Barometric compensation strategies

[Zmanco]

MS II has a correction factor for altitude that effectively increases the VE (richens the mixture) as the car goes up to higher elevations. The theory is that at higher elevations there will be less exhaust back pressure resulting in improved (higher) volumetric efficiency.

 

I live in Colorado and recently took a driving trip in the Z to California. My MS setup has the second MAP sensor so the system can monitor and adjust for changing barometric pressure on the fly. As I descended from the Lake Tahoe area (6000 feet) on highway 80 and got below 3000 feet, I noticed that my mixtures were going pretty lean. By sea level I was so lean that the O2 correction (limited to +/- 10%) was pegged and not able to fix it. As a stop gap I used the Barometric Correction of MS II Extra to add 20% at sea level (100 kPa) and that got me in the ball park.

 

While in the Bay area I read up on the built in barometric correction algorithm and found how I could change it. Keep in mind that the default algorithm was in effect leaning things out as I went down to sea level, my experience said I needed to go the other way and increase the VE values.

 

I bring this up to the group here because I'm wondering how other non-MS EFI systems handle the issue of correcting VE as altitude changes? So far I have been quite impressed with the design of the MS systems - obviously there are some pretty smart people developing them. But this issue, at least from my one data point, suggests they were completely going in the wrong direction on barometric correction.

 

So to those of you with experience with other EFI systems, what algorithms are used to correct for VE as altitude changes?

 

PS: I averaged 26.9 mpg for the trip typically running between 75 and 85 mph. Not bad for a 4.11 diff!

[1 fast z]

When you RISE in elevation, you want the engine to LEAN out, NOT richen. It is not because of Back pressure in the EX but rather Atsmospheric Pressure differential. If your engine is tuned for 6000 Feet, and you drop down to sea level, it is basicly like you have a supercharger that puts 3 PSI to your engine, which means you need to richen it up when you drop in elevation.

[Zmanco]

So what you're saying is that at a given combination of MAP and RPM, the engine's volumetric efficiency is greater at lower elevations than at high. That's what my intuition says too and my experience bears out.

 

But here's what bothers me about this: the higher barometric pressure at sea level only exists upstream of the throttle plate. It seems to me that at sea level I wouldn't have to open the throttle up as much for a given MAP for the reason you just gave (supercharger effect of extra barometric pressure). But from the cylinder's perspective, why would it be more volumetrically efficient at any given MAP and RPM at sea level? I'm not seeing any good reason for it.

 

BTW, I'm just trying to understand what's going on here - not trying to argue that VE doesn't go up at lower elevations.

 

Also, does anyone know how this is handled by non-MS EFI systems? Is the correction a linear function as megasquirt models it?

[cygnusx1]

You also have less air density at higher elevations. Less O2 in the air per cubic foot.

 

It's also relative....at sea level you are "intaking" and "exhausting" into the same atmospheric pressure. this is ALSO the case at 10,000 feet. It's a wash.

 

What counts is the ABSOLUTE pressure. PSIA. Pressure relative to a perfect vacuum.

 

 

At lower elevations you have more O2 and more pressure relative to the perfect vacuum of 0.

 

 

Having boost is like having your intake below sea level and having your exhaust pipe above sea level...sort of . It's cheating!

 

I also have the dual BAR sensor on my MSII. Unfortunately, I don't get to maximize it's use because I usually drive around between 20 and 1200 feet. Interesting thread!

[Zmanco]

I've been discussing this on msextra.com and it was recommended by one of the developers to use % Baro for the fuel algorithm instead of speed/density. I've since made the switch and find that I now have the full range of the AFR table as well as VE table to use (before I only used about 80% of the table). Also, in theory, as I go down (or up) the richening that we all add to the VE table as we approach WOT will scale according to the ambient barometer setting.

 

So far I only have one test run (weather needs to clear up before I go out again) but at this point I would recommend this approach for anyone who drives across a variety of altitudes.

 

This option is also available for Ignition. For my NA engine I've decided to leave that controlled by MAP since I found I had to back timing off more at sea level than at altitude all else being equal. For a turbo motor I'm not sure which would be the best approach.

 

Still unanswered is how the other FI systems handle correcting for changing altitude (my original question in this thread). In the next few weeks I hope to take a drive that will see me going up above 10k feet which will give me some good data to see if the change I made to Barometric Compensation helped.

[cygnusx1]

I use two point baro correction in MSII with dual map sensors and the default values. Here is an excerpt from the Megamanual which has me confused.

 

2-Point Barometric Correction - if you have selected two-point corrections, you can set the baro corrections here (if you have selected 'tables. baro correction is set under 'Tables/Barometric Correction'). Correction for barometric effects is performed using the linear function:

 

Barometric_correction = At_total_vacuum + (Rate * barometer) / 100

 

At Total Vacuum (%): 'At total vacuum' contains the total correction at a barometer reading of 0 kPa.

Rate (%): 'Rate' contains the percentage per 100 kPa to scale the barometer value.

Using the default values of 147 and -47, we see that for a barometer of 100 kPa, we have 100% correction (i.e., no correction). Barometric_correction = 147 + (-47*100) / 100 = 100%. Note that the values of 47 and 147 are not coincidental - they cancel at 100 kPa to give no baro correction, as you would expect.

 

 

This confuses my little mind. I have been out of school for a long time but I didn't think it was THAT long.

 

147 + (-47*100) / 100 = 100%

 

How do they get 100 out of that?

[m4xwellmurd3r]

multiplication, then division then subtraction.

-47*100 is -4700

-4700/100 is -47

147+(-47)=100

 

simple order of operations stuff from pre-algebra (elementary stuff basically...)

[Zmanco]

This confuses my little mind. I have been out of school for a long time but I didn't think it was THAT long.

 

147 + (-47*100) / 100 = 100%

 

How do they get 100 out of that?

The equation is correct - perhaps you read the "*" as "+"? Make sure you solve within the parenthesis first

 

In any event, the main issue I see is that the -47 value determines the slope of the function. In practical terms, that means that the default settings increase VE as barometric pressure is reduced. My own experience says this is wrong and that VE should be reduced as barometric pressure is reduced (opposite slope).

 

As I think I said earlier here (or was it on msextra.com?) this doesn't appear to affect very many people so I won't obsess over it, but I did want to make others aware that if they do drive across a wide variety of altitudes, my own experience suggests that the default correction in MS2/E is incorrect and needs to be changed significantly.

[cygnusx1]

Yup been out of school too long and too many exhaust fumes. Forgot that the division operation came second after the parenthesis. Time to go back and review some basic math! See what computers and calculators can do to your brain?

 

http://members.aol.com/algbrahlp/eqcalc.htm

 

This online calculator gets it wrong too.

 

 

 

So if I drive to the top of a mountain and get a 89kpa barometer reading, my VE table will multiply by:

 

147 + (-47*89) / 100 = 105.2%

 

So, richer at higher elevations? That seems wrong. It should go leaner. Let me see if I can work on "new, correct, defaults".

[Zmanco]

It's the exhaust fumes

[cygnusx1]

Here is a good read:

 

http://web.archive.org/web/20051214181804/http://www.saabclub.com/242/altitude.htm#physics

[cygnusx1]

Using the default math built into MSII for 2-Point Bar Correction, I feel it would be more appropriate to use the following numbers.

 

At total Vacuum %: 80

Rate %: 20

 

EQUATION IN MSII

(at total vacuum%) + (rate% * 100)/100 = baro correction

 

 

 

EXAMPLES:

At sea leavel:

80 + (20*101kpa)/100 = 100.2% correction (normal)

 

 

At 10,000feet:

80 + (20*66kpa)/100 = 93.2% correction (leaner)

 

 

For less correction, use a higher (at total vacuum) and lower (rate).

 

For more correction, use a lower (at total vacuum) and higher (rate).

 

Both numbers need to add up to 100

 

I hope I redeemed myself for forgetting the operations order of algebra for a minute.

 

 

The mystery for me is that if you operate in closed loop, the O2 sensor should maintain your AFR map no matter what the altitude without additional corrections? No?

[Zmanco]

Based on my experience, there was an 8-10% difference between sea level and 6k feet so I'd suggest 57 / 43. This will effectively not change anything at sea level and give you an 8.6% reduction at 80 kPa barometric pressure which is typically what I see at 6k feet.

 

Here's a simple spead sheet I built to help work this out. The middle scenario is unique for me in that I had already tuned my VE tables for a barometric pressure of 80 kPa and didn't want to retune them, hence the different offset value (Total Vacuum in MS2/E).

 

The mystery for me is that if you operate in closed loop, the O2 sensor should maintain your AFR map no matter what the altitude without additional corrections? No?

I think in theory that's correct, but if the table isn't perfect (and it never is) then it's possible that even with +/- 15% control there may not be enough range to fully correct. Also, I found that if the VE map is too far off, then the O2 correction is constantly changing and since it's always missing and then over correcting it feels like a lean/surge at steady light throttle such as when cruising.

Barometric correction calculations.zip

[cygnusx1]

Daniel, are you running a turbo? I think the situation is different between turbo and N/A.

 

 

EDIT: You are running an N/A so I would tend to err on the richer side with a turbo at altitude.

[Zmanco]

No Dave, I'm NA. I too think it would be different. Isn't the main resistance to exhaust flow on a turbo from the turbo itself and and its plumbing? Seems like that would be far more significant than the backpressure of a low restriction NA exhaust.

[cygnusx1]

Yes, the having an idle turbo on your exhaust manifold is worse than having a supertrap muffler.

 

It's pretty cool that you have the opportunity to watch how elevation effects your AFR's. I think that is info that is hard to come by for our motors. It would be interesting if you could datalog it. I'm sure the Pike's Peak race teams are well versed in this stuff!

[Zmanco]

I'm hoping to get a local drive together in the next few weeks that will go up into the high country, probably at least 8500 feet. Of course, we'll have to hope that weather is clear that day

 

I'll post the datalog here and on msextra.com.

[RTz]

Still unanswered is how the other FI systems handle correcting for changing altitude (my original question in this thread).

 

Its commonly done through a dedicated MAP sensor and interpreted through a user definable 2D map to trim the base map. Not ground breaking stuff.