Barometric fuelling and timing correction

[Philip Lochner]

This thread refers: http://www.msextra.com/forums/viewtopic ... 94&t=30599. From this you can see that I have "a thing" about the barometric effect on a normally aspirated high compression engine.

Those with normally aspirated, high compression* (HC) engines (Typ 9:1 or more), that drive in areas with widely fluctuating barometric pressure should be interested in this. I have found that tuning HC engines where I live at 1400m AMSL (4600') and then driving them to areas at 1000m (3300') or lower results in the engine starting to detonate / ping / knock. This then requires for the ignition map to be "lowered" (retarded) as a whole. I wanted this adjustment to occur automatically which this code now does.

Those that tune their cars at low altitude and then drive to higher elevations can benefit by advancing the ignition timing as you climb to higher altitudes, thus minimizing power loss due to altitude (barometric pressure).

I now have customized MS2Extra "3.2.1.PL" code that has been modified to include Barometric Timing Correction in much the same way as Barometric fuelling correction is applied (see attached screen shot). This code is identical to 3.2.1 but differs only in the following ways:

1) A menu item titled "Barometric Timing Correction" has been added to the "Extended" menu which allows both positive (for going to higher altitudes than tuned at) and negative numbers (for going to lower altitudes than tuned at);
2) Fuelling corrections based on IAT were implemented in 1% increments with 3.2.1 and is now implemented in 0.1% increments with 3.2.1.PL;
3) Fuelling corrections based on Baro were implemented in 1% increments with 3.2.1 and is now implemented in 0.1% increments with 3.2.1.PL. This was done just to smoothen out the baro corrections (which sometimes fluctuated by 1%)

If you are interested in this code you are welcome to PM me.

* Low compression engines (typically less than 8.5:1) seem to be "un"affected by barometric changes (at least those that I worked on).

[TheSilverBuick]

That's pretty cool! I upgraded to MS3 so can't take advantage of your code, but nice to see something for it. I live and drive regularly around 6,500ft - 7,000ft, and every time I leave town it's all down hill from there and sea level isn't out of the question. I've ran into the same issue you mentioned as my compression is 10.3:1.

I've seen the baro read from 103kPa down to 64kPa (above 11,000ft!) so features like this are nice!

[.boB]

I have the same issue. My engine has 13.2:1 static, and 8.3:1 dynamic compression ratio. I live at 6,400' ASL; that's about 81 kPa at rest. When I drive to a lower altitude (especially when it's hot) detonation is a real issue.

The issue is cylinder pressure. Too much pressure and you'll get detonation. There's no easy way to measure pressure inside the combustion chamber. Measuring intake pressure is pretty easy, though. And you can infer cylinder pressure changes from changes in intake pressures. Pressure goes up at low altitudes, and while under heavy load.

My primary ignition map is RPM x Throttle position, and it's fairly conservative. The secondary ignition map is RPM x Manifold pressure, and it's additive. During light throttle cruise when pressure is low, it adds timing. When you mash the throttle or pull a hill, you get gradually less timing. I'v only been as low as 4,000 ASL, and the engine performed flawlessly.

When I get time, I want to play with a single timing map that's RPM x pressure only, and not use a secondary map at all.

[hardline]

MS2 had this feature, but for the life of me I can't find it in MS3. Is this something that was dropped? Reason I would want this is that the burn rate will change depending on air density, and since I am running %Baro for my fuel and spark calculations, it uses the same ignition values independent of altitude.

[wes kiser]

Why not just run speed density for spark?

[hardline]

Cause I am a moron sometimes...... Done.

[wes kiser]

No need to be hard on yourself or so quick to cave. There are guys on this forum that ask for complete firmware re writes for 0% functionality change because they can only envision something working "their" way.

I personally feel that spark advanced based on MAP pressure is the right way to go from a physics modeling standpoint.

[Philip Lochner]

Here probably isn't the right place, but folks running Baro has best chime in as to why baro based retard is needed.

Whilst I understand and respect the desire to "stick with theory" I am of the (blond) opinion that there are things at play in the internal combustion engine which can only be explained with advanced thermodynamics and not "basic theory". But if we can not explain them, it does not mean that we should not deal with them, surely!!

I REMAIN FULLY CONVINCED (more so than before) that both fuelling compensation and timing compensation as a function of barometric changes is an IMPORTANT feature that should remain part and parcel of the MS firmware.

I will revive the thread where we discussed the subject with an essay I wrote for my local 4x4 forum as to why an ECU should have these features.

For now PLEASE, PLEASE, PLEASE do not delete these features. I was wanting to upgrade to more recent releases as I'm seeing AE enrichment happening without supporting TPS or even TPSdot activity, but if these features were to be deleted, it would mean that I'll be stuck on 3.2.1 for the rest of my MS "career". Then I spotted this thread....

The fact that the MS manual still advocates that fuelling should enrich as you climb in elevation confirms the reluctance the devs have in accepting that opposite actually happens - despite various squirters in the US confirming this (not just a blond at the @rse end of Africa). Anyone who does not drive a squirted car in varying altitudes is unlikely to understand (or accept) this issue.

[Philip Lochner]

I happen to notice this subject arising YET AGAIN in http://www.msextra.com/forums/viewtopic ... 30#p356445 and here http://www.msextra.com/forums/viewtopic ... 01&t=49072.

The possibility of Barometric fuelling and timing correction being dropped scares me in a big way and I hope to offer below the best explanation this blond is capable of as to why the MS firmware should offer both fuelling and timing correction as a function of barometric changes for the sake of us poor squirters who want to get the best out of our engines at varying altitudes. Whilst barometric fuelling correction is of importance to all engines (more so normally asiprated engines), Barometric timing correction is of PARTICULAR IMPORTANCE TO HIGH COMPRESSION ENGINES but is still relevant to low compression engines.

What follows was really a veiled "advertorial" (for Megasquirt) I wrote for my local 4x4 forum as to why one should use Megasquirt (with these barometric features) and not a commercial ECU - many (all??) of which DON'T offer these barometric features. I've had to re-write it quite a bit for this forum to "internationalise" it as I wrote it from a very local perspective.

With the info I present below, I hope to equip you with information to help you make an informed decision based on my own EFI experience with Megasquirt-2 which I have been using (since 2006). If you are of the opinion that I am mistaken please air your view. Who knows, I might just learn something.

So here goes:

BASICS 1 : Geography

The coast is at a low elevation of 0 meters AMSL (above mean sea level) and has a barometric air pressure (the pressure of the air we – and cars – breathe at the coast) of about 100kpa. (1bar)
Johannesburg is at 1500 meters AMSL (a higher elevation) but has a lower barometric air pressure of around 85kpa. (0.85bar)
Barometric pressure varies with elevation.


BASICS 2 : In the old days when cars still had distributors…

Some of you, of more “senior standing in society” may remember our dad’s (some dads at least) making sure to take the necessary spanners on a trip to the coast, knowing that

a) either the car will start pinging/knocking/detonating as lower altitudes are reached, or
b) the car will be down on power as lower altitudes are reached.
They used to pull in at a garage and adjust the distributor for less advance (retard) going down to the coast or advance the timing as they came up to Johannesburg.

Why is this??
a) A car tuned for optimal performance at Johannesburg will have more advanced timing than the same car tuned for optimal performance at the coast (assuming the same octane fuel is used).

The higher the compression ratio of this car the higher the likelihood that the engine will start pinging/knocking/detonating at lower altitudes.

If the car was tuned at the coast and it comes up to Johannesburg, power output would decrease, not only due to the lower barometric pressure (“less air”) but also due to the spark timing being too far retarded for the lower barometric pressure as the car rises in altitude.

SOME of that lost power could be recovered by advancing the timing at Johannesburg. I remember 3º as being the generally used (conservative) amount by which to advance or retard. This figure would vary from engine to engine though.

b) Likewise, a car tuned for optimal performance at Johannesburg will have more advanced spark timing than a car tuned for optimal performance at the coast. This means that if this same car is driven to the coast , its timing will become too far advanced at lower altitudes and hence the spark will come before it should for optimal power output and the engine will lose power (maybe not perceptibly so) and may also become heavier on fuel.

We all know that your average car feels significantly more powerful at the coast. Why? Mainly because the barometric pressure there is higher than at Johannesburg, which means more oxygen gets into the engine and more oxygen (combined with more fuel) delivers more power.

Basically driving to the coast is like strapping a turbo to your non-turbo’ed car and having that turbo boost at 0.15bar (because 0.85bar at Johannesburg + 0.15bar increase on the way down to the coast gives you 1.0bar of pressure.)


BASICS 3: Physics

Scenario 1: I have a car at the coast (Baro = 1bar = 100kpa). The throttle is depressed such that the pressure (often referred to as “vacuum”) behind the throttle is 50kpa. The pressure difference across this throttle is 50kpa and this results in a certain air flow across the throttle into the engine with a certain amount of oxygen molecules entering the engine. Let’s call this amount of oxygen molecules X.

Scenario 2: We now take the same car and place it on top of the Alpine mountains where barometric air pressure is 70kpa (0.7bar). We now depress the throttle such that the pressure behind the throttle is again 50kpa. The pressure difference across this throttle is now only 20kpa and this results in a certain air flow across the throttle into the engine with a certain amount of oxygen molecules entering the engine. Let’s call this amount of oxygen molecules Y.

Why should X and Y not be the same? (Because they are not…)

Please note, AND THIS IS VITAL, in scenario 2, the pressure difference across the throttle is now only 20kpa as opposed to the 50kpa in scenario 1. The lower pressure difference in scenario 2 will result in LESS AIR (less oxygen) entering the engine DESPITE the pressure in the engine being the same as in scenario 1.

In other words, an engine with a manifold absolute pressure (MAP) of 50kpa at the coast IS NOT at the same operating point as the same engine with MAP = 50kpa at higher elevations and therefore requires different fuelling at different elevations.

I believe there are other dynamics at play as well relating to the speed with which the air is entering the engine which affects volumetric efficiency (through the momentum of the air) but this is beyond my level of knowledge.

Lastly: Hot air is less dense than cold air. This means that hot air has less oxygen molecules per liter than cooler air. When an engine breathes hot air, it will produce less power than the same engine breathing cooler air (assuming proper fuelling for both situations).


BASICS 4: Lambda sensors

When the ratio of air to fuel is 14.7 parts air for 1 part of fuel (by weight) the ratio is called stoichiometric.

Lambda (λ) is the ratio of air to fuel (by mass) divided by the stoichiometric ratio

λ = AFR / 14.7

There are basically two types of lambda sensors:

Narrow band sensors
Wide band sensors

Narrow band oxygen (NBO) sensors are only good for measuring AFRs (lambda) in the immediate vicinity of 14.7 AFR. The output they produce basically says one of two things:

The AFR is leaner than 14.7;
The AFR is richer than 14.7

A narrow band sensor does NOT say what the AFR actually is. They are thus only good for adjusting fuelling when an AFR of 14.7 (λ=1) is desired. They are no good for measuring or adjusting fuelling when richer (less than 14.7) and leaner (more than 14.7) AFRs are called for.

Please note that an engine does not produce most power at 14.7 (which typically happens at 12 – 13 AFR), nor does it produce the best economy at 14.7 (which could be from 15 – 17 AFR)!

Wide band oxygen (WBO) sensors is a much more effective way of adjusting fuelling as they report the actual AFRs being produced by the engine. A WBO would thus allow an ECU to adjust fuelling regardless of the AFR being called for.

WBOs are more expensive and possibly more sensitive to cold shock (water being splashed on them). They normally need some kind of controller ($$$) to linearise the sensor and are also damaged by heavy metals found in some oils and leaded fuel.


BASICS 5: Closed loop control vs Open loop control

Closed loop control means that the ECU measures the output (e.g the AFR in the exhaust as measured by the WBO) and then adjusts the amount of fuel delivered in order to change the AFR to approach the desired AFR. Likewise it could rely on knock sensors to retard timing in order to reduce/eliminate pinging.

Open loop control means that the ECU is relying only on inputs to determine outputs (the amount of fuel delivered and timing advance etc), but does not measure the results of its outputs to modify its parameters.


So WHAT does all the above have to do with me just wanting my engine to run?

BZZZZT!!! Wrong question!


So WHAT does all the above have to do with me just wanting my engine to run well

BZZZZT!!! Still the wrong question !


So WHAT does all the above have to do with me just wanting my engine to run well under all circumstances ??

Bingo!! That is the right question


Here is the difference between the above:

It runs: The poor bloke who prompted me to write this has an engine “that runs” – but clearly only barely so… He gets 300km to a tank when he should be getting 500km plus. And its as lazy as a…(let me not bring politics into this)

It runs well: He gets 500km plus on a tank and it performs well - but only at Johannesburg. It pings like crazy and it “surges” at partial throttle when he goes to the coast.

It runs well everywhere: He gets 500km plus on a tank and it performs to the best of the engine’s abilities … at all elevations, at all ambient temperatures, WITHOUT requiring adjustment of any kind to be running optimally, with and without a lambda sensor!!


So what do the 5 Basics have to do with me buying an ECU for my car?


Basics 1: Cooler air vs warmer air

•With regard to fuelling:

The ECU should be able to adjust fuelling as a function of inlet air temp. There are only three ways an ECU can do this.

- It has to use a MAF (Mass Air Flow sensor) or
- it must have an inlet air temperature (IAT) sensor, or
- it must adjust fuelling with feedback obtained from a lambda sensor. A MAF would automatically compensate for air temp by virtue of measuring less dense or more dense air.

Fuelling can also be adjusted for with a narrow band lambda sensor but only if the lambda being called for is 14.7 and only if the sensor is working. Few older engines idle at 14.7 (but my ‘05 Volvo do). A wideband sensor would be more appropriate for adjusting fuelling regardless of the AFR being called for.


•With regard to timing:

The reason I include timing adjustment is that I have found high compression engines tend to ping/knock/detonate easier with hotter inlet air (hot summer day vs cold winter day). Low compression engines seem less sensitive for inlet air temp.

There is only 1 way and ECU can adjust timing as a function of inlet air temp and that is by means of an inlet air temp sensor. Neither a MAF nor a Lambda sensor would be good for this requirement.


Basics 2: The old days


The ECU should be able to adjust timing as a function of barometric pressure. There are only two ways of doing this:
1) The ECU should have inputs for knock sensors (these detect pinging and the ECU can then retard timing to compensate) and be able to use them or
2) with a continuous barometric pressure sensor.
However, retarding timing using knock sensors is only good for preventing pinging/knocking/detonation. It is no good for finding or presenting the optimal timing needed by the engine at a particular barometric pressure. Timing adjustment as a function of a continuous barometric sensor is a much better for this purpose.

One could of course set the engine’s timing as for sea level and then spend 50 out of 52 weeks driving a car at Johannesburg which is sub-optimally tuned for Johannesburg….if you want to, OR

Instead of whipping out your 13mm spanner to adjust timing, you could whip out your PC to adjust timing, but my point is that YOU SHOULD NOT HAVE TO!


Basics 3: Physics

The ECU should be able to adjust fuelling as a function of barometric pressure if your car will be experiencing such barometric changes ( and which car won’t?? ). There are only three ways for an ECU to sense such barometric changes:
1) using a MAF (air at higher pressure is more dense and the MAF will sense this) or
2 ) continuous barometric sensor or
3) a Lambda sensor.

Most aftermarket ECUs don’t use MAFs but rely on MAP and I don’t like the Lambda approach because even though narrowband sensors are cheap, they can only be used in a very narrow area around 14.7 AFR and wideband sensors are rather expensive and can be damaged by cold shock and leaded fuel (or substances – oils - that contain heavy metals)


Basics 4: Lambda sensors

An ECU that uses a WBO is by far a more competent ECU than one that uses a NBO – provided the ECU then runs “closed loop”. Personally I prefer an ECU that does NOT rely on either a NBO or a WBO to manage an engine optimally because what happens to the engine when that NBO or WBO fails? Can you still drive the car without issues?


Basics 5: Lambda sensors

A barometric sensor is so much cheaper than a NBO or WBO, is much less likely to fail and thus, if programmed correctly, caters for both barometric fuelling adjustment and timing adjustment and offers better reliability with no worries of cold shock to the oxygen sensor when wading. It is therefore possible to have an ECU run completely “open loop” and STILL provide optimal fuelling and timing without the need for costly knock sensors or oxygen sensors.


CONCLUSION:

In choosing an ECU, you would be wise to choose an ECU as follows:
With regard to fuelling:

* Choose an ECU that can adjust fuelling as a function of inlet air temperature
* Choose an ECU that uses a MAF for fuelling OR
* Choose an ECU that uses a Manifold Absolute Pressure sensor (MAP) for fuelling BUT then it also SHOULD have a continuous barometric sensor AND and IAT sensor and the code to use it for fuelling correction and be programmed accordingly to cater for barometric (changes in altitude) and IAT changes! OR
* Choose an ECU that uses a Manifold Absolute Pressure sensor (MAP) for fuelling BUT then it also SHOULD have a wideband oxygen sensor for it to make adjustments for fuelling and be programmed accordingly!

With regard to timing:

* Choose an ECU that can adjust timing as a function of inlet air temperature (particularly with a high compression engine);
* Choose an ECU that can accept and use knock sensors to retard timing in the event of knock detection (accepting that this is no guarantee for OPTIMAL timing) OR
* Choose an ECU that has a continuous barometric sensor to adjust timing as a function of barometric pressure which can yield more optimal timing under all circumstances IF this barometric correction is programmed correctly!
* A MAP based ECU with an IAT sensor and continuous barometric sensor should thus allow being tuned to the extent that it does not require either a NBO or a WBO after it is tuned in order to run highly optimized at all elevations and all ambient temperatures.

A full time barometric sensor allows an ECU, if programmed correctly, to cater for both barometric fuelling adjustment and timing adjustment. It is therefore possible to have an ECU run completely “open loop” and STILL provide optimal fuelling and timing without the need for costly knock sensors or oxygen sensors.

[Philip Lochner]

Not knowing where to post on MSXtra on a subject that is equally relevant to both MS2 and MS3, I posted my essay here: http://www.msextra.com/forums/viewtopic ... 34&t=52326

[Philip Lochner]

Please have a read of this: http://www.msextra.com/forums/viewtopic ... 34&t=52326

[Philip Lochner]

A timing advance retard table vs Baro - similar to MAT based advance/retard.

I never could understand why, when travelling to the coast (dropping from 1500m / 4500ft) to sea level brought with it the onset of pinging, even at partial throttle despite stating out with timing map with 0 pinging. I thought that having an absolute pressure based timing map would eliminate altitude related pinging and that I would only have to tune the high kpa bins once I get to sea level. It is important to understand that I tuned the engine at 1500m where I live and can not tune bins higher than 85kpa because that is WOT at this altitude.

When you have 60kpa MAP at higher altitudes, is it not the same as having 60kpa MAP at sea level? After doing the sums as reported here (http://www.msextra.com/viewtopic.php?f= ... 20#p203139 - and assuming they are indeed correct) I now understand that 60kpa MAP at higher altitudes, may not be the same as having 60kpa at sea level by virtue of different cooling drops over the TB causing different air densities, which could also be contributing to different AFRs and ultimately different timing requirements. Combine this with the effect of changing back pressure at the exhaust....

Fortunately I realised that changing my trigger offset would move the entire table up or down and this is how I sorted the problem on the way to the coast - PC on the wife's lap

Further to the above posting, I have posted this (http://www.msextra.com/forums/viewtopic ... 34&t=52326) hoping to present a case as to why BOTH fuelling AND TIMING barometric correction is needed and should be offered by all FW versions of MS.

There is talk of timing correction in particular being dropped and I'm having a hard time explaining to the devs (and convincing them) why this feature is definitely needed. I was so convinced (and still am) that I need this feature that I PAID a code writer to add this feature for me!! I was having detonation issues on my high compression Rover V8 even when dropping from 1500meters to 1000m.

[Philip Lochner]

Further to the above, please have a read of this: http://www.msextra.com/forums/viewtopic ... 34&t=52326

[Philip Lochner]

For timing, altitude doesn't really matter.

This has not been my experience at all. I believe that barometric timing correction is needed for both low and high compression normally aspirated engines, but more so for high compression engines. Please have a read of this: http://www.msextra.com/forums/viewtopic ... 34&t=52326

[TheSilverBuick]

I run standard speed density for the timing table and BaroCorrection for the fuel table so timing is tied directly to air pressure. A few years ago before being able to separate them I ran into pinging issues when I'd drive off this mountain and would manually back the timing off, now it handles itself.

[Philip Lochner]

A few years ago before being able to separate them I ran into pinging issues when I'd drive off this mountain and would manually back the timing off, now it handles itself.

The worry is that now the devs are talking about dropping the baro based timing correction feature!! What to do???

[jsmcortina]

Talking about?
Anything we were planning on doing has been done. Have you tried the current release 3.3.1 ?

James

[TheSilverBuick]

As in a separate additive or subtractive timing curve like IS need for fuel? If so that really isn't needed in my experience of driving from ~78kpa to ~103kpa and using standard speed density on the timing table. If they are talking about going back to only having one option of Speed Density or BaroCorr that switches both fuel and spark tables together, that would be a step backwards IMO.

[Philip Lochner]

As in a separate additive or subtractive timing curve like IS need for fuel? If so that really isn't needed in my experience of driving from ~78kpa to ~103kpa and using standard speed density on the timing table.

Whether you "need" it or not is a function of:
1) whether you have a high or low compression engine;
2) how well you want the engine to be optimised at 78kpa and at 100kp and everywhere in between.

Low compression engines will not complain either way but that does not mean that its running optimal at an elevation other than what it was tuned at.

Also, if you tuned it at sea level (high and low compression) and you drive to higher elevations it will again not complain (ping) but again, it will not be running optimised.

[Philip Lochner]

Have you tried the current release 3.3.1 ?

James

Nope, the download sites only seem to have 3.3.0 (which does not seem to work = error 404)

Where do I get 3.3.1? Keen to try it. Search does not search on 3.3.1...