Barometric Pressure Reading with only a single MAP sensor

[freudie1]

Apologies first off if this is the wrong forum for this (can't find anything really applicable).

I'm working with a GM/Delphi ECU from a Harley v-rod (rumor has it it's from a marine ECU that Delphi slightly modified for HD).

At any rate, my concern is around barometric pressure reading.

The ECU has a single pressure sensor, the MAP sensor. Couple that with the fact that the ECU is really a hybrid of alpha-n for VE and speed density for AFR and spark.

Now....I believe the ECU reads the output voltage of the map sensor with key on engine off to take a base reading of the ambient air pressure.

That much I can handle (did I mention I'm installing a 2 bar sensor in place of the 1 bar? Yes I know I need to scale the voltage at KEO such that the 2 bar sensor will show approximately 4.75-ish voltage to match what the 1 bar would in this condition).

What I want (read: NEED) to know is if the ECU is reading baro (recall with ONLY one map sensor for pressure readings) at any other condition. I have been told the following (of which I have no idea who knows the truth...):

It reads only at key on engine off.

It read at idle (which I believe would be 1/3 the voltage of what really ambient is...)

It reads "all the time".

It takes a reading when you got WOT.


Dilemma: First, anyone that can share some insight into this I appreciate it. Second, IF it's reading at WOT (or really any other time besides key on engine off) I am at a loss for how I can make this work....supercharging the bike that is. FYI: The ECU only has a 100 KPa OS such that I indeed plan on scaling the tables such that 60 KPa and up will equal boost for AFR/timing/etc values.

For the MAP sensor offset at key on engine off I was planning on using a Split Second map voltage modifier to add (or subtract if needed) an offset. What I can't account for is the engine on scenarios (IF they exist). Unfortunately I don't have/know a PID that states "Baro pressure" for this ECU (only MAP value and voltage)....


Thoughts?

[aaronc7]

I have a 2003 C5 Z06 running a gen3 512K pcm and I can tell you that:

-it reads at initial key on engine off
-it reads at WOT as defined by a set of tables (ie below x RPM and above y TPS, etc.) Basically a lower rpm WOT area where in theory the intake wont pose a real restriction and MAP will = baro
-it reads "all the time" and updates baro IF current MAP > current baro
-max baro is right around 104 kPa

I have efi live and it has a baro PID and after any WOT activity it pegs at 104 kpa.

this affects the injector IFR tables, as they are scaled to "manifold vacuum" ie. BARO - MAP. edit: and the injector deadtime/offset table

most people who run boost have a boost referenced FPR, so that IFR table is "flat" anyways so that is no factor

I have not found any other tables affected by baro value

Not sure how this affects your setup but I hope this helps

[ChopperDoc]

Most PCM's will only take the reading before engine start. Barometric pressure at zero Mean Sea Level (MSL) and 288K (15* C) is 101.3 KPA, or 29.92 inches mercury, or 14.7 PSI. Once the engine is started and from there on, the MAP is no longer a "baro" reference. It is absolute, meaning there is no reference. Manifold "Absolute" Pressure means just that. It reads exactly what is in the intake at any given time.

The difference between baro and MAP, as noted above is referred to as "vacuum," but in reality this is not an accurate or correct statement from a physics standpoint. It's just easier to say vacuum than to say "differential pressure in the manifold" because that is the reality. Engines do not "suck" air in. No, in reality the atmosphere pushes itself in through the intake because it is at higher pressure than what is in the manifold, aka "differential pressure" or PSID. That's how it really works, for an N/A car, as well as boosted cars when not in boost.

An N/A car cannot exceed the baro at anytime, otherwise the airflow would cease, because it wouldn't be "pushed" in. In transient periods it is possible to nearly match it with air being "rammed" into the intake, and from the turbulence inside. However the flow through the engine benefits from Bernoulli's Principle, as the faster it moves, the less pressure it creates. This effect I would assume is the reason you can "ram in" more than the baro for very short periods under WOT and continue to have air flow in (because it's moving really freaking fast), but it's just not possible for the MAP to match the baro for long periods even at WOT (N/A only obviously), because that's not how physics work. Mind you also, that the location of the MAP itself also plays a role, as pressure will be different (substantially lower) in the intake than it is in the back of the manifold due to it moving. So even though the MAP can read at or near baro at WOT, it doesn't mean that at the intake entry point it isn't at like 80 kpa from the speed of the airflow, which would further complicate the idea of WOT = baro. This difference could actually be calculated if the airspeed was known, using the lift formula.

So, once the engine is started, MAP is just MAP, it is simply tuned accordingly for the conditions the car is to run under. Later generation tunes have high and low altitude calibrations, but aren't typically used. The factory uses the MAF sensor for the simplicity and range of operating conditions. It already accounts for pressure changes and heat simply by the way it operates. By using a heated element or "hot wire", it accounts for density of air and temperature much better than VE calculations, which are only mathematical formulas based on a lookup table in % of VE, as we all know and love lol. To get the VE correct, many things must be accounted for, particularly heat bias corrections, and if going to higher altitudes, may require a new tune or sensor calibration. MAF based won't suffer as much as VE will with changing altitudes and conditions.

OP - I see where you are going with it, I'll try and keep this turning in my mind and maybe have an answer for you if someone else doesn't get to it before I get back to you. I'm trying to figure out if you can trick a MAP into thinking max is 60 kPa, but I like your second idea of throwing extra voltage at it to get it above that. I think that's your best option once you get the linear and offset values correct to max out at 60. Normally, when you turn the key, that number will be displayed in the scanner with the engine off. The extra voltage on a relay is also how some guys set up NOS tunes using 2 bar OS's on N/A cars to gain access to a "second" VE map at the flip of a switch. I think you already answered the direction you need to go here.

[freudie1]

I have a 2003 C5 Z06 running a gen3 512K pcm and I can tell you that:

-it reads at initial key on engine off

Excellent. That's easy to "fix" via the voltage modifier box I referenced (i.e. all I need to do is verify output voltage of the oem 1 bar map sensor at KEO and use the difference in voltage as the value I need to add to raise the 2 bar voltage to the expected 1 bar voltage).

-it reads at WOT as defined by a set of tables (ie below x RPM and above y TPS, etc.) Basically a lower rpm WOT area where in theory the intake wont pose a real restriction and MAP will = baro

OK! So now we are getting to the "fun" part. My theory here is that at said low load cells I should be seeing little to no boost (data logging will show me where said "low" load cells). I say this as wherever I am NOT seeing boost load cell wise I can simply use the voltage modifier (it's rpm sensing as well!) I can add appropriate offset voltage values to keep the values the same as what the ECU would see IF i were to be using a 1 bar map sensor. In other words, I believe with appropriate data logging and smart voltage offset load cell values I can essentially trick the ECU into seeing a 1 bar map sensor at low load cells, and a 2 bar map sensor at cells where boost exists (and of course put in my values for 60 KPa and up as values I see fit for boost conditions despite the ECU knowing only voltage = 60 to a max of 100 KPa).

Of course my next "hunt" will be to attempt to ascertain what formula is used to trigger one of these map = baro readings (i.e. your comment about "(ie below x RPM and above y TPS, etc.)". <---Doubt I will ever get said answer BUT I believe I will be close enough via my aforementioned low load value offset idea.

-it reads "all the time" and updates baro IF current MAP > current baro
-max baro is right around 104 kPa

So...this should be easy to account for. Let's say my baro reading at KEO (i.e. ambient/baro pressure) is 95 KPa (which is approximately 4.75 volts from a 1 bar sensor). 4.75 volts from a 2 bar sensor is 14 psi. My application will see max boost in the range of 6-10 PSI. So in other words this scenario should never apply as I would need to exceed 14 PSI of boost before the 2 bar map sensor would natively output a voltage > 4.75 volts (once again, not happening in my scenario).

I have efi live and it has a baro PID and after any WOT activity it pegs at 104 kpa.

The low end ECU on my Harley does not appear to have this PID (or at least not in any scanner I can find...possibly was never known by software writers)..

[freudie1]

Most PCM's will only take the reading before engine start. Barometric pressure at zero Mean Sea Level (MSL) and 288K (15* C) is 101.3 KPA, or 29.92 inches mercury, or 14.7 PSI. Once the engine is started and from there on, the MAP is no longer a "baro" reference. It is absolute, meaning there is no reference. Manifold "Absolute" Pressure means just that. It reads exactly what is in the intake at any given time.

ChopperDoc,

I believe indeed my ECU (recall it's a Delphi unit for a Harley) does read more than at just KEO. It's also utilizes a single MAP sensor (1 bar of course) hence my desire to know how it shares a single pressure sensor for both sensing baro AND run time intake pressure.

At any rate, I think the OP nailed it. That jives with what I have heard from various other "in the know" people and even better I think it's not going to be that hard to account for assuming I use a voltage scaling/modifier device. In fact....EVEN IF it reads at WOT (i.e. WOT = Read the map for baro!) I should for the most part be high enough natively voltage wise from the 2 bar MAP sensor that I am essentially not going to see much variance (i.e. at 10 PSI, the 2 bar will output about 4 V which equates to 86 KPa in 1 bar voltage....so IF the ECU where to read the map at that point and store new baro value as 86 KPa, I would in most situations be about 10 Kpa off baro wise assuming my 1 bar sensor usually "sees" around 96 KPa ambient)....

[freudie1]

I have a 2003 C5 Z06 running a gen3 512K pcm and I can tell you that:

-it reads at initial key on engine off
-it reads at WOT as defined by a set of tables (ie below x RPM and above y TPS, etc.) Basically a lower rpm WOT area where in theory the intake wont pose a real restriction and MAP will = baro
-it reads "all the time" and updates baro IF current MAP > current baro
-max baro is right around 104 kPa

I have efi live and it has a baro PID and after any WOT activity it pegs at 104 kpa.

this affects the injector IFR tables, as they are scaled to "manifold vacuum" ie. BARO - MAP. edit: and the injector deadtime/offset table

most people who run boost have a boost referenced FPR, so that IFR table is "flat" anyways so that is no factor

I have not found any other tables affected by baro value

Not sure how this affects your setup but I hope this helps

I have a 2003 C5 Z06 running a gen3 512K pcm and I can tell you that:

-it reads at initial key on engine off
-it reads at WOT as defined by a set of tables (ie below x RPM and above y TPS, etc.) Basically a lower rpm WOT area where in theory the intake wont pose a real restriction and MAP will = baro
-it reads "all the time" and updates baro IF current MAP > current baro
-max baro is right around 104 kPa

I have efi live and it has a baro PID and after any WOT activity it pegs at 104 kpa.

this affects the injector IFR tables, as they are scaled to "manifold vacuum" ie. BARO - MAP. edit: and the injector deadtime/offset table

most people who run boost have a boost referenced FPR, so that IFR table is "flat" anyways so that is no factor

I have not found any other tables affected by baro value

Not sure how this affects your setup but I hope this helps

Bringing this back to life....I've been consumed with other projects.

I believe this is exactly what my ECM does as well (same generation/time frame/behavior observed).

SO.....without further delay, here is what I believe will "lock" the BARO read function:

Using a programmable controller (Arduino is my poison of choice for these projects) I propose:

1. Take output of 2 bar map sensor (recall I am using a 2 bar map sensor for an ecu that was designed for a 1 bar only) and wire to analog input of Arduino.
2. Input RPM line to Arduino as well.
3. Now...if RPM =<1000 (i.e. KOEO condition) then output 5 V (max voltage which is max kPa value) to ECU. If RPM = >1000 RPM (i.e. engine running) output actual voltage coming into Arduino (from MAP sensor) to ECU. I could ALSO add additional conditions such as <3000 RPM modify voltage of 2 BAR output to 1 BAR values.....not sure if this matters as the KOEO value would be the max value and regardless of where additional "baro" reads are performed while the engine is running, it will never be greater than 5V due to sensor range max).


What does this do? Based on the logic of it reads "all the time" and updates baro IF current MAP > current baro" this essentially locks the baro read to 105 kPa and will never change (once again it's not possible to exceed 5 volts from a 2 bar map sensor).

Ideal? No. Predictable....yes.

Thoughts?

NOTE: This is for purposes of using forced induction with a Delphi ECU that was only offered with 1 BAR resolution. My thought is that baro compensation with a turbo/supercharger is not really needed (and a pain to tune around). Ideally I'd love access to the value to disable baro compensation (every behind the scenes Delphi ECU I have seen DOES have such an option), but I don't (and probably never will unless someone here is friendly with the Delphi MT05 platform).