New Fuel Pressure Regulator not adjusting

[subisubisu]

I'm still wrestling with getting my just-rebuilt OBXT engine up and going smoothly.  I'm working with a respected remote tuner and the car is running and driving, but I am experiencing a bucking behavior in the 1800-2500 range, generally where I'm changing gears as I'm trying to keep it out of boost until I am sure that it's not going too lean, too fast for too long, and risking detonation.

So, in considering the air-fuel ratio, seeing an undesirable value could be the result of an air problem, a fuel problem, or both.  I've found and fixed a couple or three post-MAF vacuum leaks.  Now when I spray starter fluid anywhere along the intake path, the idle is unaffected, except for the intake itself.  When I spray it in there as a control, it bogs the idle down almost to stalling out before recovering.  So that gives me some confidence about the air side of things.

I had doubts about the fuel side, since I have a new AEM 50-1215 fuel pump and DW 740 injectors in what was otherwise the stock fuel system.  I have seen in RTFF that some folks have problems with resonance, "stumbling" and other ill effects, and it seems that the fuel delivery system in the stock configuration is pretty tuned, making me expect that if I changed elements of it, I could well have disturbed the tuning.  And the stock system doesn't have any sensor or way to read fuel pressure feeding the rails.

So, I replaced the stock FPR with a Radium adjustable unit.  Post-install checks showed no fuel leaks, and the car starts, idles, runs and drives with the new set-up, which includes a Radium-supplied fuel pressure gauge in the send line.  However, the adjustable part is not as I expected.  Loosening the jam nut and rotating the set screw multiple rotations doesn't affect the fuel pressure gauge at all.  Yes, I disconnected the vacuum signal.  I attached that to a vacuum gauge, and it sits steady at 17 inHg of mercury, which Google tells me is about 8.35 psi.  And the fuel pressure is sitting steady at about 18 psi.

I understand that the regulator is supposed to maintain a delta pressure.  Removing the vacuum hose means the delta is compared to atmospheric pressure, and for the LGT/OBXT is supposed to be 43.5 psi.  Generally speaking vacuum is supposed to be about 10 psi, so I should be seeing low 30's on the fuel pressure gauge at idle with the vacuum line connected; low 40's if it isn't.  If I am seeing 18 psi with vacuum of 8.35, then the delta pressure would seem to be in the 26-27 psi range, which seems too low.

I am scratching my head.  Lean condition can be too much air, yes, but also too little fuel, and the readings I'm seeing from my first experience with getting data from mine are not what I am led to expect.  Am I making an obvious mistake here?

[KZJonny]

Did the FPR not come with instructions? Maybe an email to Radium would be a good place to start.

May also help to change the vacuum source from the intake manifold by cylinder 2 to the same line that provides vacuum reference to the BPV. Others have had some luck with this.

Lots of folks on here have done this mod, myself included, only often using the STi FPR, just ‘coz it’s cheaper than and aftermarket one. But the Radium stuff sure is nice!

[subisubisu]

Thank you!  It's hard to know how much data to give.  In my case, I did already cap off the #4 reference and moved the tap for vacuum to the BPV by putting a second Company23 aluminum tee (I assume that's what is referred to as the Turkeylord tee in Covert Russian's thread you provided) there; I already replaced the blue tee with another one.  I could shorten up the hose, I guess, but since I am not seeing meaningful differences with the changes I am making, I am pretty sure I have not yet "stumbled" upon the root of the problem - heh heh.

I am continuing to research.  Just saw a video where the solution was putting a check valve in the return line, since the problem was diagnosed as the fuel pump check valve in the tank causing a leak.  I am also seeing my fuel pressure drop to zero, not hold, so presumably that is another significant symptom.  That video cited the injectors, regulator and pump check valve as possible culprits; all are new units in my case, which I would hope means they are good but in practice means that they are not proven.

[KZJonny]

I may not be understanding the system well enough, but my general impression had been that the FPR is the principal item which provides back-pressure to the fuel system. So, correct adjustment would be the thing (assuming a healthy fuel pump and no blockages on the supply side from the tank) that gives you the pressure you need at the injectors, etc...

In the case of a stock or STi FPR, they are not adjustable, to the lock nut and screw are irrelevant. You just have to assume that they are made to a spec and tested good before being shipped out. 

Since the pressure regulation is provided by a spring and diaphragm, which in the case of an adjustable FPR can be dialed in with the adjustment screw, have you tried to bottom out the adjuster (gently) then back it off by a turn or so? If the result is still something in the 26-7 psi range, then I suspect you're correct, and have a bigger problem than the Radium FPR. Without spendin a bunch of time looking into the particular unit you have, I would assume based on personal experience at work with all sorts of pressure regulating devices, that they all work within a range, and the one you have is suited to the application...

Not knowing exactly what you've tried, that would a logical place to start before getting to installing additional hardware like check valves in the system. As you say however, new and not working quite right does also mean untried and unknown.

[subisubisu]

Thanks again, I'll try bottoming out the diaphragm and see if it rebounds into adjustability.  I took on this project to learn and accomplish things, and I am getting a huge return on my investment in that respect - a seemingly endless parade of

What is it doing?

What is it supposed to be doing?

Why isn't it doing what it's supposed to be doing?

[KZJonny]

24 minutes ago, subisubisu said:

Thanks again, I'll try bottoming out the diaphragm and see if it rebounds into adjustability.  I took on this project to learn and accomplish things, and I am getting a huge return on my investment in that respect - a seemingly endless parade of

What is it doing?

What is it supposed to be doing?

Why isn't it doing what it's supposed to be doing?

It's just a thought. I am personally very much a 'rule one thing out at a time' kind of person. Probably comes from spending much of my education career and the like in biology/biochemisty. Not a field that works out well with the scattershot approach!

Now, that said, I will muddy the waters a little. Do the FPR thing first, but it might be worth also looking to see if you are getting sufficient voltage to your new pump. It is not highly likely to be the cause, I would think, but it could be another place to investigate. These cars are old, and wiring get tired and corroded. I'm currently chasing down a wheel sensor not sending any signal to the ABS controller. No reason it shouldn't be, and replacing it with a known good unit did not fix the problem.... bad wiring somewhere.

You may have already come across it, but there are a number of ways to improve voltage that actually gets to the pump by upgrading the wiring to it, and some people switch to the STi Fuel pump controller.... Not night and day, but it's larger wiring and has much less drop between the battery and the pump. Some people indicate they got a full extra volt by sizing up the wiring. Lots of threads out there on it, but this is the big one. I have all the stuff to do this, just waiting for end of driving season to install in my car. Not strictly needed for a variety of reasons, but I'm personally trying to do all I can to max out the performance of my stock side feeds on the cheap, since either bigger ones or a top-feed conversion are much much more difficult, and probably a little more than I really need on a modified VF40.... I haven't run out of fuelling quite yet, but.... there is always a lingering temptation to find a VF52 frame and have it souped up, hah!

 

[Max Capacity]

Have you reset the ECU since installing the changes to the fuel regulator system ?

I did the simple "T" fitting from the blow off valve hose and capped the port on #4 intake runner on both my cars, then reset the ECU. 

 

Pull the negative battery cable for 15 minutes. 

Put cable back on, turn Key to "On" for 10 seconds

Then start the engine, drive the car like you normally would until it's fully warmed up. 

 

After doing that, both cars did not have the stumble. 

 

 

[subisubisu]

Yes, the ECU has been reset.  I in fact had the battery disconnected while changing the fuel line and regulator - I didn't like the idea of electrical power and maybe spark so nearby, even though there is no electrical connection in the fuel lines I was working with.  And I did another reset using RomRaider later, after experiencing the bucking.  I wanted to see how the fuel trims behaved.  Bucking is my word for what I am experiencing; it's hard to know exactly what other folks are experiencing as they are describing it.  I see hesitation, bog, stumble, and I'm not the only one to say bucking.  It's a quickly cycling behavior that jerks the car strongly as I try to accelerate, simultaneous with high (lean) AFR readings on the gauge.

I am thinking it might be significant that it's not holding pressure with the key outside of the run position.  I know that designs vary among cars and that not every description of an automotive system necessarily applies to the one in front of me, but otherwise the descriptions I'm reading seem to match the OBXT system in components and layout, and I'm not seeing pressure maintained in the system where it seems like I should.  That sounds like a leak, and since I have no indication of a leak in the engine bay, I am suspecting something in the fuel tank.  However, that is based on very little "research" and no experience.

[KZJonny]

28 minutes ago, subisubisu said:

I'm not seeing pressure maintained in the system where it seems like I should

Correct. You should see pressure held. I want to say mine will hold most of 25-30 psi on the inline gauge overnight, or through the course of a work day.

[Max Capacity]

Yeah, sounds like the pump is not holding pressure. I see there is talk about two o-rings on the pump, do you have both of them ? 

[subisubisu]

57 minutes ago, Max Capacity said:

Yeah, sounds like the pump is not holding pressure. I see there is talk about two o-rings on the pump, do you have both of them ? 

I installed the pump a couple of years ago as this project has dragged on before finally getting the engine started this summer.  So, I don't recall details of that sort.  I will have to pull out the rear seat and dig the pump out to find out.  However, I have spare hose and fittings on hand.  A $7 check valve could be swapped into line to see what good it might do with less mess, effort and risk of breaking something else.  Of course, if the long-term fix is in or at the pump, so be it.  I just recall being very apprehensive about breaking that plastic bucket and its little parts.

[subisubisu]

OK, pump was out and is now back in, but...

I measured a strong 10v (!) at the pump.  So, not 12v, which is what I expected, but would need to look longer at the circuit diagram to verify.

I did indeed have two O-rings at the pump interface to the inner bucket.  I have single O-rings at each of the three other places too, according to this video:

I got everything buttoned back up and turned the key past accessory to the run position (but not all the way to crank) by reaching in the window so that I could quickly hawk the gauge in the supply line.  It went up to 40 or a bit beyond, and then started falling.  It was well below 20 when I turned the ignition off and pulled the key and gave up for the night.

10v is disappointing, but the pump appears to run on it.  Not holding pressure in the system definitely appears to be wrong.  I might try a check valve in line in the engine bay, since I have one coming tomorrow (but USPS isn't super-reliable) or since the check valve is supposed to be in the pump, I could decide the pump is failed and order a new one (DW this time, probably, to mix it up) to get here probably Wednesday.  I can also try driving it again tomorrow, but not holding pressure in the line certainly appears to be a fault, so I don't expect change.  Just for completeness, I guess.  I could also try putting 25 or so more miles on it and doing the first oil change and check for sparklies.

[subisubisu]

Looking at the diagnostics from the vacation pics for the fuel pump, it seems that 10v is the minimum.  Below 10v, for a non-operating condition, you replace the pump.  It also gives direction to measure resistance across the harness between the pump and the relay, which I can do.  But the inability to hold pressure is distinct from any voltage measurement - it's purely mechanical.  When the check valve comes, and I get it inserted into the supply line in the engine bay (downstream of the pump outlet where presumably the check valve is located, so IF there are no other leaks in that portion of the run it ought to be same-same from the fuel circuit operation standpoint), I hope to tell if I get proper operation.  If so, I am thinking it will mean fuel pump replacement due to a failed check valve.  If not, continue the leakage search.  I haven't been thinking injectors so far because I would have expected some sort of rich indication instead of the lean readings I am seeing, and the positive numbers in the long term fuel trim.

[KZJonny]

If the check valve works for you, and the pump appears to have failed, there'd be no better time to throw the high performace wiring kit at it.

https://www.summitracing.com/parts/dwk-fphwk?seid=srese1&gclid=CjwKCAjwyY6pBhA9EiwAMzmfweY0-7IjGR89l7ZKe3Rpx0dHytbfPtETQ0-BYdTG_NUcTxTNqQjl7BoCc_4QAvD_BwE

Or from a million other places.

Provided you wire it through the Fuel Pump Controller, you still get the stepped duty on the pump but more voltage to to it. 10v is pretty low for a motor designed to run on 12v...

People have had problems with being too rich after doing this install + the STi FPC...

 

Maybe part of your lean condition is the low voltage combined with the lack of a check valve? Ie: the pump doesn't have enough voltage to start off with, and then it is also working overtime to keep the fuel pressure high, since it bleeds out pressure as soon as it cycles down. Doesn't change the diagnostic sequence you are following, but it's curious anyway.

Edited October 9, 2023 by KZJonny

[subisubisu]

Thanks for the link.  Summit gets some of my money on the regular.  I didn't consider the 10v good news, but it doesn't cause a leak.

And the check valve did arrive, and I did get it installed, and the pressure bleeds down anyway.  So now I guess I need to suspect an injector, despite not seeing rich indications.  Maybe one cylinder leaking and going rich is not enough to prevent registering overall lean if the other cylinders are operating at reduced pressure?

[Edited to remove indications of idiocy.]  I will have to investigate how to check the injectors, and whether there is something short of lifting the TGVs from the heads that will show me whether there is leakage from a pressurized system in one or another cylinder.

There is no fuel in the regulator vacuum line that would indicate a leak there through the diaphragm.  No fuel smell in the engine bay unless I disconnect a fuel line or pull an injector, or anywhere else for that matter.  I haven't run into anyone talking about leaks in the return side of the system except for the regulator diaphragm, but I don't really know how that side works, and I didn't really examine it when pulling and reinstalling the fuel pump yesterday.

Edited October 9, 2023 by subisubisu

[subisubisu]

New fuel pump wiring is on the way, but it's not the cause of the presumed leak that I suspect is causing the fall-off in pressure.

I have pulled, vaselined and reseated all of the new DW740 injectors, and still the pressure spikes when I turn the key to Run, and then falls off to zero within a minute.  There  is no fresh raw gas smell anywhere in the engine bay or cabin.  I don't know where to go next, but I have some thoughts:

Option 1 - bite the bullet and start stripping the top of the engine until I can separate the TGV bodies from the heads and pressurize the fuel system and see if there is a leak anywhere.

Option 2 - fire the parts cannon and replace the bucket in the fuel tank.

With no fuel smell, I strongly suspect that the leak is in the tank, but can't discount that a cylinder or two is getting soaked, and I have done such a good job of sealing the vacuum system that I can't tell.  That just made me think -

Option 3 - maybe pull the throttle body hose and see if I detect fresh fuel smell from the manifold when pressurizing the fuel system?  That wouldn't tell me what exactly was leaking, but it would tell me that it's at the engine end and not at the tank end, wouldn't it?

[ETA] Option 4 - Replace the new injectors with the original I pulled, just to see if it holds pressure, but not to run with the new ECU map.  Alternatively or additionally, I could do the same with the fuel pump.  Ideally I would see a different result after changing SOMETHING.  Seeing no change tells me I haven't yet located the problem.

Edited October 16, 2023 by subisubisu

[subisubisu]

I was stumped about how the system could be expected to hold pressure until I realized that the tank is probably supposed to pressurize, too, isn't it?  So if I have an air leak in the tank, filler neck, filler cap, pump bucket gasket seal, etc., then the system won't hold pressure?  If that's true, it will give me more things to inspect, and places to sniff for gas fumes.

[subisubisu]

Sigh.  I guess the tank doesn't really hold that pressure, so again I am a loss to understand how the system really operates.  But...

Progress!  I went for option 2.  I couldn't find a Subaru fuel pump bucket assembly that didn't include the fuel pump itself - $380 or so for the complete assembly, and $260 for just the OEM Subaru pump.  So, I elected for a $90 aftermarket one on Amazon.  It looks pretty similar, but not identical.  I went ahead and installed it as it arrived instead of trying to put the AEM 340 class pump into it, reasoning that I don't know what part of the fuel pump might not be right, so let's change it all and see what happens.

What happened was that it primes to a higher pressure and holds it for a while, before it bleeds off slowly over minutes, but not hours.  Progress!

AND, when I drove the car, no bucking.  Zapping the throttle shows AFR going a bit rich, then a bit lean, then back to stochiometric.

However, the regulator still isn't regulating.  It settles at a higher pressure now, almost 70 psi is indicated.  And it doesn't vary at all with the set screw, or with removing and replacing the vacuum hose (which is indeed pulling vacuum).  I have had the regulator apart and back together a couple of times now, and I don't see any failures inside.  The diaphragm is fine.  It's not a complicated device.  Could the spring be too strong?

The other oddity I am seeing is in the RomRaider instant fuel trims.  The long-term trim for the idle range is -9% or so, which seems reasonable - fuel rail pressure pressure is too high, so the injectors provide too much fuel on the schedule, so the computer is compensating by decreasing.  However, the #1 and #3 instantaneous fuel figures are hugely different:  #1 is in the range of about +1 to 2, whereas #3 is large negative, -50 or so.

[KZJonny]

Don't think the tank is meant to hold injector pressure level of psi, no. As you found out. Couple of psi to account for Evap stuff moving around, but I think that's it?

- Could you get your hands on a scrap fuel rail? Should be able to test for injector leaks that way.... Just push in air or a non-flammable liquid via a hose and some clamps.. Plenty of folks do DIY injector cleaning or the like this way.

- Did you by chance leave the additional check valve in the system after installing the FPR? I woulda assume not, but it's one of those "doh!" things that could be partly explaining the ridiculoulsly high pressure.

- DId you write to Radium yet? See if maybe there is a model number on the pump or spring of some such that has you barking up the wrong tree? Like, if that FPR doesn't operate in the range you're trying to get it to, then you might be spending money and spinning your wheels diagnosing? Might not be likely, but the wrong part CAN end up a box sometimes, or maybe you got send the wrong thing etc....

I am also curious about the specific fuel bucket you got, and how the build quality seems to be? Mine is fine enough, but there are some age cracks in the "lid" and I don't think there is any cure for that.... I managed to get it to seal pretty well with a non-hardening sealant where it meets the tank, but that can't last forever, and as you noted the cost of an OE unit is a little silly, since I really only need the one part. But for $90 USD I could consider throwing a whole new bucket in... that is assuming the aftermarket pump will fit, but I suppose you may figure that out shortly?

[subisubisu]

Thanks for the thoughtful response!  I run out of ideas here on my lonesome.

I do have fuel rails for my deferred plan of conversion to top feed, but the car is set up with side feed.  I need to think about what else I can do there.

I removed the hose with the check valve and built a nice new hose section, so the inline check valve is sitting on a bench.

I called Radium.  Spoke to a very pleasant and patient person who didn't give his name.  He gave me a couple of ideas about things to check.  He said that the usual cause of what I'm seeing is backward installation of the regulator, but I described my installation (and I paid attention to the instructions!) and it seems like I have it right.  I will disassemble the regulator again and check to see that the input orifice is properly seated.  I connected to the car in the same way it was when I got it, and I am supposing that all that is correct - highest  hard line at the firewall and at the engine are the supply.  A little lower is the evap line, and the lowest hard line at either end is the return.  In the tank, the supply is the nearest connection to the center of the cap on the side with two connections, and the return is the single connection.  The hoses in that compartment don't stretch to the wrong connection, which is good planning.

Below is what I ordered.  A comment (meaning some stranger on the internet) said he replaced the pump with his high-flow pump and things are working fine for him.  Build quality didn't raise any eyebrows.  There are some minor-seeming differences from my presumably original one - the hoses have a cloth sheath instead of the corrugated plastic of the original, and what I recall of the original is that the return line dumps into the bottom of the bucket and runs in a passage across to a flapper valve on the opposite side (still within the bucket) but the new one appears to vent back directly into the tank.  I confess that I haven't checked to see whether it's properly reading tank fill level - I've been hyperfocused on my fuel pressure situation.

https://www.amazon.com/Fuel-Assembly-Subaru-Outback-Legacy/dp/B07NQCX75F/ref=asc_df_B07NQCX75F/?tag=hyprod-20&linkCode=df0&hvadid=459441675016&hvpos=&hvnetw=g&hvrand=15082739349785110974&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9012722&hvtargid=pla-942777239009&psc=1

Edited October 23, 2023 by subisubisu

[KZJonny]

Welp. If I can think of anything else, I'll post it up?

I am out of ideas for the moment, and it's not my car, so I can imagine it must be a little frustrating for you.

 

Only other thing I can think of to offer up is that I've got at least a set or two of yellow top side feeds in boxes somewhere. If you need a set to test with, and wanted to pay for return shipping, you could use them to compare to your current set. They're my spare parts since I plan on staying with the side feeds, or I'd not want them back.

For a little more money tho, you could maybe get yours tested somewhere close to you? Most major cities will have a place with an injector testing bench that can look at LPM and spray pattern as well as run a cleaner through etc.... I want to say the place nearest me does side feeds for like... $35/per unit CAD.

I have no idea where 'the rocket city' is, but it's probably further from Canadia than you'd want to have things shipped.

[subisubisu]

The Rocket City is Huntsville, Alabama, so quite a way from our neighbors to the north.  Thank you for your kind offer.  I don't intend to ship anything up thataway in the near future, but feel welcome to send cold fronts south if so inclined. 

I have a couple of things to try with what I've got right now in the next couple of evenings (bar interruptions).  This failure of adjusting the regulator to make the pressure budge has the flavor to me of something that can be fixed, and the Radium guy gave me direction to look in.

Of course there is no law that says that only one thing can be wrong, and I am already seeing that I had multiple things wrong.  I am also seeing that I can't count on the original parts/installations to be "known good" just because the car ran and drove when I took delivery.  Trust nothing, test everything is the direction I am being pushed in.

Edited October 23, 2023 by subisubisu

[subisubisu]

OK, I'll wait now for the admins to change my rank from Collaborator to Idiot.  Despite checking and rechecking, I overlooked that I had connected the FPR backward.  Once I swapped input and output, it works fine.  Don't let me near your cars.

Once I figure out where I had originally placed the now-loose wire that feeds my RomRaider wideband O2 logging, I will finally get some good(-ish) data logged and off to the tuner.  At least the car is driveable now, and I want to get on with the tuning.  Later on I can think about swapping the fuel pump back to a high-flow model, if it looks like I am limited there with the 740 class injectors.

To summarize:  I'm an idiot, Radium tech support was great, FPR working properly.  Read, understand and follow the damned directions.

Edited October 23, 2023 by subisubisu

[KZJonny]