Thank you, Sir!!!

Hmm. Simple question. Could it just be the pump? I have what I believe came from the xk8 single, or, the xkr double pump design... As I've described, the guy who gave me this thing is an usual bird,,, has these pumps laying around.

​​​​​​All the readings I posted, are the readings. The pump runs strong when the ground (jimmied) is connected... As soon as I connect it, it dies... Car runs a bit then stalls... It must be the pump,,, can't run on a pulsing power supply on the ground side? But, in my little brain,,, that just doesn't add up...

​​​​​Reading here, I've heard people report (I think) 70psi at off (or sitting idle) I can't remember... And at running, about 55psi. Mine STAYS at 70... Can't be good. At the same time, I don't have a reading when driving.

I hate digging around in that damn tank, switching pumps and the line and basket, but I don't think I have a choice. Must be the pump... I just don't understand why it wouldn't turn with a steady 12v and a pulsing ground... To remind,,,, the OE pump was completely stuck, not spinning, gummed, so I put in the xk older style pump.

​​​​​

 

OK. How are you measuring the fuel pressure? The normal running is the 55 psi you've stated. If it's via OBD, then this is what the ECM is 'seeing' and it will signal the pump to back off as it thinks the pressure is more than needed.

Do you still see 70 psi when you come to the car 'cold' i.e. before you 'jump-start' the pump?

 

When the pump is off,,, it's upper 30s or just over 40... But nothing is being determined by the ECU, pressure sensor or PWModulation... It's residual pressure... Maybe held there by the 'buffer' in line on the left hand your side of the engine...?

I've never checked at cold, but can and will, tomorrow. I would guess it would be at zero,,, but, even in the cold, she fires right up... So nicely I get excited, lol... The exhaust mods are growlllly...
​​​​​

 

Mmmm. Possibly if it's not the correct one. Difficult to know what checks the FPM may do to determine if it's good or not (FPM can send a 'fuel pump inoperative' signal back via the 1Hz heartbeat).

If it would help I can produce some drawings that will show you what these PWM signals should look like viewed with a scope?

 

That would be interesting under any circumstances.

 

Jay did you mean what if pump designed not to work on less than 12volts?

If so test her out. Use a resistor in series with the line to ground. All you need to know is pump current draw and your desired test voltage (8volts?).

If you need a hand figuring out resistor size just tell us a number for current or say the fuse size.

If I misinterpreted the statement excuse me and just skip this.

J

 

Good Morning!

Thank you... I think it would be interesting to know. Hmm,,, all of this because I don't want to spring for the 'right' pump.

I tried the 3rd module. No change. Runs with make shift ground,,, pump stops, car will stall, without make shift ground.

In my thinking, it's like a light switch... In fact I've seen a few videos where an instructor will run a set of lights using pulse width modulation... Adjusting duty frequency (I think),,, the lights will go from pulsing to (what steams like) steady, even though there are gaps - they are imperceptible. All that to say, I don't see why a pump wouldn't respond (speed or slow) to being switched by cycling or modulation of the ground. But, as you know, I am a novice when it comes to all this.

I have seen a few examples (again, YouTube) but yes, for this specific application, I would love a diagram. I work better with visual learning and materials...

I have been so 'busy' with the car this weekend. But John (thanks) sent me some instructions on how I might use my DMM to read (or get close) what I am seeing here...

Thanks a bunch!

 

If all else fails, thats an idea, I think. What about heat? Will it grow at the pump connections in the tank, or only at the resistor outside of the tank AT the resistor?

I would like to see if I can get this thing working as it should, factory... A member from Staten suggested he would allow me to use his scope but I would have to do the install on PC, learning curve using it etc...

There must be an in tank PWM aftermarket pump out there somewhere. What a PITA...

 

Heat only at resistor. Just a fuel pump test suggestion.

J

 

Got ya, thank you... Much appreciated!

 

I've been trying to see if I can figure out how to use this damn DMM to read for Hz... IF the FPM sent an "inoperative" to the ECM,,, would I have a code? I've never seen any codes associated with fuel pump problems... I do have intermittent "restricted performance" and a CEL,,, but thought that was due to O2 sensor and lean running... It's like,,, the more I think I know,,, only teaches me how much more I don't know

With these cars, there is a huge difference between knowing that something WORKS (or doesn't) than knowing WHY it does or doesn't... So many mysteries in the Xk...

Even IF it burned the pump out prematurely,,, it seems that if a pump had 12v going to it, and was triggered by ground, it would run that way. On off with intermittent grounding.

Of one of the white wires is sending info back to the ECU,,, and the other white is taking the signal to turn ground of and on (the 3.xx),,, maybe 3.xx volts is simply not enough to trigger the FPM internal grounding mechanisms... Maybe it has to be 5 or something... Next time I remove the jimmied ground wire, I will have a the DMM connected to the 3.xx white wire to see if pulling the ground will LOWER at the rail pressure sensor causing the sensor to flux,,, causing the ECU to flux the 3.xx white wire at the FPM? OMG what a thing.

​​​​

 

I was reading up on the pulse modulated pumps and evidently a regular pump will not work (or work propely) in a pulse modulated system.

 

Hmm. Will it (a non PWM) work, at all, is the million dollar question...?

I have what I was TOLD would work (I haven't hooked it up yet),,, is a Walbro 255LPH pump GSS342,,, but have been reading which says it will not work,,, in other places... 75bucks...

I get a little frustrated but,,, thus far,,,, I am into this thing for a new AGM Battery,,, this damn pump,,, 6gallons of ATF (3 I haven't used as of yet),,, and some skinned up knuckles... I BETTER stay on the side of super grateful... What a beautiful beautiful car...

 

I've attached a pdf showing a representation of how an oscilloscope would display the FPM heartbeat (BT18-4) and the ECM control signal (BT18-3) in a normal situation.

Going back:
Reading at pin 4 does sound like it's pulsing.
Steady voltage at pin 3. Could be steady, or could be the meter averaging out the pulses. Would be useful to know if that varies with fuel rail pressure.

That is something I'm considering - if it is indeed a steady voltage. Looking for finer detail on how this works but I haven't encountered anything as yet. Of course, a working car to pull readings off would help

There are three or four P-codes relevant to this but the car isn't throwing any of them?

 

I found this in the 2003-xk-service-training-guide pdf in the stickies:

"
Fuel Pump
To support the installation of the 4.2-liter engines and continuing the drive for fuel consumption benefits, a new fuel system including fuel tank, fuel pump, pipes, breather pipes has been installed.

The XK now benefits from the use of the return-less fuel delivery system very similar to the X-Type and S-TYPE models. One major change on supercharged (S/C) variants is the use of only one fuel pump instead of the usual two as used on all previous S/C models (XJR, XKR and S-TYPE R).

The fuel pump uses the same principle of operation as the on the X-TYPE where the ECM monitors differential pressure across the fuel injectors and the injector pulse width and duration to accurately calculate the fuel quantity being delivered to the cylinders. It uses this to demand a specific fuel flow rate, which it communicates to a fuel pump driver module located in the right hand side rear wheel arch.

The ECM use a frequency of 150 Hz (PWM) signal during average conditions, varying its duty cycle between 4 and 50% to control fuel delivery rate or turning it to 75% to turn off the pump."


I have updated the pdf to show that duty cycle in the 150Hz section.

We know that the pump you have in situ isn't the correct one. If the FPM has also detected it, then it likely will tell the ECM. ECM could well respond by telling the FPM to shut the pump off.

It may be a coincidence, but your 3.73 volts measured on pin 3 is 74.6% of 5 volts.

Conjecture on my part, yes - but it fits the symptoms (lack of OBD codes excepted). Absent a correct pump to swop-test, my next move would be to scope those two control signals.

 

Damn.Damn.Damn,,,, Michael!
YOU.ARE.THE.MAN!!!

The 75% Find is amazing and I am thinking more than a coincidence... I ran out of sun the other day (and steam)... What I should have done and can't wait to do is, even if it's temporary,,, disconnect that white plug running across the top back of the tank and connect that Walbro Pump that's sitting in the box to R and YR. It's going to be one of those things where,,, if it works I'm going to feel REAL GOOD and REAL STUPID for having plugged along, as long, and the answer to my problems was right there! In argument for myself,,, I only got it Saturday.

The graphic and chart is amazing and gives me a visual representation of what to this point has just been left to imagination. It is helpful, and I am super grateful. Thank you...

WAS WRITTEN ~~~ The ECM use a frequency of 150 Hz (PWM) signal during average conditions, varying its duty cycle between 4 and 50% to control fuel delivery rate or turning it to 75% to turn off the pump."

SO,,, help me see if I understand it... Will it run at 4% when pressure is high or adequate (slowing the pump),,,, varying to 51% when pressure is low (speeding the pump) ---- ok,,, scrap all that... Q: what is the relationship between this percentage,,, pump speed,,, and pressure... At 4% for instance,,, is that when the pressure is high and the pump running slowly? What is 75% and what is happening real time to produce the number,,, and why would it be beneficial to SHUT DOWN at 75%? Pump running to hard?

Also,,, let's say a rail pressure sensor is bad... Does pushing a signal of 5volts to the ECU basically shut things down, as well?

I'm sorry, bare with me here...
​​​​​

 

Even the new pump,,, I'm learning, might be the wrong pump and may not give me any more clarity...

 

Please bear in mind that I've made certain assumptions (e.g. logic signals @ 5V not 12V) and done some (hopefully educated) guessing. It is one possible 'take' on what's happening.

That would certainly be logical, but it depends on how the FPM interprets it. The extra information in the training guide makes it clear that the relationship between duty cycle and fuel pressure isn't entirely linear, and why.

The 75% duty cycle is just the designer's convention by which the ECM tells the FPM just to stop pumping completely.

Not sure how a bad fuel pressure sensor is detected. Could be the ECM confirms that the sensor output varies over time and/or in response to its demands for more/less fuel, or is outside of expected range.

A scope will show you in real time what the inter-module dialogue is, and help to confirm or refute my thinking - which is the pump at this stage, QED.
Still curious about the absence of any related DTCs.

We've come a long way from a few spanners and a test lamp in the toolkit, er - I think...

 

Oh I get it,,, and goodness knows that is what I am doing as well... Trying to think it thru, making assumptions and "educated" (uneducated on my part might be a better way of saying it) guesses as to what is going on...

To me, I think the vital piece of information lies in getting a figure and understanding what would be the correct steady voltage on that one (non-fluctuating) white wire. For some reason I don't think the FPM is receiving or achieving enough voltage to engage its function... I am guessing here but,,, from the ECM, over a certain value of say 5 or 6volts, that the FPM can then determine a rate at which to pulse the pump. I think that 3.xx Volts is toooo low to function. Problem is I don't know why the ECM would be sending toooo low a voltage.

At the same time (I think) the fuel pressure sender digital potentiometer gizmo seems to work by sending a signal or the ECM on a scale of somewhere around 0 to 5 volts... Why not the FPM?

As you said, I need a working car. And I have one. But I have to line up with its owner. Busy schedules here in NYC.

Or,,, the pump could not be correct.

Or,,,, mice have been at work and one of the white wires has been nibbled down to a single copper strand.

Or,,, ALL of that could be happening at the same time, lol

What a THING.

I will be back out to tinkering on Saturday morning.

Please know your time, attention and efforts are greatly appreciated and have definitely deepened my understanding of this system and what I'm looking at... Helps a LOT. Can't overstate that fact. Thank you...

I'll be reporting back with any new info ASAP.

 

Key to this is that the signal on BT18-3 from the ECM should NOT be a steady voltage. It is switched between 0V and logic high (we'll say that's +5V for now) 150 times a second. The ECM also varies the amount of time that the voltage is at logic high during that 1/150th of a second (the width of the pulse). Either your DMM is averaging that out to give you (in this case) ~3.7 volts, or something is broken. Difficult to 'see' what's happening here with a DMM.

not sure what you mean here - is it why doesn't the fuel pressure sensor notify the FPM directly, or why doesn't the FPM send a signal to the ECM?