NA S4 used for SCCA club racing. Rules don't allow any modifications outside the ECU box, and no extra connections to the ECU, including vacuum. The S4 comes with a boost sensor measuring intake vacuum. Can a MS use the S4 stock boost sensor instead of the MAP sensor? Or can a boost sensor output be transformed into a MAP look-alike signal?

 

I'd have to figure out what the signal from that looks like... the answer is "maybe, but you or I or someone will have to do some extra work to figure it out"

 

Does "Rules don't allow any modifications outside the ECU box" mean you have to use the stock MAP sensor? If not, you could use the MS Motorola part externally mounted in the stock location and use the stock wiring to carry the signal.

-Mike

 

Yes, the rules don't allow using any alternate sensors so I'm stuck with the stock configuration.

I'm not sure how the stock boost sensor works, but it certainly provides an intake vacuum signal to the ECU. I believe the S4 harness provides +5 and GND to the boost sensor and gets a voltage signal out of it. Is that how the MAP sensor works? I expect the place to start would be to chart the boost sensor signal vs vacuum. I could also get a MAP sensor and do the same with it if you don't already know its output.

Is the MS code organized so that the boost sensor input could be read, transformed, then supplied as the MAP input? I'm an EE and embedded system programmer (although nothing automotive yet), so I could certainly help if I had a little coaching.

If I've understood what I've read in this forum, the MAP sensor should be the only obstacle in using a fully stock sensor configuration - right?

Thanks,

 

Marty, yes, there's a lookup table that can be changed at compile/assembly time that would allow the use of another MAP sensor. Mapping of the stock sensor output should do the trick.

renns, you out there??? (He would probably know more than I on this... )

-Mike

 

Marty,

What you describe can be done fairly easily. Here's the steps:

1.) Characterize the stock MAP sensor by measuring its output voltage across the full range of expected MAP values.

2.) Using a tool called "kpafactor.exe", create new include files to be used in step 3. You will be actually creating two files, one called "kpafactor.inc", and the other called "barofactor.inc".

3.) Download the desired msnsextra code version. It has the re-compile tools included in the zip file. Replace the kpafactor and barofactor files in the is distribution with the ones you created in step 2.

4.) Re-compile to create a new embedded code file (msnsextra.s19) including the custom map sensor files.

5.) Edit the msnsextra.ini file as needed (vague, I know...). This file is what the desktop/laptop tuning software uses to decode the run-time data from MS. To display stuff proper in psi or kPa, it needs to know the conversions as well.

Now you can proceed with the steps I presume are listed in the FAQ, regarding setting up for Mazda coolant and air temp sensors, and then flashing the code over to your MS. If you have some familiarity with embedded apps, you'll have no trouble at all. Just post any questions here or at msefi, and they'll be sorted out in short order.

 

The S4 boost sensor has a linear output from 3.5 V at 0 kPa to 0 V at 85 kPa.
http://www.i2k.com/~mdoane/BoostSensorResponse.jpg
I couldn't find kpafactor.exe, but I did find the .inc files and looked at the assembly code. Looks like it should work fine, so I guess I'm about to enter the fray.

Here's the plan:
* V 3 MS I kit and MegaStim
* Add the second VR conditioner in the proto area
* Mount in stock gutted ECU case
* Plan the harness connector to MS hookup when I have the parts in hand, but somehow using a 37 pin connector seems necessary so MegaStim can be attached
* Control spark and fuel

Anybody see any showstoppers?

Thanks,

 

I thought for a second the slope was reversed (ie lower voltage at higher maps), but that link shows 85kPa Vacuum... You should have no problems.

Nothing I can see. I can send over a copy of kpafactor.exe by email if you like.

 

With the vacuum line disconnected I get 3.46 V. With 24 inches of hg vacuum applied I get 0.15 V. I didn't know if that was reversed or not, but looking at the code it shouldn't matter.


That would be great - mdoane@i2k.com

Thanks,

 

You're going to have to probably get at least one more point of reference for that map sensor to accurately figure out a curve...

 

I took 13 points, every 2 inches of hg from 0 through 24. The graph in the link has all the points on it. I just mentioned the end points to describe the slope.

 

ahh ok good