will increase - but by more than 44%. I'd guess it goes something like this:

6.9V x 4.0A = 27.6W
9.9V x 6.0A = 59.4W

So, although the voltage at the pump only increases by less than 50%, the power (wattage) at the pump increases by around 100%, i.e. it doubles. That looks a lot more reassuring eh.

All things considered I'd still come to the same conclusion as before - the pump is variable-voltage by design, and probably (although we have no way of proving it) operating within specification (because you have no reason to believe otherwise, because the engine performs well at all times) and the problems have all been caused by physical contamination within the fuel tank following the tank repair.

If you still want to investigate further, then I'd suggest you do the two checks which I outlined in post #21, measuring first the voltage between pump connector positive terminal and battery positive terminal, and then the voltage between pump connector negative terminal and battery negative terminal. On one side or the other - probably but not necessarily - the positive side, you will get a reading of around 6.6V - in the event that the pump is indeed variable-voltage then this tells you which side the control module is on and you can set about locating it by taking voltage readings along the electrical path. Or, in the event that it's a fixed-voltage pump intended to be fed at full supply voltage, then it tells you on which side of the circuit your fault lies and you can set about tracing it.

There is one exception to that, which is that the pump may be fed from a dedicated, isolated supply, in which case the voltage will be 'floating', meaning that it will not be positive with respect to battery negative, nor negative with respect to battery positive, but just electrically floating, in which case you would get nonsensical readings when taking the voltage readings in post #21. That would be a complex piece of design though and I doubt that is the case, but if you get completely unexpected, illogical readings then that would almost certainly be the reason why.

The only hint I have as to a possible location for the pump driver module (if it actually does exist) is a snippet which I found which refers to a '98 escort sport', whatever that may be. It says The fuel pump drivers module is located inside the vechicle, behind the Left hand wheel housing. It's on another forum and it would probably be impolite to be posting links to other forums here so I'll send you the link via PM.

Well, I'm am happy to learn this info. '97 was such a year of big change in the Escort I am surprised the fuel system changed again in '98.

I now suspect your fuel rail does not have the regulator. Did you have a chance to look for it (just for "info")? Do you see a pressure sensor instead?

I took a look in "Powertrain Control/Emissions Service Manual" for 1999. This shows three different fuel systems (return, mech returnless, & electronic returnless) used in all gasoline USA Ford cars/trucks for that year. Under the "electronic returnless" system description, it describes a "fuel pump driver module" (FPDM) that varies voltage to the fuel pump module (FPM) based on control info from the PCM.

This must be the system on the '98.

I've tracked down a couple of photos of the Fuel Pump Driver Module (FPDM) for the Escort. One is a clear photo which applies to a '1998 Ford Escort SE'. I don't know what the policy is here on posting links, so again, I'll send them in a PM.

For the benefit of anybody reading this in the future, it's a small black box which looks like it measures less than 4" high by 3" wide by less than 1" deep. The black box looks plastic, although it may be metallic. It has a bare metal mounting lug at bottom left and another at top right, and it has a large white label attached which, helpfully, says 'FPDM' in large letters, with a Ford part number beneath, and a large barcode beside the lettering. In the clear photo, the part number is F8CF-9D372-AC and in the blurred photo it would appear to be F8CF-9D372-DA. At the bottom of the unit, an electrical connector, around 1" in width protrudes downwards from the enclosure. Presumably, the wiring harness plugs onto this connector. I came across a mention of the FPDM which said that the location in a '98 Escort Sport' is 'inside the vehicle, behind the left hand wheel housing', although that is as yet unconfirmed.

12 Ounce:
thanks for the info on variable voltage...that seems to fit our assumptions/ theory. It would sure be nice to know over what range that voltage is supposed to vary....

Selectron:
As always, your help is greatly appreciated. I too am beginning to feel more and more like the inital problem was an obstruction in the pump. Curious that it would occur just as i pulled the car into the garage and shut it off (since i first noticed it when tryign to start the next day). Perhaps this is due to the whole static friction vs. moving friction phenomenon?

If i get a chance tonite, i will do at least the first step of back-tracing that you have suggested (measuring voltage from battery + to fuel pump connector + and battery - to FP connector - (though i am currently uncertain how to interpret the results).

I am still hesitant to try to pull the pump out of the tank. Seems like a fairly straight-forward process (though i dont have the correct tool for removing eh locking ring on the pump housing/ fuel tank). It just seems like it could get messy. Generally speaking, i dont like to fool with fuel. Especially when it is doen from inside the car!

i am slowly ragaining confidence in the reliability of my car,
D.

Interpreting the results is straightforward enough because there are only five possible sets of results, the last two of which are highly unlikely:
1. On the positive side, you record 6.6V, with 0V (up to 0.5V is okay) on the negative side
2. On the negative side, you record 6.6V, with 0V (up to 0.5V is okay) on the positive side
3. On the positive side, you record some voltage, and on the negative side you record some voltage, which when added together total 6.6V
4. You record 0V on the positive side, and also 0V on the negative side
5. You record some random, perhaps unstable voltage on the positive side, and similar on the negative side, which when added together are nowhere even close to 6.6V - could be way above or way below
Interpret those results as follows:
1. This means that the control module is on the positive side of the circuit (if variable-voltage pump), or the high-resistance fault is on the positive side of the circuit (if fixed-voltage pump)
2. This means that the control module is on the negative side of the circuit (if variable-voltage pump), or the high-resistance fault is on the negative side of the circuit (if fixed-voltage pump)
3. This would suggest that the pump is not variable-voltage but is instead fixed-voltage, and has a high-resistance fault on the positive side of the circuit, plus another high-resistance fault on the negative side of the circuit
4. This would suggest that the pump is variable-voltage and fed from a dedicated, isolated power supply, in which case no further checks are possible, beyond locating the control module and ensuring that it is physically well-secured at its mounting lugs, and ensuring that the connector from the wiring harness is firmly seated.
5. This would also suggest that the pump is variable-voltage and fed from a dedicated, isolated power supply, so interpret as for #4
In order of probability, when taking the voltage readings, this is what I would expect to see:
1. Definitely the most likely
2. Less likely, but still possible
3. I think this is very unlikely
4. Only a very remote chance of this
5. Only a very remote chance of this too

As a footnote to that, I would add this:

The voltage checks won't tell you if you have a variable-voltage pump or a fixed-voltage pump - they will only tell you whether the missing 6.6V is upstream from the pump connector positive terminal, or downstream from the pump connector negative terminal. In order to proceed further, you would then need to determine which type of pump you have, by locating the FPDM - and I'm convinced that it is the variable voltage type.

In the event that it's the fixed voltage type then tracking down the high-resistance connections is straightforward, by taking voltage readings along the path, looking for the point at which the missing 6.6V is being dropped.

In the event that it's the variable voltage type then you can only check that the FPDM is securely mounted and that the wiring harness connector is securely fitted. If you were able to positively identify, and access, the two feed wires to the pump, at the FPDM end, then you could verify that the voltage leaving the FPDM is the same as that which arrives at the pump connector. There isn't really any scope to test beyond that, without access to a spec sheet detailing what the normal voltage range from the FPDM output should be.

I've removed a few tanks. Always clumsy/dirty work. You never have enough clearance under the vehicle to work with ... unless its a truck. The tank doesn't weigh much when empty, but there will be road soil on top waiting for you. The lock nut-ring that clamps the pump module to the tank is easily removable with wooden/plastic (avoid sparks) block driven with a hammer. CCW to loosen.

The next time you have a punctured or rusty tank, you may consider driving stateside and buying one from AutoZone. They have to be ordered ahead. Not very expensive.

Here's some information, does support the idea that 1998-1999 Escorts had a different fuel system compared to 1997's. (There's an access hole to the fuel pump under the rear seat???). Some elect info, but not much.
.
http://www.autozone.com/servlet/UiBroker?ForwardPage=/az/cds/en_us/0900823d/80/13/c3/70/0900823d8013c370.jsp

So i found the FPDM, located as described (behind the rear driverside wheel well...just under the rear speaker. I am glad that Selectron posted soem pictures or i would have never found it...

I aslo ran some tests:

With the car on i measured voltage as suggested:
Positive Battery terminal to Positive side of fuel pump: 6.4-6.6 V
Negative Battery terminal to Negative side of FP: 0.1 to 0.2 V

Then i tested the FPDM (with car running):

Feed line coming into FPDM: 14.1 V (relativelyconstant regardless of engine speed)

Line exiting FPDM going to FP: 7.9 to 8.1 while idleing, increases to over 9V when i stepped onthe gas (in neutral).

It seems like this is in fact a variable voltage system with the FPDM controlling the voltage going to the pump. Since Selectron found one of these for sale on Ebay and since the bidding ended yesterday and nobody else bid, i would up buying an extra (cost me about 12$ CAD with shipping...well worth it to satisfy my curiosity).

Once i get it, i will make a comparison and post the results. As well, i think i have convinced myself to pull the pump out of the tank and have a look-see. This wont likely happen till January.

cheers till then,
d.

Thanks for the update, dllewell. The voltage readings are as expected, and confirm two things: firstly that the FPDM is responsible for the 'missing' 6.6V (missing almost certainly by design, but you'll be able to confirm this by substitution when you receive the replacement), and secondly we now know for sure that you have healthy return connections on the negative side of the circuit (from pump connector negative terminal to battery negative terminal). We know that, because resistance = voltage divided by current. The voltage was from 0.1V to 0.2V so we'll call that 0.15V (midpoint) and we know that the current has been measured at 4.1A, so:

resistance = voltage / current
R = 0.15 V / 4.1 A
R = 0.037 Ohms

That's a good ground connection, so no need for further checks on that side of the circuit.

Likewise with the supply feed into the FPDM - at 14.1V it is within 0.1V of full supply voltage, thus indicating good, low-resistance connections on both supply and return sides, so no further checks needed on that.

It was a stroke of luck, finding the FPDM at that price, so it will be money well spent in that it will put your mind at rest if it gives the same voltage readings as the original, which I fully expect that it will. You will then be able to concentrate on what will likely prove to have been the root of the initial problem, which is the contents of the fuel tank.

If it were me, as I said earlier, I'd check out that new filter after four weeks and see what state it's in, compared to the state it was in at four weeks after the tank repair. With a bit of luck, the worst of the contamination may have been filtered out by the old filter, and the new one may prove to be relatively clean once again. I hope so, because otherwise you'll have to put the multimeter away and roll your sleeves up and set about checking out the tank. I'll keep my fingers crossed that it doesn't come to that though and I hope it will continue to run well and reliably.

I finally tried the new (used) driver module that i Purchased from Ebay:

The voltage to the pump is virtually the same as with the old module as expected.

As far as i can tell, case closed.
Thanks for all you help selectron and 12ounce

D.