Fuel pressure regulation

On earlier FJR models there was a fuel pressure regulator that was referenced to the intake air pressure.

Fuel rail pressure - Intake air pressure = constant = 36 psi

On the newer FJR models the pressure regulator has been moved to the fuel pump in the tank. The regulated pressure is now 47 psi.

Why is the reference to the intake air pressure no longer needed?

The Gen I regulator uses a vacumn reference signal to vary rail pressure. High vacumn (idle) opens the regulator to it's maximum, reducing fuel pressure to it's lowest, while 0 vacumn (WFO), closes the regulator, allowing max fuel pressure and flow for the injectors. I am not yet familiar with the newer system. The pump may cycle dependent upon rail pressure, the ecm may take rail pressure into account when driving the injector pulse (shorter dwell time at idle etc), I don't know. 47psi is a pretty high pressure to carry, but I imagine the injector spray is finer as a result, which means better atomization and less pooling, emissions, and less fuel usage per horsepower, which should translate into better economy.

The purpose of the vacuum regulated rail pressure is to make the pressure across the injector constant rather than making the rail pressure with constant with respect to the atmosphere. This means for a fixed injector pulse width, nearly the same amount of fuel will be sprayed into the port. I never liked this idea too much because I really doubted that the regulator compensated for the dynamics in the system quickly enough. This is unchecked and untested, however, so dont quote me on it. My gut feeling says the reason for the change has to do with drift in the properties of the regulation circuit, difficulty dealing with its transient response, or both.

The more classic port fuel injection systems use the vacuum referenced fuel pressure regulator to keep the fuel pressure drop across the injector always constant as was mentioned. As intake vacuum drops the pressure goes up and as intake vaccum increases the fuel pressure decreases. It is plenty fast to compensate for transient conditions, trust me. Put a gauge on it and watch it. The return style systems allow the injector to deliver the same exact amount of fuel for a given pulse width (injector opening/closing event) regardless of manifold vacuum since the fuel pressure was tracking it accordingly. Rather than distrusting the return style systems you really liked them without knowing it.

The switch to return-less systems is driven mainly by cost (less fittings, pipes, lines, holes in the fuel tank, hardware in general) and the ever tightening evaporative emissions standards. The return style systems are always circulating fuel thru the (hot) fuel rail thus increasing the heat input into the fuel in the tank. The hotter the fuel in the tank, the more vapor is created to be dealt with by the evaporative emissions system (the charcoal canister). Elminate the return fuel and the system elminates a lot of vapor generation. There was never an issue that I have heard about with system pressure drift or anything like that with return style systems. In addition, with the returnless systems there less line connections to (potentially) leak and a simpler system to build in the assembly plant with less lines to route and connect. Anytime you can elminate parts the reliability goes up as there are less parts to fail.

As mentioned in the other post https://www.fjrforum.com/forum//index.php?s...c=20185&hl= the return-less fuel rails do have to contend with the water hammer effect in addition to the ever-changing pressure drop across the injector changing the reference injector pulse...which is a constant in a return sytsem but now varies for each load and speed point with the return-less systems. This means that for a given injector pulse width the injector would deliver more fuel at high vaccum conditions (closed throttle) and less fuel when manifold vaccum is zero (full throttle conditions.)

The ECM is programmed with an injector pulse offset or compensation for each speed load point in the basic fuel map to account for the changing pressure drop across the injector. So, the ECM will "delete" some fuel from the calculated pulse width at high vaccum conditions and "add" some fuel to the calculated pulse width at low vaccum conditions so as to compensate for the lower amount of fuel delivery.

Aside from the advantages of the return-less system mentioned one of the key enablers of the return-less systems were ECM's with porcessors fast enough to do the extra compensation calculation for each cylinder at each speed and load point. Older ECM's of yesteryear were simply not fast enough and did not have enough computing capacity to calculate and deliver the injector correction for each cylinder event. Modern day ECM's have dramitically more computing power and much faster speed (just like in the desktop computer market) allowing cylinder by cylinder compensation and correct in real time....for each cylinder...for each event.

It was not as simple as just deleting the fuel pressure regulator and calling it good.

The other nice thing about return style systems was that the fuel injectors have effectively more "range". Since the injector "on time", or pulse width, has to be artificially increased when the vacuum in the intake is zero (full throttle or WOT) to compensate for the lack of fuel pressure increase the injector just lost some dynamic range. As a result, it is common to see larger injectors fitted into systems that are converted to return-less.

Two simple reasons for the fuel pressure increase on the returnless systems. On is to offset the loss of dynamic range just mentioned. If the pressure is increased in the fuel rail the injector will now have to open less to flow the same amount thus offsetting the loss of flow at the low vaccum conditions. In effect, the pressure goes full time to what it was at when the fuel return type system was compensating for low vaccum conditions. The other reason the fuel pressure is increased is to fend off vapor lock. Since the fuel is dead headed in the rail and not constantly replaced by cooler fuel circulating thru the rail the fuel has a tendency to pick up more heat and "boil" or flash into vapor. Especially in a hot soak condition. As long as the rail maintains fuel pressure during shutdown the higher pressure will prevent boiling and vapor lock. Along this train of thought, keep the water hammer effect in mind. As the negative pulses travel up and down the rail it is desireable to always keep the fuel pressurized to a point that vapor cannot form. If the fuel pressure too low then the even lower value seen during a low pressure spike could allow vapor fo form. Keeping the pressure higher offsets the low pressure spikes sufficiently that the fuel cannot spike low enough in pressure to allow vapor to form.

Good info Jestal. I always wondered about the transient response of those things. Thanks

I'd bet the primary motivation was the heat issue, as mentioned in the other thread. I havent ridden an 03-05 for comparison, but I was wondering if they did something like that cuz I couldnt believe how cool the tank stays on my new FJR compared to my dads ST.

It would seem if larger capacity injectors are used that you'd reach a point for conditions of low fuel requirements (idle, etc.) where the injector 'open' time interval would be very small. Maybe that's why in some systems two smaller injectors are used for each cylinder rather than one large injector.

Maybe the more powerful ECU's also allow for the deletion of an atmospheric pressure sensor. The newer FJR's don't have one. Maybe the ECU can calculate an approximate atmospheric pressure using the intake pressure, engine rpm, and throttle valve position.

This makes for a fascinating read, thanks all! (I just wish I could fully absorb the subject matter without getting a splitting headache... )

I have a question here that may seem off the wall...

Elsewhere on this Forum a strange behavior of Gen.II FJRs has been debated, that is their tendency to "stumble" upon acceleration, especially at a certain elevation (typically 3,000+ ft).

Not surprisingly, Gen.II FJRs are showing the same unpleasant symptoms here in Europe. Some riders are even compelled to kill the motor during a climb and fire it again to get the motorbike to run smoothly. This fix seems to work (while obviously being an unacceptable solution in the long run)

Could this phenomenon be connected to the engineering changes you just described or am I totally off track?

Stef

Constant Mesh said:

It would seem if larger capacity injectors are used that you'd reach a point for conditions of low fuel requirements (idle, etc.) where the injector 'open' time interval would be very small. Maybe that's why in some systems two smaller injectors are used for each cylinder rather than one large injector.
Maybe the more powerful ECU's also allow for the deletion of an atmospheric pressure sensor. The newer FJR's don't have one. Maybe the ECU can calculate an approximate atmospheric pressure using the intake pressure, engine rpm, and throttle valve position.

Click to expand...

That is exactly the problem with larger injectors. The lower the pulse width (like at idle and decels) the less accurate the fuel delivery so it is desireable to size the injectors to be just large enough for the system under consideration (basically dictated by the power level of the engine and the available on-time) so that they can still operate in their repeatable range at idle. Most production systems are not really pressing the limit of idle/overrun condition pulse width limitations but it always has to be considered.

More likely a cost consideration. This was discussed once before. The deletion of the barometric pressure sensor saves money on the sensor, the wiring and connectors, ECM input connections and manufacturing time/labor. Less complexity and mass, too. It really has nothing to do with ECM speed or capability. There were "baro-less" sensor systems in production 20 years ago with very slow ECM's. Baro-less systems take a snap shot of the MAP sensor reading when you turn the key on....between key on and crank....to measure the barometric pressure. The "baro-less" barometer term is updated while the engine is running when throttle openings reach the point that the induction system is no longer throttled. This depends on RPM and throttle angle but it doesn't take a "fast" ECM nor much computing power. They are also updated anytime the MAP exceeds the stored baro term regardless of throttle opening. Like when you would start higher elevation and ride downhill for a long time without opening the throttle to WOT. The stored baro term might be quiet low and as soon as the system sees a MAP reading higher than the stored baro it updates to the larger of the two since MAP cannot be higher than baro...at least on this planet.

The problem you mention teerex could certainly be related to the baro-less barometric pressure model that Yamaha has in their system. I may be giving them too much credit for actually having the baro-less model working correctly or adequately enough to prevent driveability issues. I suspect that they are probably close on driveability issues to meet emission compliance with the cat and to protect the cat on engine over-run condition which makes the engine borderline lean on some part throttle operating conditions and then when a barometric pressure "error" in the baro-less modeling of the baro term compounds this leaness it makes for poor driveability.

I still believe that Yamaha may have been seeing an issue with catalyst melting and/or overtemping on decels so they incorporated decel fuel cutoff in the 06/07 models and/or made the use of decel fuel cutoff much more pervasive in the calibration. This leads to the hesitations and herks and jerks people experience when going from closed throttle to open throttle at speed. If the injector pulse widths were marginally low at certain overrun conditions for the injector to operate cleanly then a baro term error could exacerbate the problem. Just guessing here but it seems to fit my preconceived notions of what is happening.

It would be interesting to see what happened in those cases if the rider simply went full throttle in a higher gear....sort of lug the engine for a brief interval (just seconds) to see if the baro term would update and alleviate the issue without turning the engine off and restarting. Certainly the baro-less term updates with key on using the MAP reading but there must be an additional update with the engine running and the throttle well open. The update may not happen if the engine is held at high RPM since it takes full WOT to get the inductin unthrottled at high RPM...like a lot of people would do during a spirited climb. The baro-less update could be inhibited by the higher RPM operation. Returning briefly to lower RPM and full WOT by lugging briefly in high gear might update it. Maybe not...... This sort of thing is pretty elementary to electronic fuel injection systems and it is surprising to see yamaha stumble (no pun intended) on this. Certainly not new ground as much of this sort of "technology" was done in the auto industry 20 or 25 years ago. Certainly shouldn't have to ride around the baro-less sensor calibration. Poor execution on Yamaha's part.

Looking at the '06 R6 bits and '07 R1 bits in Yamaha's parts catalog reveals:

The R6 has a two hose fuel return system. The pressure regulator resides at the fuel pump. The fuel rail has a pulse dampener between injectors 2 and 3.

The R1 has one fuel hose and is a non-return system. The pressure regulator resides at the fuel pump. The fuel rail has a pulse dampener between injectors 2 and 3.