How-To: Painless 3 BAR MAP Sensor upgrade

[mrQQ]

Right, but what units is the distance in? Because both 65,77 x 1 and 296 x 0.22222 is the same "slope" so which is more "right"?

[SSinstaller]

If you do the math you will never have a scalar B = 1. So IMO that's always the wrong value.. x = voltage , y = boost in Kpa

Logically if a 1 was acceptable the engineers at Mazda would have just eliminated an unnecessary variable from the code..

[mrQQ]

well, ax+b=0 fits any given linear expression. So there is only one multiplier in that, and you can define any slope you want with it.

a = slope
b = offset

so what is one more doing in there? esp since judging by everything they are simply being multiplied with each other.

[SSinstaller]

I honestly have no idea what the ECM is doing with Scalar B, but from personal experience and everything I've seen on here--regardless of which MAP sensor you're running it's more accurate when scalar b ≠ 1.

[mrQQ]

So what is more accurate:

43.85 and 1.5
6.577 and 10
131.54 and 0.5
3288.5 and 0.02

?

because in the end they all describe exactly the same "slope".

[SSinstaller]

If your only worried about the slope they're all equivalent. But were not dealing with a line extending to infinity in both directions...

If I had to guess as to the actual purpose of Scalar B it would to place limits on the sensor range.

[mrQQ]

That's exactly why I asked what is B used for.. till we know, we're just guessing.. has anyone ever proven that it matters? Till not, 1 is as good as anything else..

[SSinstaller]

I know in my car my AEM sensor was not accurate using 1 for scalar b.

I also remember people having issues with their sensors not reading the full vacuum range when using a value of 1, but that may have been a sensor limitation.

IMO, there are enough people running these things now that there should be a consensus on the right way to do this...

[mrQQ]

Well I might agree that it would make sense to keep B as close to stock as possible and only recalculate A using that. Just to not throw ECU off much if it uses B in some way.

[SSinstaller]

From what I've seen whenever scalar b is calculated using the slope/distance method the results are very close to the OEM value..

[mrQQ]

I still don't understand what you mean by that.

[SSinstaller]

The sensor spec sheet usually has the full accurate sensor range shown.. The lowest accurate output and the highest accurate output is used. Scalar b can be calculated by taking the slope of the line determined by your datapoints divided by the distance between them..

[mrQQ]

well, yes, but that's just one scalar.

For example, 845 sensor lists 11kPa at 0.25v and 307kPa at 4.75v. Voltage range 4,5v (4,75-0,25), pressure range is 296 (307-11), so its 65,78 kPa per 1v. So that's your scalar A. What is B?

[rfinkle2]

My thoughts on how to convert the Bosch data to ATR from this data:

p2-p1 / slope of linear formula = offset


---------------------------------------------------------------------------------------------------------------


For the 3 bar sensor (#845)...

P1 -> 11 kPa at 0.25V
P2 -> 307 kPa at 4.75V

y = 65.778x - 5.4444

component A: 296 (307 kpa - 11 kpa)
component B: .22 aka 65.778 / 296
offset : 5.444



Tuning for 3.5Bar Map sensor with cobb ap

[mituc]

Originally Posted by rfinkle2 My thoughts on how to convert the Bosch data to ATR from this data: p2-p1 / slope of linear formula = offset --------------------------------------------------------------------------------------------------------------- For the 3 bar sensor (#845)... P1 -> 11 kPa at 0.25V P2 -> 307 kPa at 4.75V y = 65.778x - 5.4444 component A: 296 (307 kpa - 11 kpa) component B: .22 aka 65.778 / 296 offset : 5.444 Tuning for 3.5Bar Map sensor with cobb ap

@Steve@VersaTune - you need to see this, we can't go higher than 250 on A. We do load based tuning with VT most of the time, but still...

[mrQQ]

Now this make sense somewhat, because it would say what each voltage step is. That is its the same as 1/voltagerange. I guess ecu could make use of something like that.

[rfinkle2]

It had better makes sense. The values are as calculated by Mazda for the 2.5 bar sensor.

I took the stock 2.5 bar data and worked with it in order to make sure the stage 0 values (ots cobb / mazda rom) matched and got a linear formula from:

point 1: 20 kpa @ .4v
point 2: 250 kpa @ 4.65v

This gives the formula: y = 54.118 - 1.6471
recall stage 0 map sensor settings (factory rom 2.5 bar)

component A: 230 =(aka p2-p1)=(aka 250 kpa - 20 kpa)
component B: .24 = (slope/ p2-p1) = 54.118/230
offset: 1.65 (negative y intercept)

[mrQQ]

Woot! @Steve@VersaTune; please increase the limit!

[RPM]

Originally Posted by The Nut Originally Posted by mituc Is there any source to get the 3BAR sensor harness? The bitsise-motorsports.com site seems to have some issues, any HTML file it has available is offered for download apparently because the content-disposition HTML header is set to "attachment". It's Bitwise, and Edge has them.

We have them now, as well. And for much less!

Bosch MAP Sensor Plug 'n' Play Adapter Harness @ therpmstore.com

[Motocooney]

I'm using the Bosch "845" 3 bar sensor as used by the OP.

The scalar values listed dont work for me.

When I use:

Scalar A 65.78
Scalar B 1
Offset 6

I get a dashboard boost reading of ~14.21!?!

The values that work for me are:

Scalar A -65.78(negative)
Scalar B -1(negative)
Offset 6.5

This gives me a reading of zero and a proper reading at idle.

Are these original values supposed to be negative?
And if not, what could the problem be?

Thanks!

[Motocooney]

In ATR there is a box to check/uncheck under options
"Show standard units"
Should this be checked or unchecked?

Obviously, the scalar numbers in the OP work well enough for everyone else.
Why would mine be different?

[Motocooney]

14.21 seems to be the barometric pressure
Why would my boost reading when logging with ATR show baro KOEO?

When I unchecked the show standard unit box, it shows .04psi.

[btstarcher]

component A: 296
component B: .22
offset : 5.444

Show Standard Units should be checked.

[Motocooney]

Originally Posted by btstarcher component A: 296 component B: .22 offset : 5.444 Show Standard Units should be checked.

Thanks, buddy

Not sure how I got owned by something so simple but
It seems to be working now with the OP values and an offset of 6.5.

Thank you for replying, dude.

Happy Halloween!

[btstarcher]

I honestly never even checked to see if it worked; Fob proved the relationship mathematically, and I trust math. And Finkle.

[TiGraySpeed6]

Yeah, that's from @rfinkle2; math there, and it works a treat- I've been using those settings for a while now all over the country and have had consistent results.

I've been meaning to add them to the OP forever - should probably go do that .....

[NCspecV81]

Anyone done any math for a 4 bar? How much different will the values be? Would the offset just change?

[smckenzie04]

I have tried the values in the AP help file and the ones posted here but I never get any better than -12.XX PSI with key on engine off. Also my BAT's are 130-138 when the car is ice cold and outside temps are in the 50's. Any suggestions?

[aromig]

Check your adapter harness wiring, the one I bought had two wires flipped and caused that problem. I changed them around and was golden. Pic of harness with crappy drawing to illustrate how wires were reversed is attached. Ignore scribbles for proper alignment.

uploadfromtaptalk1419390611209.jpg

[smckenzie04]

So orange goes on the outside and red next to blue? I'll try to get a peak at it (such a PITA to get to) later tonight.

My new problem is I'm getting 0 pressure from the HPFP and the car won't start. I'm on the stock HPFP internals(only have intake and cat back for now) but fuel pressure has always been fantastic. Any suggestions?

[smckenzie04]

This is what my 3 bar map sensor harness from Edge looks like and I have no idea how I would change that.

[smckenzie04]

After some digging I found out fuel pressure comes back when I unplug the map sensor. I am baffled how one could effect the other but I'm going to try and swap the original sensor back in tomorrow in the hopes of getting it back on the road. I don't need the 3 bar for a few months anyway.

[aromig]

I don't think those wire colors should cross. Basically, the shape of the the plug is different, but the pins should stay the same. I used a pair of needle nose pliers to swap the pins on one end. Original pic from vendor attached.

uploadfromtaptalk1419408809987.jpg


Second pic just for shits 'n giggles.

[FSRRM]

I know this is a thread that I am bringing up from the dead, but it still is very useful and relevant.

I want to double check if those magic numbers work OK for the Bosch 0 281 002 845

Cobb recommend this:

Range: 3 bar
Component A: 65.78
Component B: 1.00
Offset: -5.41

Your magic numbers are this:

Component A: 296
Component B: 0.22
Offset: 5.444

I have a BNR S4, Bosch 845 3 Bar, HTP 3.5, GrimmSpeed EBCS etc etc

Originally Posted by TiGraySpeed6 Yeah, that's from @rfinkle2; math there, and it works a treat- I've been using those settings for a while now all over the country and have had consistent results. I've been meaning to add them to the OP forever - should probably go do that .....

[TiGraySpeed6]

For the Bosch 845 3 BAR, the 296/0.22/5.444 numbers are a safe choice.

Feel free to try both- they're both fine for most elevations. Your goal is to get close to 0 with the key on, engine off (KOEO) in the boost reading on your AP or in ATR dashboard. The original post includes instructions using the Cobb numbers to move up or down as needed if your elevation is substantially off of sea level.

[FSRRM]

Excellent I'll I've that a go today, I've got a S3 to tune today so this all makes perfect sense.

[FSRRM]

Awesome.

Component A: 296
Component B: 0.22
Offset: 5.444

Worked a charm! Thank You

[stonecold0469]

So how bout for a 4 bar map sensor???

[tastyratz]

I know, necro bump... but I wanted to update this old thread with some relevant info to help others.

I noticed a mention of offset changes before the rfinkle numbers were posted but nobody posting their experience after. Maybe my post will save the next guy a few flashes?

I loaded the rfinkle scalars and 5.44 offset showed KOEO bouncing between -0.37 and -0.44

I changed the offset to 4 and result was -0.23
(1.44 reduction resulted in a change of 0.17)

I changed offset to 3.60 and got to -0.16
(0.4 reduction resulted in a change of 0.07)

I then Changed offset to 2.75 and got to -0.01, nice
(0.85 reduction resulted in a change of 0.15)


Each change did not translate evenly so the offset does not appear to be completely linear. While rfinkle's numbers are regarded as great, I might still recommend people still take the time to tweak their offset. I was almost 1/2psi off so the extra accuracy is welcome.