[SOLVED] Can't Hold Fuel Pressure

3/29/2023 Update: CorkSport internals were seizing. Installed AutoTech internals and the car runs normally again.

4/21/2023 Update: CorkSport confirmed the internals were bad after testing them at their lab, however were unable to warranty them since I had a shop purchase and install these on my vehicle.

Hey everyone, back again with another fuel pressure issue.

For the last month or so, my car has been having an issue where it's unable to maintain desired fuel pressure levels under WOT, dropping linearly into the 1400s at times.

Fuel system wise, I've got:

• CorkSport internals
• Corksport injector seals
• OEM PRV
• OEM HPFP
• OEM injectors (serviced)

These parts have all been installed new/serviced within the past few years.

Troubleshooting wise, I swapped (with no luck):

• 3 different PRVs
• Spill valve
• Ignition coil harness
• Verified issue happens both on 91 and E30 tunes

I also took apart the 1yr old pump and "cleaned" the nothing inside, internals looked surprisingly scored (to me), but CorkSport didn't seem to think it was a problem.



I've noticed my first pull after warming up seems to be the least problematic, with further pulls bringing the fuel pressure down quicker. When turning the car off after being driven, the fuel pressure drops quickly, hence me trying many different PRVs.

I've heard the sensors don't fail in the way that I'm experiencing, so I'm hoping it's not a leaky injector or the in-tank pump. Pump pressure is normal with the KICO test, showing 60ish psi. I have not tried a fuel volume test.

I've attached two datalogs showing the fuel pressure drop along with a video of the PRV test.

If the PRV test is truly that simple, and I assume the PRV is good, then am I left with the check ball or leaky injector?

Edit: I forgot to mention my fuel system squeals like a bitch on cold starts and only goes away once the car is warmed up. It's incredibly annoying.

(these cars are nightmares)

 

Did you clean the spill valve

 

Yeah it looked really good as it's only a year old. No black gunk.

 

I took apart the HPFP and rebuilt it once more, this time using Naptha along with Redline fuel system cleaner and taking extra care to ensure things were clean and lubricated. Unfortunately, no improvement.

Fuel pressure still starts to drop during pulls, not necessarily WOT ones. I saw as much as 257psi drop in 20 milliseconds.

Short of the sensor itself, I'm unsure what else to do.

Here's some charts of fuel pressure from two different tests.

Key On Engine Off:


This seems pretty normal based off CorkSport's troubleshooting, with the EVAP system letting the pressure drop slowly.

Key On Engine Off after idling to 219° coolant temp:


Not sure why the pressure drops so fast at first, but it never did buildup anywhere close to 1850. I only let the car idle to 219, so perhaps the pressure would be higher if I had driven around spiritedly?

Datalogs attached as well.

 

Pertaining to the PRV test, for the hell of it, here's my data along with 3 other sources I found on YouTube, two being around 30 seconds and the other being a ~10 min log. My car certainly doesn't build pressure as fast as these sources.

Grey: bigshoms3 on YouTube
Gold: b.sean on YouTube
Thin Blue: Unknown Username on YouTube
Orange: My car after warming up to 219° coolant temp
Thick Blue: My car cold, key on engine off

 

Seems like you've run through most of the common items, may be worth exploring the ITFP, maybe line blockage somewhere?

 

I have been considering the ITFP as well, but after tons of reading and words from @Enki on a few, it really seems like seeing ~60psi KOEO is all that's needed, as the HPFP does most the work. Everything below the 60psi seems to lead to the relay/resistor,EVAP system, or ITFP.

Correct me if I'm wrong though. Either way, I'm not in a position to service that on my own currently.

The way I understand it, there's 4 points of pressure loss for the high pressure fuel in the rail. This is assuming the data is good, else it could be electrical related.

1. HPFP feed line / check ball
   • Looks fine, but I haven't tried swapping. Pump is a year old.
2. Pressure relief valve
   • Have tried 3 additional PRVs, 1 being brand new
3. Pressure sensor
   • Have not changed it, considering as next step
4. Injectors
   • Serviced within past few years, will get an idle log to check out fuel trims for leaky injector symptoms

 

Regarding the blockage, is this something I could take compressed air to the feed line into the HPFP or anything?

 

Did you blip the throttle just as you shut it down?

 

I didn't, but I did plan to redo this test after a harder drive. I'll try blipping the throttle instead, seems much easier.

 

Yes blip then shut down, should give you a higher starting pressure. Also just because the ECTs are 220 doesn't mean it won't dump heat like crazy if you don't get the entire engine bay hot. Doing a cold start up to 200 deg idling isn't going to cut the mustard, for example; the injectors, rail, hardline and HPFP need to be up to temp as well. If your oil is hot the HPFP will be too, but that takes a lot longer to heat up than the coolant does.

 

Alright, did another fuel pressure test after a 30 minute drive and with a blip of the throttle prior to shutting the car off. The results are in green below. It's the exact same pattern as prior, just a higher starting pressure, of course. Still unable to hit 1850psi.

Grey: bigshoms3 on YouTube
Gold: b.sean on YouTube
Light Blue: Unknown Username on YouTube
Green: My car after 30 minute drive, throttle blip, car shut off
Orange: My car after warming up to 219° coolant temp
Dark Blue: My car cold, key on engine off



Additionally, I grabbed an idle log after the car had been warmed up and noticed a significant negative STFT. This has me thinking that I potentially have a leaking injector, as it would seem the engine is seeing additional fuel. I am open to feedback, though. Log attached.

 

At 1600 psi, your injector is not leaking. If it were, even a little, it would *tank* even with a tiny leak. Incompressible fluids (hydraulics) change pressure very quickly with small changes in volume, like with temperature; compressible fluids (pneumatics), however, do take a long time to drop pressure with tiny leaks.

 

Any ideas on what my issue might be then? Sensor is the only thing left of the four mechanical parts, short of an electrical issue.

 

Is your ambient air temperature sub 60 deg F?

 

Didn't get the ambient at the time, but today is 57-65 degrees. Datalog shows intake temps of 129 and BAT of 113, albeit idling.

 

Another update with more data.

Red: My car after spirited driving and pulling over to record fuel pressure with car off
Purple: My car after spirited driving and going home to record fuel pressure with car off
Grey: bigshoms3 on YouTube
Gold: b.sean on YouTube
Light Blue: Unknown Username on YouTube
Green: My car after 30 minute drive, throttle blip, car shut off
Orange: My car after warming up to 219° coolant temp
Dark Blue: My car cold, key on engine off



The red one is by far the most interesting to me as it almost looks like the PRV is doing something, albeit a bit shy of the 1850 mark.

Additionally, I've attached 3 WOT 4th gear pulls that vary in pressure drops.

 

It's not ideal but it's fine.

 

Are fluctuations of fuel pressure dropping such as below normal?

Down 126psi over 179 milliseconds and then back up 129psi over 182 milliseconds seems odd.


Trying to determine if I should replace the sensor or troubleshoot something else next.

 

That could be the PRV opening up just enough to let a *drop* of fuel out.

Edit: The camshaft runs at 1/2 crank RPM, has THREE lobes to drive the HPFP, and has VVT so who knows when it's going to push the piston forward or at what rate (faster if advancing, slower if retarding), etc.

Also assuming you were cranking it here, correct? If so, then the spill valve and check valve come into play; yet more variables. Get the car hot, in a controlled environment (where ambient is say 70-80F and then see how it acts. Other than that, even 1750 PSI you see here is actually fine.