DNMakinson

iTrader: (3)

This line of inquiry got me thinking again about manifold referenced FPR vs Atmosphere referenced.

If you tune a Manifold referenced VE table at sea level, then the MAP to Fuel pressure remains the same at all elevations. Baro would only need to correct for the back pressure issue. However, if you have a non-return, atmosphere referenced FPR, then the absolute fuel pressure will be lower at the higher altitude, thence the differential across the injectors will be less.

For example:
At 100kPa MAP, the NB system will have 60 psi fuel to MAP differential at sea level (assume sea level Baro is 100kPa).
The same 100kPa MAP on that car, when at 8000 feet elevation will have a Baro of 75kPa, and therefore an absolute fuel pressure that is low by 3.7 psi (25kPa).
That would yield a pressure differential of 56.3 psi, fuel to MAP differential. The resultant fuel flow, for the same pulse width would be SQRT(56.3 / 60) or 97% fueling. Down by 3%.

A MAP (Manifold) referenced system would not see this reduction in fueling.

This is true whether NA or boosted. Another reason to use Baro correction, or Manifold referenced FPR.

noname4me

Is there an issue using BOTH Baro correction and a manifold referenced FPR (i.e. like the OEM FPR)? Does this cause the MS3 to double correct for changes in base Baro pressure? If yes, logically one could just use a manifold referenced FRP and skip Baro correction. (I believe the answer is no... Baro correction should be used with a manifold referenced FPR, but MS firmware/logic is a PITA at times...)

Thanks

18psi

iTrader: (76)

Pretty sure you still use baro correction, but a referenced return-style system is a whole lot easier to deal with and plays nice with the engine/ecu

DNMakinson

iTrader: (3)

I am assuming both is advantagous. I mentioned what I did as some say that baro is not needed with a boosted car, but there is still a place for it if atmosphere referenced fueling. Also, I'm looking for 18's correction curve for this set-up and wanted to point out that the curve will be somewhat different for a return style fueling system.

18:
1) I presume this car is returnless?
2) How do you plan to calibrate the sensor?
3) How do you plan to set the correction curve if you only have the car at sea level? Is the owner going to add the curve when he goes to altitude?

Savington

iTrader: (31)

I think you forgot about the pressure in the manifold itself. The difference in pressure between the fuel rail and the manifold is what forces fuel through the injector. If you reference the regulator off of atmosphere, that's what ensures that the differential stays even as you go up and down in altitude.

~414kpa = absolute fuel pressure
~100kpa = manifold/atmospheric pressure

414-100 = 314kpa differential pressure

Now go up 8000ft.
414kpa - 25kpa (atmospheric reference) = 389kpa absolute fuel pressure
100kpa - 25kpa = 75kpa = atmospheric pressure and manifold pressure at WOT

389-75kpa = 314kpa differential pressure, same as at sea level.

The barometric correction compensates for the reduction in pressure on the exhaust side, which has to be compensated for separately. IOW, you need both.

DNMakinson

iTrader: (3)

Savington. The problem with my example is that it pertained to a boosted car.... running 100kPa MAP (Manifold Absolute Pressure) at both sea level and 8000'. I realized later that I had used a bad example for a NA car, but since no-one responded, I did not correct.

The problem with your example is that the VE table will be in a different row (100kPa at sea level, 75kPa at 8000'), and therefore not the same pulse width called for. If you take the VE table into account, then again, you will be under-fueled.

To redo your example, but correctly at 75kPa MAP at sea level and 8000' (thence the same VE table cell for fueling), it would look like this:
~414kPa absolute fuel pressure
~75kPa MAP (not same as atmospheric, but throttled to make same row in VE table for the next comparison at altitude.

414-75 = 339kpa differential pressure

Now go up 8000ft.
414kpa - 25kpa (atmospheric reference) = 389kpa absolute fuel pressure
100kpa - 25kpa = 75kpa = atmospheric pressure and manifold pressure at WOT

389-75kpa = 314kpa differential pressure, different from value at sea level for the same cell (row of cells) in the VE table.

I think that is the correct example.

noname4me

I believe that the proverbial apples and oranges have gotten mixed up...

This thread started on the topic of barometric correction/reference for the ECU, but then the topic of FRP's was pretty quickly added/mixed into the discussion. Also, we have been a bit sloppy in terminology... "barometric" pressure refers to atmospheric pressure specifically. At least in regards to the FPR's, every OEM one that I can think of is referenced to intake manifold pressure (not barometric pressure), if the FPR is air pressure referenced in the first place.

So Savington's example is about Fuel Pressure, not MAP as seen by the ECU. By referencing the FPR to the intake manifold pressure, it assures that the pressure differential between the fuel leaving the injector nozzles and the intake tract/manifold is consistent, no matter whether you are in Death Valley or cresting to Continental Divide. For example, it would always be 314kpa in Savington's scenario.

Getting back to Megasquirt ECU's, MS firmware/code behavior, VE tables, etc...

With an intake manifold pressure referenced FPR and a Speed Density ECU strategy, I don't actually quite see the point of having the MS ECU reference barometric (i.e. atmospheric) pressure. Take away the FPR's reference to the intake pressure and I can understand why the ECU would need to compensate for the changes in effective fuel pressure/delivery and thus need a barometric reference. A baro reference for the ECU also makes sense with an Alpha-N type ECU tune. However with Speed Density, I personally don't get it...

Using Savington's example, the MAP reading from the intake would be 75kpa @ 8000ft and thus the MS would use the 75kpa row from the VE table. If it "corrected" for the 25kpa change from the higher altitude and did in fact use the 100kpa row, the wrong VE value would be used! Afterall, if the pressure is 75kpa in the intake, it is 75kpa in the intake! The reason why it is 75kpa in the intake doesn't change the fuel needed to achieve a certain AFR, as that is always a matter of the number of O2 molecules in the intake charge air.

Hopefully someone can shed some light on how the Megasquirt's use the baro reference with the typical Speed Density strategy and an intake manifold pressure referenced FPR. There could very well be a great reason that isn't obvious (to me)...? Maybe something to do with the difference between FuelLoad vs MAP?

18psi

iTrader: (76)

99+ NB's use a return-less system.
OP's car is still returnless, and therefore needs the correction.

Savington

iTrader: (31)

Actually, David is right for everything except WOT. At WOT, the atmosphere-referenced FPR will deliver an accurate fuel pulse, but at anything below WOT, it will be a little lean.

The need arises from the reduction in backpressure in the exhaust. You're essentially drawing a vacuum on the exhaust system, which will improve VE.

noname4me

Sorry for the long post... I have put a lot of thought into some of these topics while tuning my MS3X over the past 3 years. If the post is too long, please ignore and go onto something less verbose.


1) If the NB returnless FPR isn't referenced to the intake pressure, then the fuel pressure differential will vary based on the intake pressure. However the delta in effective fuel delivery between different intake pressures will just end up baked into the VE values in the ECU/MS when the car is tuned (manually or using VEAL).

Yes, if one actually injected fuel using the same PW@ different intake pressures with a constant fuel pressure in the rail, you would see the AFR change based on intake pressure. In the real world we tune usually for some specific AFR, so the required change in PW to account for the change in pressure differential and effective fuel delivery ends up being included in the VE changes from one row to the next in the VE table.

2) The OEM FPR's for return type systems are usually intake manifold referenced so that the required PW values/durations don't end up going lower than what the injectors can do in a stable manner (with 80's-90's injector tech).

As MAP drops at idle, the pressure differential actually increases unless the FPR adjusts, meaning that shorter injector PW's are needed to deliver the same amount of fuel vs higher intake pressures (closer to WOT/atmospheric pressure). This makes idle fueling even harder for the old tech injectors, unless the FPR reduces the rail pressure to keep a constant differential injection pressure. Mazda OEM manuals actual call this out as the reason for the FPR - intake vacuum switch, which is between the intake and the FPR on NA return type systems.

3) I wasn't commenting directly on the example given by DNMakinson. I agree that fuel pressure differential would change as his math shows in his specific scenario, but that isn't the behavior one would want is it? Also, comparing <WOT @ Sea Level to WOT @ 8000ft makes the example/logic hard to follow.

I don't believe one wants the fuel pressure differential to change based directly on baro pressure, without regard for the actual intake pressure. Remember that with an intake referenced FPR, the differential is held constant.

Higher elevations are going to mean that WOT is <100kpa, so that lower rows with lower VE values are going to be used from the VE table, reducing the fuel injected and compensating for the reduced air/O2 in each intake charge. Having a baro referenced FPR lower the fuel pressure differentials AND hitting lower rows on the VE table sounds like a recipe for a lean running engine.

4) If lower baro pressures reduce exhaust back pressure significantly, wouldn't that tend to make for a more efficient engine (and the need for more fuel), since pumping losses would be lower?

5) We really need to take a look at the logic of how the MS code uses the barometric reference pressure. That would help explain a lot...

6) To have a safe running engine at different altitudes, the key is going to be to use the Incorporate AFR option and not use MAP as the Y axis for the AFR table. Afterall, one probably still wants certain AFR's at WOT, even with the intake pressure lower than normal @ Sea Level. This is why I was asking about FuelLoad... The VE table should be based on MAP, though that raises questions about VE varying based on throttle blade position not being taken into account.

BTW... The idea that a home/hobbyist can truly put together an accurate VE map which represents the true volumetric efficiency of the engine without all kinds of "external" factors being reflected in the VE values is not realistic. If we had accurate VE values, then all of the theoretical modifications/calculations would work much better than they do in the real world. So instead, we need to rely on feedback from the O2 sensors and hack the tunes until the tune is close enough for the O2 feedback to correct.

Thank you, great discussion!

DNMakinson

iTrader: (3)

Returnless, by definition, means atmospheric referenced.
The primary fueling Always uses the Manifold Pressure for SD. This discussion is about a second sensor that reads atmospheric pressure and adds a small correction based on that sensor's reading.

EDIT: Rest deleted as too much to address this early in the morning.

noname4me

I don't believe that a returnless system "must" use a FPR referenced to atmosphere. A FPR could be set to work against a fixed, internal resistance like a spring or sealed internal chamber and not have any external air pressure reference. Springs are common with carb FPR's for example. I do not know the inner workings of the NB returnless FPR and am not claiming such. If it is open to the atmosphere, then the OEM NB ECU is programmed to take the change in fuel pressure into account.

I understand the 2nd MAP sensor being used to read the barometric pressure. I have one installed in my Rev built MS3X. I had it added as an option when he built my MS3X 3+ years ago.

Actually opening up Tunerstudio and the default MS3 project (don't have my car's tune on this computer), the VE table is actually using "fuelload%" as the Y-axis and AFR Table 1 is using "afrload %". So neither is using the MAP signal directly/alone.

Note: Under Basic/Load settings, these is an option for a "Non-linear Barometric Correction" table, which I have never personally used before in my MS3 tunes.

I need to read up on the latest MS3 1.4.x documentation and MS forums. The key is going to be understanding specifically how the MS code uses the baro pressure data to influence the actual fuel PW commanded at any specific combinations of baro pressure, intake pressure, and RPM.

I 100% get that changes in baro pressure will have an impact on engine efficiency. I just doubt that one can simply use a FPR with atmospheric pressure to vary the rail pressure and thus fuel pressure differential directly on baro pressure, without any other consideration for baro pressure in the ECU tables/logic. Obviously this is why there is an option with the MS ECU's to have the 2nd MAP sensor provide baro data.

DNMakinson

iTrader: (3)

You will find another place that shows that "fuelload%" is actaully MAP reading, This is selectable, but I'd be completly surprised if yours was set to anything other than MAP.

codrus

The NB's regulator is inside the tank, so it's not really atmosphere referenced. At best, it's internal tank pressure-referenced, which is not quite the same thing.

--Ian

DNMakinson

iTrader: (3)

I wondered if anyone would mention that. So, when the tank is pressurized, the situation is even worse.
i definitely will be going return system.
EDIT: Brain fart… problem is less if tank is pressurized.

18psi

iTrader: (76)

Split from this thread:
https://www.miataturbo.net/megasquir...3-basic-87969/

To keep discussion in there on point.

Created this thread for barometric pressure/fuel pressure regulator reference/fueling discussions



@DNMakinson let me know if you want me to re-name this one or update the thread title or something

noname4me

My tune is actually set to "Use Primary Load (Algorithm)" for the AFR and EAE loads. I will have to dig in and research exactly what "Algorithm" is used by the MS3 1.4.x code in this case.

Again with my tune, Primary Fuel and Ignition Loads are set to "Speed Density", not "Percent Baro". Per the help bubbles in TS 2.6.14, "Percent Baro" = Speed Density + Barometric corrections. From the help bubble for Primary Fuel Load:

"The calculation method for fuel. The choices are:
Speed Density (uses MAP sensor)
Percent Baro (Speed-Density but with barometric pressure difference included)"

BTW... The help bubbles in the Calibrate MAP/Baro screen in TS are written better than the online manual. From the Barometric Correction, At Total Vacuum(%) help in TS:

"The barometeric correction to fuel at a total vacuum.
For new installs set these two settings to zero and the barometric correction curve to 100% across the board.
Imported settings from older code versions use a vacuum of 147, a rate of -47 and have the curve at zero."

In the case of my tune, I do not believe any barometric correction is actually being applied since my Barometric Correction curve is set to 100% across the table and I have zero'd out the Barometric Correction settings in the MAP/Baro Sensor calibration. However, I need to determine exactly how "afrload1" is calculated to verify. Of course in the big picture, I actually would like my Miata to run correctly at varying altitudes, so I need to work out how best to set all of these configuration parameters and tune the Barometric Correction curve correctly.

Does anyone have a good how-to or reference?