DIY RRFPR

[THAWA]

hmm, we still need to find out what sizes to use for the pistons to keep it at a good ratio, I guess I also need to find out what ratio I need to run.

[vrg3]

If we were to modify a stock FPR we'd need to measure the stock diaphragm area and pick one appropriately larger. Don't RS-T people use 4:1 or 6:1 or something?

Hmm... I wonder if we could find a stock FPR from another make/model car that has a much larger diaphragm... Then maybe you could frankenstein a stock Legacy FPR and that one.

Or maybe you could use the vacuum diaphragm assembly from a distributor's vacuum advance assembly or something.

*gears turning*

[THAWA]

so just to summarize, your idea about the two different sized pistons connected to the fpr would work right? Before you said a spring would be better than a rod between the two pistons, would you still say tthe same thing? If you tee'd the fpr inlet and put a checkvalve to the line going to the fpr would this keep the fpr acting regular under vacuum and the rrfpr sti push air into the fpr under boost? Where would I find cylinders that small to use? or would it be easier to use ball bearings, and where can one find those ? Instead of the spring or rod between the balls/piston, what about using water or soemthing?

[vrg3]

so just to summarize, your idea about the two different sized pistons connected to the fpr would work right?

No. I don't think it would. The air between the doohickey and the FPR would compress.

Before you said a spring would be better than a rod between the two pistons, would you still say tthe same thing?

Now I think it has to be a rod. You can't let the spring absorb the force.

If you tee'd the fpr inlet and put a checkvalve to the line going to the fpr would this keep the fpr acting regular under vacuum and the rrfpr sti push air into the fpr under boost?

To be clear -- this whole time I was thinking of leaving the stock FPR in place and fitting something downstream of it.

But a check valve wouldn't work there because I don't think the idea would work with air in there to begin with.

Where would I find cylinders that small to use? or would it be easier to use ball bearings, and where can one find those :)?

You could probably get a local machine shop to fab up simple pistons if you wanted. They routinely machine things down to the thousandth of an inch. To get even better sealing, have them machine in a groove for an o-ring.

You can get ball bearings at a hardware store. McMaster carries a wide assortment of sizes though if the hardware store's selection is insufficient.

Instead of the spring or rod between the balls/piston, what about using water or soemthing?

Yeah, you could use an incompressible fluid.

But anyway, let me explicitly explain what my idea du jour is:

Get a stock FPR. Internally, it looks something like this:
http://www.graphics.cornell.edu/~v/pics/fpr.png

Cut off the cap on top. Remove the spring (I think that's a spring between the diaphragm and the top). Keep the part of the cap that fits over the spring though.

Obtain another device that has a vacuum/pressure-actuated diaphragm. The diaphragm needs to be bigger in area than the diaphragm of the FPR, proportionally to the desired pressure rise ratio. For example, if you want a 4:1 rise ratio, the diameter should be twice as large as the FPR's diaphragm. This device could be another car's FPR, a vacuum-advance mechanism off a distributor, a wastegate actuator, an EGR valve, etc.

Cut the top of this device off. You want the cap with the vacuum nipple on it, and the diaphragm attached. Hopefully the diaphragm is also attached to some kind of rod.

Attach the rod to the FPR's spring cap. Then find a way to close the whole air chamber (maybe just wrap it in sheet metal and solder/braze/weld it shut).

As of my current understanding, this would result in a fuel pressure regulator with the same base pressure as stock but with a rising rate. The problem is that it would have a rising rate even off-boost. Maybe some kind of bypass valve that opens off boost could resolve that.

Realize that this is all theoretical and made up in my head. It's quite possible that this isn't possible or wouldn't work right.

[THAWA]

okay so this new idea is pretty much like using both of our original ideas together? Cept using a fpr instead of the 4way fitting. So do you think it could be done with a 4way fitting having the smaller part of your idea be the small cyl in my idea, and have a large cyl on top that would have a cyl/nipple leading in and all that stuff? And just to be sure it looks like to find out what size relation it should be between the small piston and the larger one you take the square root of the ratio you want and multiply it by the sie of the smaller piston. So if say the small piston is 10mm in diam and i wanted a 4:1 ratio I would need the larger to be 20mm, if I wanted a 6:1 the larger would need to be about 24.5 and an 8:1 would be a lil more than 28mm right? What's been bothering me though is, does the diameter of the fuelflowing through make a difference in the pressure? And Would this ever cut the flow off completely, like if there was x amount of boost pusing the cyl down, would it close the opening between the inlet and outlet or would the stock fpr keep enough fuel flowing through to keep it...I think I'm basically explaining in a reverse question how a rrfpr works. Basically what I want to know is, will the cyl ever be pushed so far that it closes the fuel passage or am I just retarded and not thinking correctly about how the rrfpr works? (i think i confused myself :\)

[vrg3]

okay so this new idea is pretty much like using both of our original ideas together?

I was thinking of it as using my original idea but using the original FPR diaphragm instead of the small ball bearing.

Cept using a fpr instead of the 4way fitting.

Right, since your original design was pretty much just an FPR.

So yeah. :)

So do you think it could be done with a 4way fitting having the smaller part of your idea be the small cyl in my idea, and have a large cyl on top that would have a cyl/nipple leading in and all that stuff?

Well, I really think using diaphragms is the way to go. You need the air and fuel chambers to be well sealed from each other, considering the pressure differential between them..

And just to be sure it looks like to find out what size relation it should be between the small piston and the larger one you take the square root of the ratio you want and multiply it by the sie of the smaller piston.

Yeah, I think that's right. It's the area that matters, right?

What's been bothering me though is, does the diameter of the fuelflowing through make a difference in the pressure?

Diameter of the fuel flow? The fuel pressure pushes on the diaphragm of the FPR, so it's the area of that diaphragm that we'd be concerned with.

Basically what I want to know is, will the cyl ever be pushed so far that it closes the fuel passage

Only if the fuel pump is too weak to create enough fuel pressure, right? The valve would close if boost pressure (multiplied by the diaphragm size difference and added to whatever preload the spring provides) exceeds fuel pressure. That allows fuel pressure to build until it's strong enough to open the valve again.

[THAWA]

hehe yeah I knew I was just getting confused. the closing thing is what makes a regulator a regulator! but uh, what about doing like you said with the o-ring and like a lil bit of oil around it? if you cant tell I dont want to mess around with cutting a fpr open

[vrg3]

What would the o-ring and oil around it seal? At that point I was still just talking about air against air.

Whatever you do, you're going to need to contact fuel at some point. I'd feel safer using the relevant bits of a purposely-designed fuel pressure regulator than a bunch of pieces out of the bins at the hardware store, wouldn't you?

Why are you queasy about cutting an FPR open? It seems like a pretty elegant solution to me. If you're lucky you could even fit the housing of whatever the other diaphragm you get over the FPR's body, and it could end up looking almost as if it were designed that way from the factory.

[THAWA]

hmm, yeah I guess that makes sense, just seems like itd be more work geting the air part to seal back up but its pobably better than having fuel dumped through this dealie straight into the manifold

[vrg3]

Yup. :)

A very small leak on the air side wouldn't really be that big a deal, either, especially if your stock FPR was still installed.

[DOA]

Can I suggest that if you go any further with the idea, you finalise on a diaphragm system as its essentially frictionless. Getting a piston to seal correctly for air is a complete nightmare unless you have rather more friction than the air pressures involved can move. Ive recently been playing with ideas for a boost controller and even with a full workshop of cnc machines and all the orings, piston seals and all else I could want I couldnt make a piston that would reliably move at the pressures that the turbo puts out. Not only that but having gone to try out "proper" pneumatics kit and talking to some reps from a couple of companies on the QT thay all say that the pressures involved simply arent sufficient for a piston to work reliably.
Also, I would think that 2 diaphragms with no connection between them would work just as well as with a connection if not better as it would give the damping you seem to want and the pressures involved arent high enough to compress the air enough to make a difference not to mention the effect of a rigid link between the 2 diaphragms adding inertia to the whole system or the fact that the diaphragms will move at different rates wth regard to position. I also would have thought that a hydraulic type system wouldnt have the sorts of response times that you would require due to the higher inertia of the fluid and the higher viscosity of any liquid over a gas.
Few thoughts for you to think on anyway.
OOh, one last thought, if you feel the need to bleed boost from the input to allow the system to reset then please bleed it directly back to the inlet, most of these so called boost controllers dont do this and its just a waste of good boost however small an amount it may be.

[DOA]

One other thought, have you considered the ratio of the pressures that your working with here, i.e fuel pressure versus boost pressure. Your going to need some serious pressure amplification to create a fuel restriction arent you? Your probably better off isolating the air bit completely and using a linkage to turn a ball valve or some other valve that doesnt use a piston type arangement as that would need far less pressure to simply turn than to actively push into the pressure. It would also make it easier to adjust by using a longer or shorter ball valve arm. Maybe even use a stock diaphragm cylinder similar to a wastegate to act directly on an arm on the ball valve though you would need some way of retracting the actuating rod and would have to make sure the cylinder had enough travel to move the arm.

[DOA]

*Suddenly remembers cost is an issue. Seriously, you will find it MUCH cheeper to buy something ready made off the shelf. Nice idea, but mass production has some major cost advantages and if you dont have access to lots of free stuff one offs like this end up being very expensive, I design bespoke machines all the time and customers are always amazed at the costs for one off jobs.
P.S. Sorry if the above sounds at all patronising lol, it aint meant that way.

[THAWA]

Cool beans, I can't say I agree with you on the price thing, even using the system we explained here is going to be at least 20% of the price of a premade rrfpr.

This ball valve can you explain it better? I don't quite understand how it works.

I was just thinking that since the stock fpr can raise fuel pressure normally it'd be alright to use one with more air pressure and would raise fuel pressure even more, dunno if that explains how I was thinking or not.

What kind of liquid system would you suggest? Can you give an example?

Good info though

[DOA]

Right, a ball valve is just a ball with a hole drilled through it so that when the ball turns, it either restricts the flow or allows it through, imagine it as a spherical bearing or rose joint thats put in a pipe. Instead of a piston creating the restriction by moving into the flow it turns into the flow (just like a throttle butterfly lol).
Think you may have read the fluid bit wrong there as Id recomend you dont go that way on response time grounds.
Having re-read your posts though I now follow what your up to a bit better, basically using a large FPR to boost the signal to a smaller FPR (correct me if im wrong) or alter the stock fpr's air signal. Far better to go a route like that as you dont have to play around getting tolerances right and so on and you know the bits both work.
Having said the above though and not knowing enough really about the way the engines plumbed Id have to say that the way the FPR is powered is a bit shonky if im getting the jist of how it works right (more boost signal to FPR = more fuel flowed) as it seems a bit backwards to me. What I mean by hat is that if you think about it, if the FPR needs the boost to be there in the first place to raise the fuel flow then isnt it really lagging behind the whole system at all times? Would it not be better to take a signal from the throttle pot and from a load source and feed them into a little ECU that could power a stepper motor to adjust a ball valve or even the stock FPR (after some mods of course), that way you could programme it to do exactly what you want when you want it and it would always be flowing fuel just when it was needed. Obviously thats probably a bit OTT and costly (not to mention the fact that I suspect that Im a bit out on my reconing about how far behind the FPR is from the fuel flow needs) but vrg3 seems to be a leccy wiz and I know that ECU's can be built for fuel injection from info on a website (which Ive no doubt Ill never find again lol) for very little cost.
Fookin eck, thats enough of my essays for one night, I should be sleeping lol.

[THAWA]

hmm, okay so no fluid .
I didn't think the fpr was electrical in any way, but what did you mean by the fpr's air signal? Also you're thinking about the fpr backwards. When more air goes into the top chamber it pushes the diaphragm down twords the fuel outlet and this lowers the fuel flow which increases the fuel pressure behind the outlet, like putting your thumb over a garden hose.

[vrg3]

That's a good point that a piston powered by manifold pressure won't be able to push on pressurized fuel; I missed that . A ball valve with a linkage operating it sounds like it would work, but my idea of Frankensteining two FPRs together would probably be simpler to accomplish.

About the feedback loop -- the purpose of the FPR is to maintain a constant pressure differential between the fuel rail and the intake manifold. That way the injected fuel quantity is always the same for a particular injector pulse width (and battery voltage, to be nitpicky). So the FPR isn't actually trying to flow more or less fuel to keep up with engine demand or do anything else intelligent; it's just trying to maintain that constant differential pressure. The feedback loop stabilizes very fast. The idea behind a rising rate fuel pressure regulator is to trick the system into running a bit richer at higher manifold pressures.

In fact, if you somehow used inputs like the TPS to infer the appropriate FPR opening, you'd probably end up a little ahead and end up with unpredictable fueling under transient conditions.

I think the web site you're thinking of is http://www.bgsoflex.com/megasquirt.html . MegaSquirt is awesome.

[DOA]

Cheers for the link to the megasquirt, been thinking of looking that up again for ages now lol.
By air signal I just mean the air that pushes on the diaphragm, just a term of phrase if you like.
Thinking about the whole thing again though, what you are trying to acheive is to gain a rising rate fuel pressure related to boost pressure yes? You are also talking as if you want to cut one open at some stage so....... Looking at the picture of an fprs internals, could you not simply change the spring inside that to suit the job? Maybe even replace the top of the fpr with one that could adjust the spring pressure.
One more thing, just clear my mind up, whats operating the original fpr, vacuum or boost lol, still aint worked that out lol.

[vrg3]

The stock FPR is operated by manifold pressure (boost or vacuum; it doesn't matter) plus the spring both pushing against fuel pressure. It maintains the invariant condition:

(1 * P_manifold + 36.3 psi = P_fuel).

If you put a stiffer spring in the stock FPR (or just increase the preload on the spring) you'll get higher static fuel pressure but the ratio will remain 1:1. That is, you change the 36.3 but not the 1. We're trying to achieve a situation where the coefficient of manifold pressure increases without the constant 36.3 offset changing. Ideally, also, the gain would only start happening above a certain pressure.

I found something in my garage that interested me: a wastegate actuator from an IHI RHB5 turbo. I don't know if it came off an 80s Subaru, a Legacy, a Mazda, or what... I don't know if the actuator would be the same on different cars or not anyway. But anyway, this actuator looks like it's got a diaphragm with about 4 times the surface area of the stock Legacy FPR! I think this might be the start of something workable...

Incidentally, if you do want to just increase static fuel pressure, you can increase the preload on the FPR's spring by crushing the cap of the FPR in a little bit. A large brass drift can do it. The easiest way to do it without having to guess too much is to install a fuel pressure gauge on your car, jumper the fuel pump on, and slowly crush the cap, keeping an eye on the gauge, stopping when you get the desired pressure.