Announcement

Collapse
No announcement yet.

Switching to TPS tuning

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Switching to TPS tuning

    Currently my engine is tuned using VE with MAP load sensing for both fuel and ingnition on a Haltech PS1000. However, I'm going to be installing GTi-R ITBs soon and will be making the switch to VE tuning with TPS load sensing for fuel and MAP load sensing for ignition.

    My question is how well will my old pressure based fuel map carry over to TPS? If I take the column corresponding to max boost, set that to 100% throttle, set the idle vac column to 0% throttle and linearize the TPS axis values inbetween, how close will the map be? Basically I'd just be scaling my map axis from -20mmHg/20psi to 0%/100% throttle and keeping all the VE values the same. Will that be good enough to drive and then fine tune from there? Keep in mind that because I'm tuning with VE it will automatically make adjustments for pressure so I only need the TPS table, not TPS with MAP correction.

    Thanks!

  • #2
    I'm assuming you're NA?

    As far as the TPS vs. MAP fueling - not even remotely close. The alpha N map (which is the type of tuning you're talking about) will be a function of how much mass flow gets through the throttle body at that RPM and that throttle position.
    '18 Chevrolet Volt - Electric fun hatch for DD duty!


    DefSport Koni Sleeve and Spring Perch Buy!!!
    http://www.nissanroadracing.com/showthread.php?t=5902

    Comment


    • #3
      Nope, I'll be running the GTi-R ITBs with a GT2871r.

      And I know that MAP and Alpha-N are two completely different methods but I thought perhaps there is some correlation between them. What I mean is that as the throttle opens, the manifold pressure changes so why would I not be able to take my map based on manifold pressure and sale the axis to fit throttle opening percent?

      Actually, as I'm typing this I'm thinking that 0-100% throttle would probably scale (if I can scale it that is) only with the vaccum portion of my map from -20mmHg to 0mmHg, not with the whole map of -20mmHg to 20psi as I said earlier.

      My current VE map in vacuum is basically a linear slope from 50% at -20mmHg to 90-100% at 0mmHg. Wouldn't it slope similarly with the throttle? At idle, with the throttle at 0% my VE would still be 50% and at full throttle I'd be at the maximum VE of 90-100% with the ECU automatically compensating for boost pressure in the VE calculations. The only thing I'm not sure about is what happens in the middle. Will VE ramp up similarly in relation to throttle opening as it did manifold pressure since those two are proportional to eachother or will the VE take on a vastly different curve, perhaps ramping up much faster since ITBs equalize to atmosphere quicker?
      Last edited by LuckyX2; 03-29-2013, 01:02 PM.

      Comment


      • #4
        I guess a question that more directly serves my purpose would just be to ask how VE scales with throttle opening. If it's mostly linear then the manifold pressure based map I have now will work fine when translated to a TPS axis of 0-100%. If it's not linear and let's say VE peaks and holds from 50% throttle and up then I'd have to adjust the map.

        So what does a typical VE vs TPS curve look like?

        Comment


        • #5
          Alpha N will not work with a turbo. Pressure has to be constant ahead of the TB for you to even think about Alpha N (and I still think it's janky).

          Why do you not want a MAP sensor again?
          '18 Chevrolet Volt - Electric fun hatch for DD duty!


          DefSport Koni Sleeve and Spring Perch Buy!!!
          http://www.nissanroadracing.com/showthread.php?t=5902

          Comment


          • #6
            I have been told directly from Haltech that VE tuning with throttle position as the load measurement is the way to go for a turbocharged ITB setup. Injector timing tuning with throttle position as the load and a MAP correction factor also achieves the same thing but it's easier to just use VE since that handles all the MAP corrections on its own.

            Technically, what I am doing is not Alpha-N since that's a pure function of throttle position and RPM. What I'm doing is that plus MAP corrections so it works in boost.

            The reason for using this tuning methodology is that ITBs are notorious for giving poor vacuum readings, making it hard to tune with MAP only. MAF is the easiest way but I really do not want to convert back to a MAF; I'd rather just retune with a different method.

            EDIT: Found a Haltech staff member suggesting VE with throttle position.
            http://forums.haltech.com/viewtopic.php?f=23&p=26282

            "For a car with no manifold pressure you would use TPS with Baro correction.
            for a car with plenty of manifold pressure (turbo or not) you would use MAP based VE tuning.
            For a Turbo car with no manifold pressure you would use VE based TPS tuning (MAP is compensated in the VE calculation)"

            Or in other words:
            N/A ITB => TPS
            Single throttle turbo or N/A => VE w/ MAP
            Turbo ITB => VE w/ TPS (VE automatically accounts for boost)
            Last edited by LuckyX2; 03-29-2013, 09:29 PM.

            Comment


            • #7
              Alpha N is just TPS vs. RPM - so that's exactly what you're describing. This 2D map is I suppose a VE map related to throttle position, but it's usually not described as that.

              The problem is that turbo speed and therefore pressure UPSTREAM of the throttle body is an additional variable, so you will have a really high amount of hysteresis in the actual engine VE as you change throttle position. As in, your TPS vs. RPM "VE" map is going to actually change depending on if the turbo is accelerating up to the speed required vs. decelerating down to the speed required.

              So yes, you can get the engine to run at some level with a turbo and alpha N with lots of tuning, but you're not going to get it to run particularly well except for relatively steady state (i.e. full throttle drag run) or driving mostly out of boost.

              There's a reason alpha N is considered an 80's way of EFI tuning.
              '18 Chevrolet Volt - Electric fun hatch for DD duty!


              DefSport Koni Sleeve and Spring Perch Buy!!!
              http://www.nissanroadracing.com/showthread.php?t=5902

              Comment


              • #8
                I'm not sure how else to explain it but no, Alpha-N is not what I'm describing. Just as you said, Alpha-N is TPS vs RPM. What I'm doing is TPS vs RPM and MAP vs RPM all at the same time. Think of it almost as a 3D map but I don't have to give values for the Z (MAP) axis since VE automatically compensates for that. If I was using injector timing mapping I would have to rig up the equivalent by making a 2D TPS vs RPM table and then overlaying a MAP compensation on top of it. I do not have to do this for VE however because MAP is already part of the equation used by the ECU to figure out injection time along with air temp, injector size, engine displacement, etc.

                So basically, all I have to do is tell the ECU how VE relates to the throttle position and it handles the rest. I realize VE is dependent on a lot more than throttle position but it doesn't matter since the ECU handles all the other things it's dependent on.

                Comment


                • #9
                  If you're just using TPS for fueling, then that's alpha N fueling - period. If you're using a MAP sensor for fueling, then that is the primary input, and TPS changes are used as a modifier.
                  '18 Chevrolet Volt - Electric fun hatch for DD duty!


                  DefSport Koni Sleeve and Spring Perch Buy!!!
                  http://www.nissanroadracing.com/showthread.php?t=5902

                  Comment


                  • #10
                    Again, I am not just using TPS. And that's not how the Haltech's handle it. TPS is the load sensing that the VE table is made with and then the ECU uses the MAP sensor as a modifier in its calculations, not the other way around as you stated.

                    Here's a screenshot I took for you:


                    VE with TPS as load sensing for the fuel map. I could do TPS for ignition too but MAP is fine for that since at the lower vacuum ranges where I would need TPS load sensing, the ignition timing is more RPM than load dependent anyway. The load doesn't have much effect until it's in boost and MAP handles that fine. That's another topic tho...

                    So are you following now? It's VE with TPS load sensing. Because it is VE, MAP is considered in the calculations. Were I using ignition time tuning, MAP would not be compensated for in which case I would be doing Alpha-N unless I implemented a MAP compensation myself.

                    Comment


                    • #11
                      So what do the tuning maps look like? If it actually does reference the MAP sensor it would need to be a 3D cube, which 32x32x32 would be 32768 load points to tune.



                      Lots of people have gotten good MAP signals from ITBs, I'm not sure why yours would be any different.
                      '18 Chevrolet Volt - Electric fun hatch for DD duty!


                      DefSport Koni Sleeve and Spring Perch Buy!!!
                      http://www.nissanroadracing.com/showthread.php?t=5902

                      Comment


                      • #12
                        It doesn't reference MAP in the table, rather it references MAP in the internal VE calculations the the ECU performs. So the table is just TPS vs RPM but way more than that is being accounted for in the VE calculations. I essentially get "free" MAP compensation by using VE vs injector time tuning.

                        And I'm sure I could get MAP to work by just running all 4 lines to a vac manifold and getting signal from that but when I'm being told directly by Haltech staff that TPS VE is better with turbo ITBs than MAP VE, I'm inclined to use TPS VE instead.

                        The way I see it, TPS VE should be better than MAP VE on any setup if tuned correctly since it takes everything into consideration that MAP VE does PLUS TPS. From that viewpoint it should always be possible to make a better tune with TPS VE since the tune is based on more information.

                        Here's some info on how VE tuning works on a Haltech:
                        http://forums.haltech.com/viewtopic.php?f=23&t=5463
                        Notice how MAP is part of the equation used when the ECU calculates injector time. So whether my VE table is TPS vs RPM or MAP vs RPM, MAP is always considered. The table serves only to let the ECU how VE relates to a single variable in that equation. I can choose to build the relationship off of TPS and have an extra variable or I can build the relationship off of MAP which it is already measuring anyway.
                        Last edited by LuckyX2; 03-31-2013, 08:21 AM.

                        Comment


                        • #13
                          Originally posted by LuckyX2 View Post
                          It doesn't reference MAP in the table, rather it references MAP in the internal VE calculations the the ECU performs. So the table is just TPS vs RPM but way more than that is being accounted for in the VE calculations. I essentially get "free" MAP compensation by using VE vs injector time tuning.

                          And I'm sure I could get MAP to work by just running all 4 lines to a vac manifold and getting signal from that but when I'm being told directly by Haltech staff that TPS VE is better with turbo ITBs than MAP VE, I'm inclined to use TPS VE instead.

                          The way I see it, TPS VE should be better than MAP VE on any setup if tuned correctly since it takes everything into consideration that MAP VE does PLUS TPS. From that viewpoint it should always be possible to make a better tune with TPS VE since the tune is based on more information.

                          Here's some info on how VE tuning works on a Haltech:
                          http://forums.haltech.com/viewtopic.php?f=23&t=5463
                          Notice how MAP is part of the equation used when the ECU calculates injector time. So whether my VE table is TPS vs RPM or MAP vs RPM, MAP is always considered. The table serves only to let the ECU how VE relates to a single variable in that equation. I can choose to build the relationship off of TPS and have an extra variable or I can build the relationship off of MAP which it is already measuring anyway.
                          All that post does is describe how a VE map is formulated for an engine - notice it doesn't use TPS input whatsoever.

                          The MAP sensor maybe has some slight percentage change vs. your TPS/VE map - but IMO, that's a bit of a kludge.


                          An engine is an air pump, modelling it as such works great when you know what the pressure is across the engine. The TPS setting does NOT tell you this information. You can only infer it with a relatively large error.




                          It can probably work ok over some range, but the science of EFI tuning says it cannot possibly be a perfect tune over a wide range of usage. You can sit there and get the kludge to work somewhat ok, but the easier solution is to use a primary input to the real theoretical model (i.e. MAP as an input to VE).
                          '18 Chevrolet Volt - Electric fun hatch for DD duty!


                          DefSport Koni Sleeve and Spring Perch Buy!!!
                          http://www.nissanroadracing.com/showthread.php?t=5902

                          Comment


                          • #14
                            That was the point of linking to that post, to show that MAP is part of the calculations. Just because he doesn't mention TPS in there doesn't mean you can't use it for load. I already provided a screenshot that shows I can set TPS as the load source with VE.

                            Originally posted by Def View Post
                            The MAP sensor maybe has some slight percentage change vs. your TPS/VE map - but IMO, that's a bit of a kludge.
                            Are you saying that when I switch to TPS from MAP, the table is only going to have slight percentage changes? If so, that's the type of info I've been looking for. I'd like to know how I can expect my table to change so I can get it close to that before I jump in.

                            Originally posted by Def View Post
                            It can probably work ok over some range, but the science of EFI tuning says it cannot possibly be a perfect tune over a wide range of usage. You can sit there and get the kludge to work somewhat ok, but the easier solution is to use a primary input to the real theoretical model (i.e. MAP as an input to VE).
                            MAP is always an input to VE no matter what load source I'm using. That is the point I have been trying to make here. I'm not sure how many more ways I can say it.

                            I would just like to know what some common TPS vs RPM VE tables look like, that's all. But I guess I'll just venture into it somewhat blind and find out for myself. I'll post back with datalogs and then some dyno runs later for MAP and TPS when I eventually get it on the dyno.

                            Comment


                            • #15
                              Do they not describe the math behind the VE table when you use TPS as the load?

                              And a MAP input is *NOT* always an input on a VE table. Not sure why you think that, but VE is just a way of relating how much theoretical cylinder filling you have. MAP is not an input on that whatsoever. It can be based against MAP nicely since VE scales very well with intake manifold pressure.


                              Again, you need to find out what the math is and proceed accordingly.
                              '18 Chevrolet Volt - Electric fun hatch for DD duty!


                              DefSport Koni Sleeve and Spring Perch Buy!!!
                              http://www.nissanroadracing.com/showthread.php?t=5902

                              Comment

                              Working...
                              X