HPFP: pressure holds @ 1600 psi during cold start for about 20 seconds

Z

Zogy

Greenie N00B Member
Recently installed Autotech HPFP upgrade when doing a timing chain on my MS6 (no real mods yet, just running an Accessport Stage-1 pre-canned tune). Dissembled and cleaned everything that the instructions called for during the installation. Now upon cold start the pressure surges to 1600 psi and stays there for about 20 sections. During that time the vacuum drops to about 12 inHg. Which causes a misfire and CEL (for the misfire). After about 20-30 seconds the fuel pressure drops to ~400psi, the vacuum climbs to 22 inHg and the idle smooths out and all is well. After that it runs pretty well. The full pressure is ~400 psi @ idle, and then jumps up to ~1600 psi when you accelerate normally. While sitting, if you take the REVs up to 2k or 3k RPM the pressure surges up then drops back to ~400; if you keep the RPMs steady.
I assume that kind of fuel pressure behavior in normal?
For those fuel HPFP-savy any ideas what's goin on, and how to fix it?
(one final note, I can swear the last time I shut the car off and returned 10 minutes later, the fuel pressure was at 1600 psi before starting it. It quickly settled back to 400 psi at idle once fired up. Mention that as well in case that added to the troubleshooting).
Thanks in advance for any help/thoughts you might have.
Haven't posted in a while, which is generally a good thing. 8-)
 
Pressure generally increases for higher requested loads; seeing 1600 PSI after sitting is actually a troublehsooting test for the PRV, so that's fine too.

Not sure if that's normal for a cold start, but the misfiring sure isn't. Maybe post a log of a cold start so we can see what's going on.
 
Thanks for the thoughts. I was hoping someone could check their HPFP pressure; or know off-hand if 1600 psi is normal for the first 20 seconds after a cold-start; so I can check that off my list of weird symptoms. I agree the misfire is not normal. I kinda recall some misfire prior to the HPFP change, but it was pretty subtle, and much worse now. I'm gonna order some new coils so I can check that off my list just in case (long shot maybe. but they are likely tired after nearly 20 years). I'll try to run a log of the cold start, so you can see what is going on. I'm a newb, but I think I understand the logging process. What parameters would you suggest I log? I was thinking: RPM, Boost, Fuel Rail Pressure (I think the options are: actual, sensed, dev?-whatever that is) and/or maybe injector pulse width or duty cycle. Does that sound about right, or would you suggest a different set? Thanks in advance for any thoughs.
 
Okay here is a datalog from this mornings cold start up. I did notice a few things (but don't quite understand why it's doing it).
-The Fuel Rail Pressure is high and steady @ ~1600 psi, then drops like a rock
-The vacuum is crap during that time (9-10 inHg) and then rises up to 20 inHg as the fuel pressure drops
-The calculated load hangs around 0.45 then drops to 0.25 as the fuel rail pressure drops
-The ignition timing is pulling WAY back to around -10 to -12 deg during this time
Not sure what it's telling me, but if you have any ideas I would certainly welcome them.
Thanks in advance for your help.
(I tried to log only a few items, but luckily I failed and I think it log everything, although I don't see Fuel Rail Des Press, which I think is a commanded fuel pressure?)

In general, after it warms up it runs pretty good. Maybe an occasional misfire/stumble under load but nothing too serious.
I recently replaced the timing chain. I replaced the VVT Actuator like 60k miles ago (when I did the chain last time). I am pretty certain I got the timing correct. Spun the crank to the timing mark several times after torqueing everything down and the cam alignment tool inserted with only a minor jiggle.
 

Attachments

1. Higher loads need higher rail pressure; this is likely a tune thing, but at the pressures logged, shouldn't be an issue. You can see the correlation of load -> rail PSI in the log.

2. Vacuum is related to RPM; in order to have a higher RPM, the engine needs more air to flow; thus, vacuum will be higher at a lower idle RPM.

3. Load is a measure of airflow (sorta, not quite but it's a good enough explanation for now; just know that load is calculated off of what the MAF sees); thus, higher RPMs will see higher load (because more air is flowing). See point 1 above regarding rail pressure.

4. For timing, know that it's much more responsive for the ECU to fine-control idle using timing than it is the throttle plate.

For much of the above, you can also look at throttle position %, which is controlled by the ECU for high idle (instead of having, say, an idle air valve).

Another neat test you can do is do a cold start but skip the high idle; may take a couple tries but if you give the engine a little gas while it's starting, it should skip the high idle (just a momentary blip, if memory serves...can lift after it fires and it should settle into a hot idle, even though the engine is cold).

Another thing you can look at to compare the start with "normal operation" is to get on a nice flat stretch of road, turn on the cruise control, start a log (log everything again) and then hit the + for cruise three times rapidly, and stop the log when it settles. This should show both steady state and what happens when requested load (acceleration) is increased, and how all the things you point out above are ultimately tied together.

Honestly, the only thing that really stands out in this log is your +15% LTFT for the cold start, and when the O2 heats up its still a bit lean; this suggests either some kind of PCV leak or other post-MAF intake leak somewhere, but one only severe enough to effect a cold start.

Edit:
We'd need to see a WOT log of the stumble or whatever to make a determination on that, if it's even related.
 
That's a very detailed and informative post, thanks for taking the time. I'll recheck for an air leak. If I get ambitious maybe I'll try a smoke test. The new coils should be here on Monday so I'll install them and see if that helps (I am dubious). It seems to always come up with a misfire on cylinder #2, when it's not a general misfire code. I switched the coils around and it didn't shift position (which I why i am dubious about it being a coil issue). Could it be some type if wiring/connector issue with cylinder #2?
And the LTFT is hovering around +7% routinely, so as you mention maybe some type of intake leak?
I'll try your cold start by-pass thing and see what that provides as well.
Thanks again for the analysis and info. I'll report back when I have more pieces of the puzzle.
 
Got it. I'll try to investigate. BTW the cold start bypass trick worked like a charm. I didn't think to take log it to see what's happening, but I'll try to do that in a bit.
One thing I noticed is that the LTFT seems to be commanded to max out, and then within one data cycle drops from 15 to 6. (that happens at the instant the timing goes from -10 deg to +6 deg, also instantly) I always thought that the LTFT was a slow moving correction based on the STFT readings (trying to keep it to near zero). So, by that thinking it shouldn't be jumping around that quickly. This looks to be a commanded shift rather than a response to fuel mixture. I noticed that the LTFT hangs around 7 at idle and then when you take it off idle jumps near instantly to 15, and then back to 7 at idle. That happen all the time when you blip the throttle, or accelerate. Just shows I guess I don't really understand LTFT. And why would the timing and the LTFT jump instantly? It seems like the ECM is commanding a change since it is so sudden. Not sure what triggers the change, or why the ECM puts it in that state to start with (if that makes sense).
 
LTFT is based on STFT over time, but for a voltage range (configurable in tune). No real reason to adjust the ranges though. I normally configure my ranges to be split based on the kind of driving I'm doing (idle, low speed cruise, high speed cruise and up to WOT).

Also max for both trims is actually 20 or 25%, I forget which.

As a side note, I wouldn't worry about trims too much unless they exceed 8% ish each. STFT will swing wildly depending on weather and shit like whether or not your air filter is clean, too.
 
I'm a bit confused about your comment regarding voltage range. The LTFT jumps to a pretty constant, high value (like 14.85) when I take the RPM off of idle. Does than indicate a "voltage range"? I always thought LTFT was an idle-only thing (where the O2 sensors are in closed loop). Is the LTFT used at higher RPMs as well? (sorry for the potentially stupid question, obviously I need to do some learning)
 
No, there's settings in the tune where you can specify the MAF voltage ranges where trims are to take place. As I mentioned in my prior post, I edit these ranges so the long term fuel trims for my tunes are kept in blocks sized roughly around the airflow numbers (specifically MAF voltages) that line up with the specific engine activity; Idle, low cruise, high cruise, and pre-WOT. These are all separate LTFT ranges, and if you look at what the STFTs do when the LTFTs jump like that, you'll understand more.

Add them up in a spreadsheeet and see how the total compares to LTFT and STFT over time/jumps.
 
I did a smoke test on the intake system, and the only smoke came from the EGR valve at the location indicated. I nabbed the picture from member shivverring; the had the same question: is it normal? or does it need to be replaced?. I asked him in his post, which was a bit dated, whether he ever figured out the answer since nobody replied to his post. I figured you might happen to know. Since it part of my troubleshooting process for the rough idle and high LTFT readings. Thanks.

upload_2025-8-23_13-23-16.png
 
You must log in or register to reply here.
Back
Top