Volvo Penta V6-240 Fuel High Pressure Dropping to Zero

[harringtondav]

I'll run some more .rec, at temp and different RPMs and post them tomorrow, if I can.

To attach a .rec file, open the Diacom 'record' file and right click on the file or files you want to attach. Select "Compress to...", and select 7z file.
It will take a few seconds before the 7z file shows in the folder. These can be attached.

 

[Biz603]

Try to get this back into one thread with this message.
Ok. Many good points above. I'll try to answer all of them. If I miss something let me know. I'll do the tests listed below this weekend.
1. I'm going to pick up an Ultraviolet fuel additive and put it in the gas supply. If the relief valve is dumping into the crankcase (or bypassing the piston) I will see it on the dip stick with a UV light. Will let you know.
2. I'll get a scope on the signal to the HPFP and try to record what I see.
3. Regarding the battery. Interesting points. Open to trying a different battery if it helps. I purchase the boat from a Cobalt dealer in 2017 with the following battery in it. DieHard part number 24M-AGM.
This battery lasted 6 years then I purchased the same one 2.5 years ago to replace it.
Just purchase a brand new one tonight, put it in and no noticeable change.
4. The only rubber fuel supply line runs from the tank to the fuel filter. To eliminate it and possible fuel tank screen clogged or fuel line check valve issues, I took this line out and ran a new line from the filter to a separate 5 gallon gas can. No change. Same issue.
5. Regarding the HPFP - Started with a new Volvo HPFP. Same issue. Then purchased 2 new GM pumps and tried them, replacing the fuel pipes each time. Same issue.
6. For reference, this is a model V6-240-C-J engine. Its governed to a speed (or interpolated RPM) request and tries to hold that requested RPM.
7. Just found out why the .rec didn't stick. This site didn't like them. No big. I ran new once tonight, on the new battery, will add them to this reply as .txt files. You'll have to download them and change the extension to .rec. One is a 18 minute run with several minutes of warm up at idle before increasing the RPM by 500 RPMs and holding for 1 minute before increasing another 500 RPMs, until failure. Another is just a quick throttle up. And the last is a Misfire .rec. Needless to say, all still failed at the end. I held the requested RPM high for seconds after failure to see what it would do. It didn't sound good. Engine was struggling to stay alive. Didn't feel good about holding this too long. You can see this in the first two files. I haven't reviewed them yet. They are hot off the engine.

 

[Biz603]

I've never used OBD equipment on a marine engine but I diagnose vehicle engine problems everyday. On a car P0174 would point to lean mixture on bank2, on this boat engine the 0174 code also seems to point to lean mixture on bank2. The way it detects mixture is using an O2 sensor, the boat engine seems to have O2 sensors because the OBD information includes readings for mixture.

If the O2 sensor on bank2 were bad that could maybe account for the engine going into limp mode? In both 'OBD snapshot' tables posted by @Biz603 in post #24 closed loop fuelling is disabled and the 'readings' from the O2 sensors seem unlikely to be correct because they point to an absolutely correct air fuel ratio. In practice the air fuel ratio usually flutters around 1 (where 1 is correct, above 1 usually lean, below 1 usually rich) so we might expect to see numbers like 0.997, 1.002 etc. At high engine load the engine normally gets a rich mixture so the reading is usually below 1, on a car it might go to 0.9 (you could read that as 10% rich). A lot of OBD systems on cars will create fake numbers for components they have deemed to be not working properly, they might give an A/F reading of 1 pointing to absolutely correct mixture for a failed O2 sensor.

Summing up, I'm suggesting the O2 / lambda sensor on cylinder bank 2 could be at fault...

See posts above. I replaced both O2 sensors and all the fuel injectors and the bank 2 trim fault went away. It did seem like it was in limp mode, but it turns out its just this problem we're trying to troubleshoot now. Basically the engine dies at >4300 RPM because the HPFP pressure is dropping from around 2050 psi down to below 30 psi even though the commanded RPM's stay high.

 

[muc]

I don't know how to convert these into .rec filed. Can anyone help me?

 

[harringtondav]

7. Just found out why the .rec didn't stick. This site didn't like them. No big. I ran new once tonight, on the new battery, will add them to this reply as .txt files. You'll have to download them and change the extension to .rec.

You can attach the .rec files in a 7z zip file.
Open the Diacom 'record' folder and right click on the file or files you want to attach. Select "Compress to...", and select 7z file.
It will take a few seconds before the 7z file shows in the folder. These can be attached.

 

[harringtondav]

5. Regarding the HPFP - Started with a new Volvo HPFP. Same issue. Then purchased 2 new GM pumps and tried them, replacing the fuel pipes each time. Same issue.

Interesting. I believe the VP 4.3 DI engine is a GM 4.3 liter v6 ecotec3 lv3 for the most part sans exhaust manifolds, cooling, ECM etc.
I wondered if you got the Delphi/Delco p/n from the VP HPFP. I was poking around an found the HPFP at a considerably lower cost than VP.
I buy only boat OEM parts when they're proprietary, like drive parts. But when they're stamped with other OEM p/n's like this I feel safe purchasing them and saving a lot on the cost.

 

[Biz603]

I don't know how to convert these into .rec filed. Can anyone help me?

You can also take these files and rename the ending from .txt to .rec

 

[Biz603]

Inte resting. I believe the VP 4.3 DI engine is a GM 4.3 liter v6 ecotec3 lv3 for the most part sans exhaust manifolds, cooling, ECM etc.
I wondered if you got the Delphi/Delco p/n from the VP HPFP. I was poking around an found the HPFP at a considerably lower cost than VP.
I buy only boat OEM parts when they're proprietary, like drive parts. But when they're stamped with other OEM p/n's like this I feel safe purchasing them and saving a lot on the cost.

The GM part number for the pump is not on the VP pump, but it's not on the GM pump either. After realizing the VP engine was the GM 4.3 V6, that you mentioned above, used in the Chevy Silverado 1500 (2014-2019), I ordered the GM HPFP (p/n: 12744840). It looked identical, QR codes and everything. Saved me a ton of $$'s. In fact, like you say, many of the parts on the engine are most likely GM parts. Injectors, fuel pipes / rails, spark plug coils, air intake manifold, many sensors, etc. Not the low pressure pumps, exhaust manifolds, cooling systems, or ECM.

 

[Biz603]

You can also take these files and rename the ending from .txt to .rec

Here is a zip file with the .rec files

 

[muc]

Here is a zip file with the .rec files

Based on what I see in these files.
I can see two inputs and one output go crazy.
The inputs are fuel pressure and MAP.
The output is injector pulse width.
The fuel pressure drops at the same time manifold pressure increases. Both of them happen at the same time.
The injector pulse width seems to be a reaction that happens a little later.

Because I don't fully understand exactly how this system works. I have to look for something I see bad. What I see is the manifold pressure increase from a normal number to almost barometric pressure in less than a second. MAP rises way too fast for it to be anything other then a sudden massive vacuum leak or a bad MAP sensor/MAP wiring problem.

I would replace the MAP sensor. It doesn't seem to be rising smoothly along with RPMs like I'm used to seeing and then It suddenly jumps.

Is it possible that these crazy readings from the fuel pressure and MAP could be a EMP? When you did the first repairs, is it possible that some of the wires got moved closer to the coils and spark plugs?

 

[Biz603]

Based on what I see in these files.
I can see two inputs and one output go crazy.
The inputs are fuel pressure and MAP.
The output is injector pulse width.
The fuel pressure drops at the same time manifold pressure increases. Both of them happen at the same time.
The injector pulse width seems to be a reaction that happens a little later.

Because I don't fully understand exactly how this system works. I have to look for something I see bad. What I see is the manifold pressure increase from a normal number to almost barometric pressure in less than a second. MAP rises way too fast for it to be anything other then a sudden massive vacuum leak or a bad MAP sensor/MAP wiring problem.

I would replace the MAP sensor. It doesn't seem to be rising smoothly along with RPMs like I'm used to seeing and then It suddenly jumps.

Is it possible that these crazy readings from the fuel pressure and MAP could be a EMP? When you did the first repairs, is it possible that some of the wires got moved closer to the coils and spark plugs?

Great read on the data and feedback.
I'll get a new MAP.
Not sure the wiring moved as it's mostly in a large harness running up both sides of the air intake manifold. But I will check.
I'll circle back once both are completed. It might be a couple days.
Again, thank you for looking into this deep into this with me.

 

[Lpgc]

See posts above. I replaced both O2 sensors and all the fuel injectors and the bank 2 trim fault went away. It did seem like it was in limp mode, but it turns out its just this problem we're trying to troubleshoot now. Basically the engine dies at >4300 RPM because the HPFP pressure is dropping from around 2050 psi down to below 30 psi even though the commanded RPM's stay high.

Sorry I missed that.

 

[Lpgc]

Based on what I see in these files.
I can see two inputs and one output go crazy.
The inputs are fuel pressure and MAP.
The output is injector pulse width.
The fuel pressure drops at the same time manifold pressure increases. Both of them happen at the same time.
The injector pulse width seems to be a reaction that happens a little later.

Because I don't fully understand exactly how this system works. I have to look for something I see bad. What I see is the manifold pressure increase from a normal number to almost barometric pressure in less than a second. MAP rises way too fast for it to be anything other then a sudden massive vacuum leak or a bad MAP sensor/MAP wiring problem.

I would replace the MAP sensor. It doesn't seem to be rising smoothly along with RPMs like I'm used to seeing and then It suddenly jumps.

Is it possible that these crazy readings from the fuel pressure and MAP could be a EMP? When you did the first repairs, is it possible that some of the wires got moved closer to the coils and spark plugs?

I don't have Diacom (so can't read the .rec files) and I'm not familiar with this particular engine. I'm not even fully versed with exactly how some direct injection engine fuel systems work with their multi-pressure stage fuel systems.

I do know that on some direct injection engines there is command fuel pressure and the low side pressure affects the high side pressure. Command pressure might be related to engine load. The fuelling system needs to inject a certain amount of fuel to keep mixture correct. The amount of fuel an injector will flow is affected / controlled by both fuel pressure and pulse duration. Increase fuel pressure but keep pulse duration the same and the amount of fuel the injector will flow will increase. Increase pulse duration but keep fuel pressure the same and injector flow will increase. There might be command fuel pressure but actual pressure might not be quite at command pressure so the ECU has to compensate pulse duration for fuel pressure reading - If actual fuel pressure is below command fuel pressure the ECU might compensate by increasing injector pulse duration.

If the engine has an accelerator pedal / gas pedal / boat throttle lever that acts like a kind of cruise control, by which I mean the electronics try to achieve an engine / prop speed that is directly linked to the throttle lever position (as opposed to a conventional throttle lever), then if RPMs suddenly drop below intended RPM the electronics might very quickly open the throttle (try to bring rpms up to command rpm) causing an almost equally quick rise in MAP. Map can change very quickly, almost as quickly as the electronic throttle can open, if I connect to my car OBD system with a fast sample speed scan tool and 'blip' the throttle the map sensor shows an sudden increase in map from idle pressure (around 0.3bar or 0.7bar vacuum) to almost atmospheric pressure (1bar) in a few milliseconds.

It wouldn't be surprising if map were to suddenly increase if the engine suddenly started making less power than expected due to some other problem, such as fuel pressure suddenly falling off and the engine running lean because the injectors couldn't possibly flow enough fuel to the engine despite increasing pulse duration in response to the low fuel pressure reading, this could be in response to the ECU opening the electronic throttle to try to make more power to increase rpm. In this case we should expect to see a sudden drop off in fuel pressure reading, then almost immediately a sudden rise in injector pulse duration, then almost immediately a lean mixture reading and experience a sudden drop off in engine power but a scan tool / diagnostics might not show events happening in the correct order, the diagnostics system has a sample speed for each of the values. In this case the issue wouldn't be the map sensor or petrol injector pulse, they would just be responding as expected. The issue would be the drop off in fuel pressure and the question in that case would be why is the fuel pressure dropping off. Maybe fuel pressure dropped off due to a fault with a low or high pressure pump or maybe the ECU intentionally caused fuel pressure to drop off in response to some other detected fault, like deciding to go into some sort of limp mode.

 

[Biz603]

Great read on the data and feedback.
I'll get a new MAP.
Not sure the wiring moved as it's mostly in a large harness running up both sides of the air intake manifold. But I will check.
I'll circle back once both are completed. It might be a couple days.
Again, thank you for looking into this deep into this with me.

Checked the wiring and it was a little close to the alternator so I moved it a bit. Not even close to the coils.
Put in a new MAP sensor, still the same issue.

 

[Biz603]

I don't have Diacom (so can't read the .rec files) and I'm not familiar with this particular engine. I'm not even fully versed with exactly how some direct injection engine fuel systems work with their multi-pressure stage fuel systems.

I do know that on some direct injection engines there is command fuel pressure and the low side pressure affects the high side pressure. Command pressure might be related to engine load. The fuelling system needs to inject a certain amount of fuel to keep mixture correct. The amount of fuel an injector will flow is affected / controlled by both fuel pressure and pulse duration. Increase fuel pressure but keep pulse duration the same and the amount of fuel the injector will flow will increase. Increase pulse duration but keep fuel pressure the same and injector flow will increase. There might be command fuel pressure but actual pressure might not be quite at command pressure so the ECU has to compensate pulse duration for fuel pressure reading - If actual fuel pressure is below command fuel pressure the ECU might compensate by increasing injector pulse duration.

If the engine has an accelerator pedal / gas pedal / boat throttle lever that acts like a kind of cruise control, by which I mean the electronics try to achieve an engine / prop speed that is directly linked to the throttle lever position (as opposed to a conventional throttle lever), then if RPMs suddenly drop below intended RPM the electronics might very quickly open the throttle (try to bring rpms up to command rpm) causing an almost equally quick rise in MAP. Map can change very quickly, almost as quickly as the electronic throttle can open, if I connect to my car OBD system with a fast sample speed scan tool and 'blip' the throttle the map sensor shows an sudden increase in map from idle pressure (around 0.3bar or 0.7bar vacuum) to almost atmospheric pressure (1bar) in a few milliseconds.

It wouldn't be surprising if map were to suddenly increase if the engine suddenly started making less power than expected due to some other problem, such as fuel pressure suddenly falling off and the engine running lean because the injectors couldn't possibly flow enough fuel to the engine despite increasing pulse duration in response to the low fuel pressure reading, this could be in response to the ECU opening the electronic throttle to try to make more power to increase rpm. In this case we should expect to see a sudden drop off in fuel pressure reading, then almost immediately a sudden rise in injector pulse duration, then almost immediately a lean mixture reading and experience a sudden drop off in engine power but a scan tool / diagnostics might not show events happening in the correct order, the diagnostics system has a sample speed for each of the values. In this case the issue wouldn't be the map sensor or petrol injector pulse, they would just be responding as expected. The issue would be the drop off in fuel pressure and the question in that case would be why is the fuel pressure dropping off. Maybe fuel pressure dropped off due to a fault with a low or high pressure pump or maybe the ECU intentionally caused fuel pressure to drop off in response to some other detected fault, like deciding to go into some sort of limp mode.

1st paragraph - Agreed. See above. Fuel pressure to RPM Chart. This engine is running almost exactly to that chart. Injectors are all new.
Fuel pressure in this system is maintained with the ECM. The ECM then controls the injector with a PWM signal to deliver the right amount of fuel based on that pressure. My problem, the fuel pressure drops unexpectedly above 4300 ish RPM. Even with a new low pressure, verified working, fuel pump and three now HPFP's. We are currently reviewing the injector PWM signal, but it looks good, even during the initail failure.
2nd paragraph - See above. This is a fly-by-wire system where the commanded RPM gets sent to the ECM. If I increase the RPM very slowly over 1 minute or goose it, it still fails. The difference is when I goose it, it gets to 5800 RPM (max rpm) but doesn't stay there long, like a fuel starvation issue. Again all new pumps and lines.
3rd paragraph - Agreed. See above about these items and limp mode. See 1st paragraph response above. New MAP sensor today and no change. It was going off the charts, but we didn't know if it was because the fuel pressure dropping first or a problem with the MAP first. Now we know it's not the MAP. Now were on to your last point. Not limp mode, but is the ECM injector output signal dropping first, somehow in response to another sensor issue or input signal. Overall the signals look like the system is requesting more fuel and air and it sounds like the butterfly throttle valve is opening up during failure. Back to analyzing the .rec recorded data.

Open to ideas.

Thanks!

 

[muc]

Checked the wiring and it was a little close to the alternator so I moved it a bit. Not even close to the coils.
Put in a new MAP sensor, still the same issue.

I can only see two inputs to the ECM that act weird. MAP and fuel pressure.
It was a toss up as which I would recommend first. My call was the MAP, because of the "gold rule" when a tech is guessing --- Pick the cheapest and easiest first.
I think it's now time to replace the fuel pressure sensor.

After @Lpgc comments, I went back and looked at all the ECM outputs. The ones that start changing all have a lag behind the MAP and FP. I have good confidence in Diacom's ability to output data in real time.
While this still could be a lack of or aerated fuel (is there a flexible fuel line between the low pressure pumps and the HPFP?) you are measuring low pressure right at the pump, so if there is a restriction. It wouldn't show up on the gauge.
But fuel pressure is working just like it should until RPMs get to high. It's able to reach full pressure and maintain it for a long time, it's just once RPMs get to high.

It would be nice if Diacom would monitor PWM to the HPFP. Maybe give them a call to ask? Or have you looked at all the possible monitors to see if it's included in any of the data sets?

 

[muc]

I don't have Diacom (so can't read the .rec files) and I'm not familiar with this particular engine. I'm not even fully versed with exactly how some direct injection engine fuel systems work with their multi-pressure stage fuel systems.

I do know that on some direct injection engines there is command fuel pressure and the low side pressure affects the high side pressure. Command pressure might be related to engine load. The fuelling system needs to inject a certain amount of fuel to keep mixture correct. The amount of fuel an injector will flow is affected / controlled by both fuel pressure and pulse duration. Increase fuel pressure but keep pulse duration the same and the amount of fuel the injector will flow will increase. Increase pulse duration but keep fuel pressure the same and injector flow will increase. There might be command fuel pressure but actual pressure might not be quite at command pressure so the ECU has to compensate pulse duration for fuel pressure reading - If actual fuel pressure is below command fuel pressure the ECU might compensate by increasing injector pulse duration.

If the engine has an accelerator pedal / gas pedal / boat throttle lever that acts like a kind of cruise control, by which I mean the electronics try to achieve an engine / prop speed that is directly linked to the throttle lever position (as opposed to a conventional throttle lever), then if RPMs suddenly drop below intended RPM the electronics might very quickly open the throttle (try to bring rpms up to command rpm) causing an almost equally quick rise in MAP. Map can change very quickly, almost as quickly as the electronic throttle can open, if I connect to my car OBD system with a fast sample speed scan tool and 'blip' the throttle the map sensor shows an sudden increase in map from idle pressure (around 0.3bar or 0.7bar vacuum) to almost atmospheric pressure (1bar) in a few milliseconds.

It wouldn't be surprising if map were to suddenly increase if the engine suddenly started making less power than expected due to some other problem, such as fuel pressure suddenly falling off and the engine running lean because the injectors couldn't possibly flow enough fuel to the engine despite increasing pulse duration in response to the low fuel pressure reading, this could be in response to the ECU opening the electronic throttle to try to make more power to increase rpm. In this case we should expect to see a sudden drop off in fuel pressure reading, then almost immediately a sudden rise in injector pulse duration, then almost immediately a lean mixture reading and experience a sudden drop off in engine power but a scan tool / diagnostics might not show events happening in the correct order, the diagnostics system has a sample speed for each of the values. In this case the issue wouldn't be the map sensor or petrol injector pulse, they would just be responding as expected. The issue would be the drop off in fuel pressure and the question in that case would be why is the fuel pressure dropping off. Maybe fuel pressure dropped off due to a fault with a low or high pressure pump or maybe the ECU intentionally caused fuel pressure to drop off in response to some other detected fault, like deciding to go into some sort of limp mode.

@Lpgc thank you for taking an interest in this issue, more eyes looking at a problem always help. You can look at the files if you want. Just go to rinda.com and click on the updates tab. Click the button next to Diacom and then click the download button. This will let you install Diacom software and you can view any .rec files. This is all free, if you want the extras like the service manuals and the harnesses to scan an engine, that costs extra.
Rinda taught me this when I was at a Very remote lake and the customer attempted to help me, but instead he knocked my laptop into the water. Customer used his satellite phone and downloaded Diacom to his Laptop and used my cables to finish the diagnosis and get him running again.

I would be very interested in your opinion.

 

[Biz603]

I can only see two inputs to the ECM that act weird. MAP and fuel pressure.
It was a toss up as which I would recommend first. My call was the MAP, because of the "gold rule" when a tech is guessing --- Pick the cheapest and easiest first.
I think it's now time to replace the fuel pressure sensor.

After @Lpgc comments, I went back and looked at all the ECM outputs. The ones that start changing all have a lag behind the MAP and FP. I have good confidence in Diacom's ability to output data in real time.
While this still could be a lack of or aerated fuel (is there a flexible fuel line between the low pressure pumps and the HPFP?) you are measuring low pressure right at the pump, so if there is a restriction. It wouldn't show up on the gauge.
But fuel pressure is working just like it should until RPMs get to high. It's able to reach full pressure and maintain it for a long time, it's just once RPMs get to high.

It would be nice if Diacom would monitor PWM to the HPFP. Maybe give them a call to ask? Or have you looked at all the possible monitors to see if it's included in any of the data sets?

I found the PWM in the data set in the .rec files I sent earlier. See it below at the bottom of the chart. It's measured in milliseconds. Starts at .9ms then starts to climb when the rpm drops. Kinda tells me the ECM is working if nothing else.
There is a braided flexible line from the low pressure (75psi) pump to the HPFP. It's about 14" long and has connectors on both ends with metal clips holding the connection on. It's the kind you have to press inside the ID of the connector to release it. Not sure how I'd tap into it on the HPFP end. Open to try.
I'll install a new high pressure sensor soon and let you know.

In the meantime, I thought of something to try. I believe the Diacom software allows me to shut down injectors. I will try shutting down one injector (if possible) and see what happens. Because of the reduced fuel requirements it should be able to get to higher RPM's before the pressure drops, if it's a fuel starvation issue. Or, on the other hand, if it's the HPFP relief valve opening up, it should open sooner or the same low RPM I've been seeing.

Thoughts?

Will circle back later tomorrow.

Here is the pic:

 

[muc]

I don't think shutting down an injector would give much useful info. I wouldn't know what to look for. But if you want to try I don't think it would cause problems and might be useful.
It's very unlikely the flexible fuel line is collapsing, but possible and would explain your issue. I don't know, but the auto people might might have a T that you could install by the HPFP. It would rule it out.

The PWM that I would like to see is the Fuel control solenoid.
Your injector pulse width operates a little different (but pretty close) then the other Volvo DI engines files that I have looked at.
This is different than all the other Marine engines that I worked on. All the other one's use pulse width heavily and you see big swings.
The DI engines have fuel pressure modulation so they change very little. If you watch the injector pulse width (IPW) you will see that it stays very constant until the event that your experiencing. This is slightly different from the other 4.3 DI files I have looked at. The .rec files I have from the correctly running engines that are 4.3 DI and they show the average IPW a little higher then yours until high fuel pressure reach's about 1800psi. Then IPW slowly rises to about 5ms. Your IPW stays very constant until the event and then it goes Waco then it goes up to almost cold starting IPW, this is different then other engines.

This brings up something. I haven't connected Diacom to a DI engine, so I'm not 100% sure what you see when you connect. But from helping others by looking at the .rec files and talking with them. There are sometimes many "monitors" that you can choose view/record. This is what some engines see.



I'm thinking that some (maybe all) of these might be useful in tracking down your problem. If you decide to do this as .rec files. It's important that they all be done one after the other with the engine already fully warmed up and try to make each .rec the same as possible. I suggest starting each one following this format. This is because when I look at them, I won't have to take the time to determine operating conditions as I switch between the different .rec files.

Stopped at idle, shift into gear, accelerate slowly (like 2 minutes) to reach full throttle (use the trim as you normally do), hold it there for 5 seconds and then slow down to neutral as you normally would. Repeat the next monitor.

It's possible the HPFP PWM might be covered in one of these monitors. From trying to help others I have found that some of them will have all these monitors and some will just have a few. One person had one monitor "grayed" out and couldn't access it. They called Diacom and were told that it was because Volvo hadn't enabled that monitor in the ECM and to talk to Volvo to get it reprogramed. They weren't a Volvo dealer and they found the problem without it, so I don't think they followed up.

 

[harringtondav]

I found the PWM in the data set in the .rec files I sent earlier. See it below at the bottom of the chart. It's measured in milliseconds. Starts at .9ms then starts to climb when the rpm drops. Kinda tells me the ECM is working if nothing else.
There is a braided flexible line from the low pressure (75psi) pump to the HPFP. It's about 14" long and has connectors on both ends with metal clips holding the connection on. It's the kind you have to press inside the ID of the connector to release it. Not sure how I'd tap into it on the HPFP end. Open to try.
I'll install a new high pressure sensor soon and let you know.

In the meantime, I thought of something to try. I believe the Diacom software allows me to shut down injectors. I will try shutting down one injector (if possible) and see what happens. Because of the reduced fuel requirements it should be able to get to higher RPM's before the pressure drops, if it's a fuel starvation issue. Or, on the other hand, if it's the HPFP relief valve opening up, it should open sooner or the same low RPM I've been seeing.

Thoughts?

Will circle back later tomorrow.

Here is the pic:View attachment 411946

I graphed the same data from your 18 min run.rec. What jumps out to me is your injector PW is flat under 1 ms through your acceleration to over 5000 rpm. It held close to 3ms until you started accelerating then dropped to 1ms.
Something seems wrong to me. Last summer when I was chasing poor performance and low power with @muc . my injector pulse width climbed with engine rpm to just under 5 ms at my under powered 4500 max rpm.
(Once I fixed the problem with new spark plugs the engine was back to full power and rpm. So unfortunately I didn't record healthy data.)
You are using significant power and fuel at 5000 rpm, but you injector pulse width doesn't reflect this. I'm thinking the readings from the
ECM are bad. I can't believe you're cruising at 5K rpm with idle speed injector pulse width.
It looks like your ECM was trying to compensate for the dropping rpms vs commanded power by opening up the injectors to 15ms to dump more fuel as the rail pressure was dropping.
Also interesting is as your engine was falling on its face the rail pressure briefly jumped back up to normal while the injector pulse width dropped back to a low level.
Observations, but no clue to the causes. This is a DIY forum and I hate to use the word "dealer", but is your engine still covered by the VP five year warranty? If it comes to it, ECMs are painfully expensive.