Barometric Correction

[rogerguzzi]

Hello My learned Friends

                                            I am trying to get the OLD grey matter working again on this subject which still confuses me.

So here is my question

I can understand why a TR6 type fuel injection goes rich up mountains(thin and lower pressure air but the same fuel at any given throttle opening)

So why does Spitty on Micro Squirt go weak up mountains assuming no authority is given to correct it? 

I know I have lambda sensor but if no authority given it is just a fancy gauge!

I have a second sensor to sense atmospheric pressure change and I have set the curve as attached plus I give the system 6% authority

When we were in the mountains she still is little bit lean and so gets a bit hotter so I would prefer it to be on the rich side and perhaps run a bit cooler.

I know now you are all going to say don't be so thick this is what is happening but I think my poor old 74 year old brain cells need a reboot as I seem to be going around in circles and I have read all sorts of forum posts but they just above my pay scale and they use big words etc

Confused Roger

ps plus I do not understand the correction table as the pressure drops so does the correction % I would have thought it would add to add fuel?

pps the table numbers are what one of the forums suggested and perhaps tweaked a little?

[JohnD]

Roger,

Pi goes rich because it compares atmospheric pressure with that in the intake.  It assumes that the lower the difference, the more open are the throttles.     As atmospheric falls, actual intake pressure doesn't, so the difference is less and Pi allows more fuel in.  So in fact, Pi gives MORE fuel for the SAME throttle opening.  But you know that!

     Lucas did produce a mountain version of the M/unit, it has an atlmospheric sensor under the cam housing, that allows it to compensate, but they are extremely rare.

 

I don't know enough about MicroSquirt to be definitive, but I expect a smiliar argument applies.     But the table you show does not indicate a constant correction.    For instance at 65 kPa, the corrected is 74, which is 14% higher, while 85 is corrected to 92, an 8% correction, and at 105 the correction is MINUS 4%!   But you know that, too!    Is this "6%" correction a constant, or can you apply a table, like the one you show?      

Thinking aloud.    While pressure falls with altitude, oxygen content is constant.     But each litre of air will have less of all the gases in it. At 3000ft (900metres and approx 70kPa) the effective oxygen contant is only 19% compared to the normal, sea level content of 21%.      So the ideal fuelling per litre of air for stoichiometric mixture should fall, but only by 3%.    

There's something wrong with the correction factor.  Sorry not to be more help!

John

Edited December 15, 2019 by JohnD

[zetecspit]

Roger, I have no idea about the microsquirt. BUT when I plugged a map sensor into my ECU and added a calibration (I used the same sensor as you) it used the output as part of the fuelling calcs, and I didn't need to fill a table out at all. So I wonder, just wonder, if you set your table to all 100% and test the sensor by sucking on it, what happens then? It is possible that you are doing a double correction?

 

(I didn't get this at all when I fitted it. I expected to have to fill a fuelling table in, and had to call the ECU manufacturer to confirm all was well....)

[Nick Jones]

1 hour ago, JohnD said:

There's something wrong with the correction factor.

That's it in a nutshell.  As it's going lean it's over-correcting so your curve is too steep.

I was was confused about why your curve should be "wrong" because as I remembered, when I first did this on the Vitesse (when I fitted the MS2 ECU in 2009) the firmware had the curve already in it based on theoretic values and I left it untouched.  In fact, I've never messed with it as subjectively it works fine.  I say subjectively as I'm purely judging it on a "seat of the pants" basis - I've never been able to monitor the actual AFRs when driving in the mountains as I didn't have an AFR display until this year and that stopped working just as I got to the mountains.....   It's been tested up to 2,802m.  It has also been established that above about 1500m, PI cars that would leave us for dead at sea level become prey.

Having just googled "barometric correction, Megasquirt" I find that  for many years the firmware has been supplied with the correction curve supplied "flat" (ie no correction) for user setting after (apparently) some robust debate about the validity of the original curve.  I should be able to retrieve the original curve from my ECU if it helps......

 

[rogerguzzi]

Hello All

                Thank's for the replies but I think I am a bit thick as it still confuses me! (but perhaps not quite as much?)

I found this statement which helps me understand the going lean bit!

(The reduction in exhaust back pressure as altitude increases also increases the volumetric efficiency of the engine, so with no barometric correction, you should experience leaner EGO readings. If you are experiencing richer readings then it would appear that your correction is too aggressive.)

So a bit more head and bum scratching needed as I am still confused by the curve and what it is doing?

Does it mean I am supposed to set the AFR ratio at 14.7 at X thousand meters and the curve will reduce it at sea level by altering the pulse width of the injectors?

Roger

John

Is this "6%" correction a constant, or can you apply a table, like the one you show?      

it is what % the system can alter the fuel to try and achieve what is set in the AFR table.

So in theory if the AFR table was 14.7 it cope with rich at 13.8 and lean at 15.5 so It can not go to far which would damage the engine  plus there is this table!

So you can allow for hot air? 

You can see now why the car manufacturers spend millions on EFI alone and ours is a simple ish system 

 

 

ps I wonder if I could fit a manual choke lever(joke)

 

[JohnD]

Did you see this, Roger?    https://adaptronicecu.com/blogs/modular-instructional-videos/fuel-model-in-modular-ecus

Or did the equation put you off?  It does me!

John

[Nick Jones]

28 minutes ago, rogerguzzi said:

The reduction in exhaust back pressure as altitude increases also increases the volumetric efficiency of the engine, so with no barometric correction, you should experience leaner EGO readings. If you are experiencing richer readings then it would appear that your correction is too aggressive.

Delete this from your brain.  It is complete horseshit and the opposite of the truth.  I found the same thread and there is two pages of complete drivel that should be erased from the internet. 

Also, the MAT air density is a separate thing, although also involved in the fuelling calculation.  Leave it out of the discussion.

As you go up from sea level atmospheric pressure falls as air density decreases.  This upsets the fuel calculation because there is less air available for given map reading so you end up with a rich mixture, going ever richer as you go upwards.  So the Baro correction curve is there to introduce a factor, determined by the pressure reading of the second pressure sensor to offset this.

We discussed this before in 2017....

 

[zetecspit]

Am I being thick in thinking that at 100kPa you want 100% fuel, but at (say) 70kPa you want 70%? being as it is 70% of the air pressure, so contains 70% as much oxygen as at 100kPa? Or is there something else at play?

(and I did chase the company up about lack of MAp sensor. Turns out the ecu was originally developed for a Le Mans racer, so baro correction was not required. And their target audience in the UK seems to be people who like a good hoon about, but not so likely to go Alpine touring. Their now flagship model does have a map sensor, along with inbuilt wideband. And a pricetag to reflect)

[rogerguzzi]

Hello All

               So Nick you are telling me that it goes rich the same as TR6,s?

So the correction table is a  % of the VE table at what ever point it is at and Barometric pressure and so leans it off?

So if I am seeing lean AFR the curve is to much below 100%?

This is in the setting up manual

 ITB uses a combination of MAP and TPS This mode was created specifically for naturally aspirated engines running with independent throttle bodies. It combines alpha-n (at high engine loads) with speed density (at low engine loads), using the load calculation that makes the most sense at each RPM. For example, most ITB set-ups do not have good vacuum at idle or low RPM, and slightly touching the throttle makes them lose all vacuum, but at higher RPM start to respond more like a traditional single throttle body engine. This mode allows the use of speed density set-ups at low engine loads and switches to alpha-n at high loads, with an adjustable switchpoint curve over RPM. • %baro uses a combination of MAP and a barometer reading. This setting is similar to the Speed Density setting in that the MAP sensor is used to determine load. However, instead of directly using the manifold pressure, the manifold pressure is divided by barometric pressure to give a percentage of barometric pressure. This setting can be useful for those who regularly drive at high altitudes. It ensures that regardless of barometric pressure, all table lookups operate over 0-100%. For example, if barometric pressure is 80kPa, and the engine is operating at 50kPa, the actual value used for table lookups is 50kPa/80kPa or 62.5%. All have pros and cons. Radical engines may need to use a combination of algorithms. Many engines will get good results with Speed Density or MAF. 

Roger

ps still confused but have vague idea how it works now just need a Mountain to try it out? or perhaps a vacuum pump and look at the settings? 

[Nick Jones]

2 minutes ago, rogerguzzi said:

So Nick you are telling me that it goes rich the same as TR6,s

If you turn off all altitude compensation it will go rich as you go up, though not as rich as a PI as I think there's a bit of a self-correct effect built in as Clive is reaching at.  In fact there is at least one ECU manufacturer out there who insists that no correct is needed when using MAP/VE based mapping though I think it's pretty plain this is not the case in the real world.  There are several factors that affect it physically and several more in your ECU (which vary with firmware and what strategy you are using - things like speed density/alpha-N or combination and "multiply map", or the "%baro") which means that you can only really get the "perfect result" by actual corrections at altitude.  You can't be all that far off - just need to try and find time to have a bit of a tweak of your correction curve next time you go and play in the big hills.  There's an excuse for you to get planning the next trip - as if you needed one .

I've not tried to look at this in detail before as when I added my barometric correction it just worked. Seems I was very lucky as it's much more complex than I thought it was and turns out I don't really understand it properly either......

My current firmware doesn't have any of the options mentioned above because it's old........  I could update it but it works pretty well already and I don't want to screw it up!

This seems to be quite a useful link

http://www.msextra.com/forums/viewtopic.php?f=134&t=54593

Incidentally, systems that use mass air flow measurement (flapper or hot-wire) and this includes carbs are more resilient but even they need correction when you stray far enough from the base tuning altitude.

 

[rogerguzzi]

Hello Nick

                 It looks like I have opened a can of worms!

I think for now I may just reduce the correction a bit as a bit rich is better than lean?

I have looked up pressures at different heights

10000ft = 70kpa

9000      = 72.5

8000       = 75

7000       = 78

6000       = 81

5000       = 84

4000       = 87.5

3000      = 90.5

Plus it is purely academic for most of the time but you know me I get a bee in my bonnet! (and its a F*****g long winter already) and I can plan this in the warm and not the cold garage!

Roger

Now what about a cam sensor for Sequential injection? would it make any difference or just occupy my time in doing it?

[rogerguzzi]

Hello Nick

                 On re reading this section I am thinking now perhaps should set the Barometric correction at 100% as I use ITB mode so perhaps it is being corrected twice?

I am not sure what I was using when we toured the Pyrenees as it was early days and I have made so many changes since then as I was learning about EFI and still am (but mainly forgetting more!)

What do you think? or perhaps halve the values of the correction table?           

On 12/15/2019 at 7:59 PM, rogerguzzi said:

ITB uses a combination of MAP and TPS This mode was created specifically for naturally aspirated engines running with independent throttle bodies. It combines alpha-n (at high engine loads) with speed density (at low engine loads), using the load calculation that makes the most sense at each RPM. For example, most ITB set-ups do not have good vacuum at idle or low RPM, and slightly touching the throttle makes them lose all vacuum, but at higher RPM start to respond more like a traditional single throttle body engine. This mode allows the use of speed density set-ups at low engine loads and switches to alpha-n at high loads, with an adjustable switchpoint curve over RPM. • %baro uses a combination of MAP and a barometer reading. This setting is similar to the Speed Density setting in that the MAP sensor is used to determine load. However, instead of directly using the manifold pressure, the manifold pressure is divided by barometric pressure to give a percentage of barometric pressure. This setting can be useful for those who regularly drive at high altitudes. It ensures that regardless of barometric pressure, all table lookups operate over 0-100%. For example, if barometric pressure is 80kPa, and the engine is operating at 50kPa, the actual value used for table lookups is 50kPa/80kPa or 62.5%. All have pros and cons. Radical engines may need to use a combination of algorithms. Many engines will get good results with Speed Density or MAF. 

Roger

[rogerguzzi]

Hello All

               I hope I am not boring you all with this thread? but I would like to know I have got the right idea

So my latest thinking is in ITB mode providing you have a Barometric pressure sensor it works it out for you!

So I am up a 5,000 ft mountain kpa is about 85 and assuming we are using 3,000 rpm and about 60kpa map this gives 30% reduction in fuel!(I think?)

Plus the Barometric table is giving another 9% = 39%  

So at that point the AFR table is 14.0 - 39% = 19.46 AFR

Now the system has 6% authority so 19.46 - 6% = 18.29 AFR

But this would be equivalent to sea level I have no idea how much less fuel is needed at 5,000ft I assume the 30% less?

But If I look at a different way and say the system knows what is doing but I have upset it by taking another 8% off  which would give me 14 - 8% = 15.1AFR(but still to low for an engine that maybe working hard!)

This is all  assuming my tables are somewhere near to start with!

So what do we think or is all this rubbish?

Roger

ps  pity I do not have a nice high mountain close by as I could do it by trial and error?

 

 

[zetecspit]

I am totally confused by your numbers!

But for simpllcity, the basics is that at 80kPa you have 80% of the oxygen at sea level. So IF (big if) it is as simple as that, you need 80% of the fuel. It is no doubt more complex, but my ECU uses a VE table, not sure about the microsquirt? And until I added a MAP sensor for baro I didn't use one at all.

So I would be inclined to use a linear correction table. And a bit of authority to the wideband. 

 

(your figures go teh wrong way?? at altitude you have calculated a very weak AFR.? or am I missing your point?)

[rogerguzzi]

Hello Zetecspit

                           I did it that way as I thought if the system reduced fuel by these % that is what you would get?

But am probably completly wrong but I have yet to find an explanation I can understand!

Roger

[rogerguzzi]

Just now, rogerguzzi said:

Hello Zetecspit

                           I did it that way as I thought if the system reduced fuel by these % that is what you would get?

But am probably completly wrong but I have yet to find an explanation I can understand!

Ps microsquirt uses a VE table as well but I understand that is just a % of 100% injector rate that you have told it(I think?)

Roger

Roger

 

[zetecspit]

I don't think VE tables are just a %. There is all sorts going on, and I have seen numbers over 100.....

 

With your calcs, I think you are working out the AFR if you were still at 100kPa? but you are not! (or shouldn't be) 

[rogerguzzi]

Hello Zetecspit

                            I did the calculations as of the Mega/Micro squirt information as I understood it!

I read somewhere you can have higher numbers than 100 in VE tables I think it said it just meant they were open longer!(not sure were I read that) but that may have been for Sequential injection with bigger injectors or something like that when you have to try and get all the fuel in while the valve is open(gone off that idea as it gives no advantage or not much! a bit better mpg and better idling!)

Roger

[rogerguzzi]

On 12/15/2019 at 7:59 PM, rogerguzzi said:

• %baro uses a combination of MAP and a barometer reading. This setting is similar to the Speed Density setting in that the MAP sensor is used to determine load. However, instead of directly using the manifold pressure, the manifold pressure is divided by barometric pressure to give a percentage of barometric pressure. This setting can be useful for those who regularly drive at high altitudes. It ensures that regardless of barometric pressure, all table lookups operate over 0-100%. For example, if barometric pressure is 80kPa, and the engine is operating at 50kPa, the actual value used for table lookups is 50kPa/80kPa or 62.5%. All have pros and cons. Radical engines may need to use a combination of algorithms. Many engines will get good results with Speed Density or MAF. 

Hello All

               I am still thinking about this and realize I had read a section wrong as it all runs together!

So it looks now as if I should either change to %baro mode and let it do its own thing!

Or leave it in ITB mode and adjust my correction curve as it is obviously over correcting to the lean side!

I think?

I had a play today by finding out what the barometric pressure was in this area and slightly recalibrating the map and Barometric sensors so they showed current Barometric pressure.

So I think now they are accurate at sea level (or very close?)

Roger

[Nick Jones]

I agree..... I think...... as I'm a bit confused as to where you are now. 

Looks like %Baro includes the baro sensor reading in the basic fuel calculation so the later correction is not needed (don't know whether you are still able to put numbers in the correction table for further tweaking? but I'd definitely have them all at zero to begin with), so if this is what you are using you should maybe put them to zero?

If using ITB mode then I think you just need to back off your numbers a bit as you are over correcting at present (said this earlier) but you are going to need to go uphill a bit to figure out how much.  Doesn't sound like you are are that far out though - would you even know (from the feel of the engine) if you didn't have the AFR display?

Nick

[rogerguzzi]

Hello Nick

                  Spitty is off the road for now (We live down a bridle track that gets muddy !)

All of this this is not that important but I can play in the warm!

I run ITB mode now which seems ok!

Plus next year not sure how high mountains we will go up and it was not that lean but(you know me?) I like to mess!

The thinking at the moment is to just lower the Barometric correction table as ITB seems to work ok!

Were as changing to %baro opens up a new can of worms!

Still it gives some of us things to think about?(How does that little box do all this stuff?)

We have booked at trip to India for March this the details(not cheap but what the hell!)

Roger

ps still going to Spain in Spitty!

india.pdf

[zetecspit]

6 hours ago, rogerguzzi said:

Hello Nick

                  Spitty is off the road for now (We live down a bridle track that gets muddy !)

All of this this is not that important but I can play in the warm!

I run ITB mode now which seems ok!

Plus next year not sure how high mountains we will go up and it was not that lean but(you know me?) I like to mess!

The thinking at the moment is to just lower the Barometric correction table as ITB seems to work ok!

Were as changing to %baro opens up a new can of worms!

Still it gives some of us things to think about?(How does that little box do all this stuff?)

We have booked at trip to India for March this the details(not cheap but what the hell!)

Roger

ps still going to Spain in Spitty!

india.pdf 494.28 kB · 1 download

Are you saying my earlier post was right in that it was automatically correcting for baro, plus you were also doing a later correction? That would be a first for me!

BUT if that was the case, and now in the ITB mode it doesn't automatically correct (are you sure? do check by giving a suck on the sensor with your correction table all set at zero. or 100%, whatever is no correction) then you will need the second correction, which may be correct? as before it was doing that as an unwanted extra "correction" 

Bah, too many uses of the word correction.

[rogerguzzi]

11 hours ago, zetecspit said:

Are you saying my earlier post was right in that it was automatically correcting for baro, plus you were also doing a later correction? That would be a first for me!

BUT if that was the case, and now in the ITB mode it doesn't automatically correct (are you sure? do check by giving a suck on the sensor with your correction table all set at zero. or 100%, whatever is no correction) then you will need the second correction, which may be correct? as before it was doing that as an unwanted extra "correction" 

Bah, too many uses of the word correction.

Hello Clive

                  No I think I am just over correcting in the Barometric correction tables.

I think my best way is to set in %Baro mode and put the Barometric correction tables at 100% and let it work it out its self(it is probably brighter than me!)

After re reading it all several times I realise most of the time I am in Speed Density Mode but that does not correct for altitude.

Plus with 6% authority it should sort its self out if the AFR tables are some were near correct?

Roger

ps there is no mention of Barometric correction in ITB mode unless you have set the Barometric correction tables other than 100%

pps I wonder if I could build a huge pressure chamber and lower the pressure????????

[thebrookster]

2 hours ago, rogerguzzi said:

pps I wonder if I could build a huge pressure chamber and lower the pressure????????

Doable!!

Guy Martin did this in one of his programs, I think it was the one where he did a Hill-Climb in the USA at ridiculous altitude, and needed to do the altitude training ahead of it.

He managed to rig a spare room with an exercise bike in it, because he didn't want to pay to go to a proper centre!

Phil

[mpbarrett]

you could just reduce the pressure  seen by the barometric sensor to make if think the air pressure is reduce and then see what it does to the mixture....

mike