Page 1 of 2 12 LastLast
Results 1 to 20 of 35

Thread: Saab 2.8T tuners, lets chat!

  1. #1
    Tuner
    Join Date
    Mar 2004
    Location
    British Columbia
    Posts
    90

    Saab 2.8T tuners, lets chat!

    Hey guys.

    A good customer of mine from the Grand Prix world bought a Saab 9-3 Aero with the 2.8 turbo Alloytec motor and I had a chance to begin tuning it on the weekend. It was cut short by spark blowout and misfires with 130,000km (80k miles) stock plugs, so his homework is to install new plugs with a tighter gap and then we'll get back to it.

    I noted a few things while tuning. First of all, the commanded PE lambda of .94 in the factory tune means its aiming for 13.8 AFR at WOT and I thought no way they'll have fudged the MAF table or something but sure enough my wideband showed 14.0-14.2 out the tailpipe! How the heck do these motors not blow themselves to pieces?

    I input .78 and got 11.8-12.0 at the tailpipe but even just doing that we ran into misfires . What are others running for PE settings?

    The misfires did some interesting things. After a WOT run, the PCM started commanding 1.04 lambda on B1 and 1.58 lambda on B2 and the motor ran very rough. Shut the engine off, restarted, no problem. Guy had experienced this before on a warm engine startup, apparently common with US-spec Saabs, and its fine with a stop and restart. Perhaps something that's triggered by P0300 code?

    I standardized the timing tables above 100% load and zeroed IAT/ECT adders but actual timing appears to come out at 5.5* below commanded at WOT. There's something else in there changing the timing advance that we can't see. I'll have to rework the timing tables over 100% load

    The camshaft timing didn't seem to match what's mapped out either. The car seemed consistently in the 42-48* range with the intake cam and didn't register the exhaust cam angle at all. Has anyone had success altering the intake cam tables? I noticed the Saab Turbo X had much different settings but I'm leary to try them in fear that the mapping isn't proper.

    The MAF correction table is all over the place from factory. Once I changed the table to 1 and start to correct things directly on the main MAF table the transmission started to shift firmer and quicker as the PCM saw increased airload. Might work out well with the "limited" control we have over the Aisin trans to improve the shift feel.

    The max air load settings in torque management all worked well once set to 100%. I did not get into changing the desired air load (boost control) tables yet, hopefully next time around. Engine seemed to be pushing and holding 10psi.

    If you have 2.8T tuning experience, post up and share your thoughts and ideas! I doubt there'll be many, would be great to converse with those that are here to hear your experiences.
    Last edited by Overkill; 04-18-2012 at 10:09 AM.

  2. #2
    Tuner in Training
    Join Date
    May 2012
    Posts
    27
    I am not a tuner myself but am trying to learn how to do it. I drive a turbo x and work at the dealer. I had the same misfires on warm engine startup, a guy on the saab forums said to use ngk lfr7aix gapped to .75mm, these are a step colder/hotter i dont remeber, but it took care of that problem. also these are notorious for bad coils

  3. #3
    Potential Tuner
    Join Date
    Apr 2010
    Posts
    3
    lil off topic but i have a turbo-x also and would be interested in map sharing.

    currently have a maptun stage 1 flash and happy to share that if anyone can tell me how to do that, if it's transferrable.

    i assume a ppc flash just flashes whatever maps you download to the ppc?

    i'm new to the hptuner world.
    2008 Saab 93 Turbo-X manual sedan

  4. #4
    Potential Tuner
    Join Date
    Sep 2011
    Posts
    6
    Quote Originally Posted by azreal View Post
    lil off topic but i have a turbo-x also and would be interested in map sharing.

    currently have a maptun stage 1 flash and happy to share that if anyone can tell me how to do that, if it's transferrable.

    i assume a ppc flash just flashes whatever maps you download to the ppc?

    i'm new to the hptuner world.
    we spoke about this earlier. when you connect the MPVI interface to the car for the first time you will be prompted for the editor to read the car, it will then make a copy of the file on the ECU that you can then edit. If the calibration on there happens to be maptun, then that is what the "stock" file will be.

    as for sharing maps, once you have the VCM Editor all set up, just attach the file to one of your posts, or put it up in the repository on the main page.

    The problem I have seen through modifying Vtuner's file is that the turbo setup the car has runs out of steam at around 5000 rpm with any meaningful boost level. and I noticed that IATs start to climb pretty quickly if you try to force the boost.

    Overkill, I was wondering, how would you go about calibrating the MAF on this car? Every other Maf calibrating guide is based on frequency and this car only allows for voltage.

    to answer one of your questions, the car doesn't register the exhaust cam because it is fixed. the intake cam however is variable, I have tried tuning it but have not really noticed a difference (maybe more knock if anything), but I did this via road testing and not dyno testing.

    One of the reasons you are having issues with the timing, is that the car is rarely ever only at 100% load in boost. And thus the timing changes you made don't come into play. I believe this load can be read as VE, and thus for a boosted vehicle is should be over 100% in boost.

  5. #5
    Tuner
    Join Date
    Mar 2004
    Location
    British Columbia
    Posts
    90
    Hey guys. The customer installed new plugs, one step colder, at .030" gap and misfires were instantly gone. Came back and we finished the tune.

    Ended up with 17* timing and 11.5AFR on wideband/factory lambda.

    Indeed, must scan air load % pid when matching against spark timing and cam angle. Wot timing is still 5* less than commanded, commanding 22* to get 17* actual.

    Added 5% to the air load tables but started to get surging beyond that. Seem to be hitting an airload limiter, wouldn't go higher than ~120% at high rpm, no matter what was commanded. either that or a MAF flow limiter. I tried scaling down the main table and upping the correction factor but no difference so I set it back.

    On the MAF, I set the correction factor table to all 1 and made revisions to the main maf table as needed for fuel trims and make actual lambda match commanded.

    Also be careful in PE as enrichment time is not the same as enrichment rate. I left that stock.

    The cam tables were laid out correctly once I correlated it to airload %. I advanced the cam at WOT to -17 up to 5000rpm advancing further to -12 at 6000rpm. Added some good midrange torque.

    Overall good performance gains. Drove much smoother with the timing advance more stable. Not the rip your ass out WOT at higher rpms that we were expecting but good throttle response, good midrange, good mid throttle.

    He's looking to have the airload tables remapped by a 3rd party, doing some things that HPTuners doesn't allow, after which we'll resume the tuning at WOT.

    Hope that helps someone else! Nice cars, had fun tuning it.

  6. #6
    Potential Tuner
    Join Date
    Sep 2011
    Posts
    6
    Quote Originally Posted by Overkill View Post
    Hey guys. The customer installed new plugs, one step colder, at .030" gap and misfires were instantly gone. Came back and we finished the tune.

    Ended up with 17* timing and 11.5AFR on wideband/factory lambda.

    Indeed, must scan air load % pid when matching against spark timing and cam angle. Wot timing is still 5* less than commanded, commanding 22* to get 17* actual.

    Added 5% to the air load tables but started to get surging beyond that. Seem to be hitting an airload limiter, wouldn't go higher than ~120% at high rpm, no matter what was commanded. either that or a MAF flow limiter. I tried scaling down the main table and upping the correction factor but no difference so I set it back.

    On the MAF, I set the correction factor table to all 1 and made revisions to the main maf table as needed for fuel trims and make actual lambda match commanded.

    Also be careful in PE as enrichment time is not the same as enrichment rate. I left that stock.

    The cam tables were laid out correctly once I correlated it to airload %. I advanced the cam at WOT to -17 up to 5000rpm advancing further to -12 at 6000rpm. Added some good midrange torque.

    Overall good performance gains. Drove much smoother with the timing advance more stable. Not the rip your ass out WOT at higher rpms that we were expecting but good throttle response, good midrange, good mid throttle.

    He's looking to have the airload tables remapped by a 3rd party, doing some things that HPTuners doesn't allow, after which we'll resume the tuning at WOT.

    Hope that helps someone else! Nice cars, had fun tuning it.
    from what RPMs did you advance the cam? On my 2007, the cam table is at -41* retard all the way to 3520. Did you advance from low RPMs at WOT down to -17* or somewhere in the middle? wouldn't you want to retard the cam at high RPM to reduce overlap and get more high end power?

  7. #7
    Tuner
    Join Date
    Mar 2004
    Location
    British Columbia
    Posts
    90
    His 06 was the same, -41 up to 3500rpm and trailed off to -12 by the time you reached 6500rpm @120% airload.

    I chose to try -17, started it at 120% airload and above from 600 to 550rpm. I blended it up to 5500rpm in the 60-100% airload range to make a smooth transition. At 6000 rpm, I trailed it off to -15 and -12 at 6500, which are mostly factory settings.

    The higher rpm cam retard would help cylinder fill with a later intake closing. I'll work on the cam table more on the next tune.

  8. #8
    Potential Tuner
    Join Date
    Sep 2011
    Posts
    6
    Quote Originally Posted by Overkill View Post
    His 06 was the same, -41 up to 3500rpm and trailed off to -12 by the time you reached 6500rpm @120% airload.

    I chose to try -17, started it at 120% airload and above from 600 to 550rpm. I blended it up to 5500rpm in the 60-100% airload range to make a smooth transition. At 6000 rpm, I trailed it off to -15 and -12 at 6500, which are mostly factory settings.

    The higher rpm cam retard would help cylinder fill with a later intake closing. I'll work on the cam table more on the next tune.
    thanks man that was exactly my question. I would suppose the limitation of cam retard at higher RPM would be knock correct?

  9. #9
    Potential Tuner
    Join Date
    Sep 2011
    Posts
    6
    any updates?

  10. #10
    Tuner in Training
    Join Date
    May 2012
    Posts
    27
    Quote Originally Posted by Overkill View Post
    Hey guys. The customer installed new plugs, one step colder, at .030" gap and misfires were instantly gone. Came back and we finished the tune.

    Ended up with 17* timing and 11.5AFR on wideband/factory lambda.
    Tuning in open or closed loop?

    Indeed, must scan air load % pid when matching against spark timing and cam angle. Wot timing is still 5* less than commanded, commanding 22* to get 17* actual.

    Which tables are you changing for commanded spark timing?

    Added 5% to the air load tables but started to get surging beyond that. Seem to be hitting an airload limiter, wouldn't go higher than ~120% at high rpm, no matter what was commanded. either that or a MAF flow limiter. I tried scaling down the main table and upping the correction factor but no difference so I set it back.

    On the MAF, I set the correction factor table to all 1 and made revisions to the main maf table as needed for fuel trims and make actual lambda match commanded.

    I was able to correct the maf table in the same manner using my wideband for actual afr vs the commanded and it seemed to work pretty well.

    Also be careful in PE as enrichment time is not the same as enrichment rate. I left that stock.

    The cam tables were laid out correctly once I correlated it to airload %. I advanced the cam at WOT to -17 up to 5000rpm advancing further to -12 at 6000rpm. Added some good midrange torque.

    Overall good performance gains. Drove much smoother with the timing advance more stable. Not the rip your ass out WOT at higher rpms that we were expecting but good throttle response, good midrange, good mid throttle.

    He's looking to have the airload tables remapped by a 3rd party, doing some things that HPTuners doesn't allow, after which we'll resume the tuning at WOT.

    Hope that helps someone else! Nice cars, had fun tuning it.
    I can post up my tune and logs if needed, I have a 2008 turbo x with the 2.8t

  11. #11
    Potential Tuner
    Join Date
    Jan 2019
    Posts
    5
    hi Stefan from the old continent .

    I've just purchase an Hp tuner kit for my saab 2.8t , i'm not new in this world as i'm tunning sucessfully for years on stand alone engine management for money. but I will need some Help understanding some of the menus....In my world they call a cat a cat and not an a hairy ball with claws that can scratch if not happy.

    yes ,i'm confuse with some terminology used in HP tuner. if some of you can sort me out to get me on my way because base on the name only i dont see where i can increase airflo/pressure or demanded torque etc....


    thanks for your supports guys.

    Capture 2.JPG

    Capture 1.JPG
    Capture 3.JPG
    Capture 4.JPG
    Last edited by Noos; 02-02-2019 at 08:52 AM.

  12. #12
    Potential Tuner
    Join Date
    Jan 2019
    Posts
    5
    Additionaly i can transfere my hpt file. if needed

  13. #13
    Senior Tuner cobaltssoverbooster's Avatar
    Join Date
    Dec 2008
    Location
    Nevada
    Posts
    4,452
    Figure 1.....Upper axis is Percentage of Toruqe output or Request. The values in the middle are percentage of Load Desired or commanded.

    Figure 2.....Think of this one as a torque table in terms of 100%.....100 is equivalent to 100% allowed output (no limiting)

    Figure 3.....Optimum Torque is a model of the efficiency output. based on max capable output. Typically used to tune in the sensitivity of the system. The closer the Optimum matches the Desired the more urgency there is to achieve that desired torque request from table in figure 1. If its too low then it gets really sluggish. Typically tuned to drivers desired feel.

    Figure 4.....Sets the efficiency of the engine at said lambda values. typically left stock as its already accurate and requires lots of sensors and steady state testing to verify.
    2000 Ford Mustang - Top Sportsman

  14. #14
    Potential Tuner
    Join Date
    Jan 2019
    Posts
    5
    Thanks a lot cobaltssoverbooster.. i see where i can go now.
    i will remove limiters of figures 2 and start increasing demand of loads in regards with torque of fugure 1 and see what happen .. just need to dial in my wide band to monitor all that and go go dyno.

  15. #15
    Tuner in Training
    Join Date
    Nov 2016
    Posts
    13
    I am in the middle of tuning my large turbo =, Turbo X Some of you might know me from FB groups.
    I would be grateful to Bill or anyone else that would be willing to put some much needed R&D time into our ECUs for HPT.
    If we could just get people to chime in to ask, I think it would help them along a bit.
    So much on the table.

    So there are a couple of threads at the moment open about the 2.8 but perhaps this will work here as well.
    Firstly understanding the cam advance .
    I assume positive values increase the intake degree centerline thus RETARDING the timing (reducing overlap with the fixed exhaust cam)
    So when you are at lower RPM under load, it increases cam overlap via negative numbers, and then slowly increases centerline value and thus retarding the cam up till redline.
    Factory Intake cam degree.PNG
    Let's start with that.
    I also need to be able to add a larger MAP sensor but HPT doesn't have this mapped out among other things like fuel puressure PIDs etc.

    Also, I need to go with a larger MAF, what is the best way to dial in the MAF with the limited feature set? Either stock or larger diameter housing?

  16. #16
    Senior Tuner cobaltssoverbooster's Avatar
    Join Date
    Dec 2008
    Location
    Nevada
    Posts
    4,452
    cam events follow the idea that separation is typically max around idle and shift towards overlap through the mid range and wot. the mid range will typically be higher than wot but this all depends on what the application wants.

    if that is a stock table(i dont have a file)....then i see separation at high + scalar values and overlap with high - scalar values. so starting from high loads and low rpm going to high loads and high rpm i see an overlap period that tapers down towards a separated event.

    you should have the ability to change the calibration of your factory map sensor. this would allow you to read to a further range but it wont add any length to tables axis that are based on boost pressure.
    as for the maf sensor install it in a large housing. then you can delay the enable rpm to prevent the lower range of the sensor (where its not as accurate) from being used to make fuel corrections.
    2000 Ford Mustang - Top Sportsman

  17. #17
    Senior Tuner cobaltssoverbooster's Avatar
    Join Date
    Dec 2008
    Location
    Nevada
    Posts
    4,452
    if you guys are after reading material then check out the LNF Bible Thread in my signature. Your ecu functions almost identical to the E69 that the lnf uses in terms of Torque Modeling.
    2000 Ford Mustang - Top Sportsman

  18. #18
    Tuner in Training
    Join Date
    Nov 2016
    Posts
    13
    Thanks friend.
    You should pull up a Turbo X tune from the repository... We're working with extreme bare bones... Seriously go check it out.
    Good tip on the delay, which delay are you talking about specifically where would that be located. Remember if we ask about this it's likely because we don't have it.

  19. #19
    Tuner in Training
    Join Date
    Nov 2016
    Posts
    13
    Like, for instance we don't have a frequency correction curve.

  20. #20
    Senior Tuner cobaltssoverbooster's Avatar
    Join Date
    Dec 2008
    Location
    Nevada
    Posts
    4,452
    Thanks for the file direction. I pulled one and took a look.
    This ecu is damn near the same as the LNF E69.

    The table posted above in post #12 image #1 (Torque to Load [T2L]) i may have miss informed you on so here is the clarification. In torque axis on the top is record-able through the "absolute load" or "Load" Channels in the scanner. the plotted value is an airmass value. 255 is the typical limit for this table when i used to work on the E69 platform. We believed this was directly correlated to map pressure but after testing that doesn't seem to have any proof to back it up. Think of this table as an air load request table. The higher the number the more air the ecu is going to ask for which will bring in more boost as a side effect. If you increase it towards 255 in the max load column then you need to check for stability in the tunes performance. this max load column will apply to any load value higher than the max listed (90%).

    The MAF Sensor. You can disable it but unfortunately they run better plugged in. There is no offset value in this ecu. The only way around maxing the sensor out is to install it in a large tube and calibrate it. To calibrate the maf Calibration table you need to first disable the maf Correction table first by setting the whole table to a value of 1. Tune the maf calibration curve in and then com back and apply corrections to the calibration base table to complete the maf tune accurately.


    I would work with HPT Tech Support to get the wastegate control tables included. They have to be in there since this is such a close platform to the E69. The gate controls makes calibrating boost way easier.
    At this time all you can do is raise the T2L until the boost is at a desired level. Overshoots can be managed with Max Airload Torque table (MALT).
    Disable the Under and Overboost systems as they will cause limiting issues through torque management.

    If you contact Tech with Feature Requests then you should have these tables:

    Map Sensor Calibration
    Wastegate Base Duty Cycle
    Wastegate PID Table Set
    Fuel System Commanded Rail Pressure Table Set in MPa
    Fuel System Injector Injection Angle Table Set

    Dont forget to mention the similarity in tables to the E69 LNF ECM. They may be able to work faster at addition tables if they know the formats of the tables they are looking for.
    Last edited by cobaltssoverbooster; 02-02-2019 at 09:42 PM.
    2000 Ford Mustang - Top Sportsman