Auto rescaling not just scalers but also maps

Discuss possible software enhancements/changes here.

Moderator: Matt

Matt
Site Admin
Posts: 8222
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Re: Auto rescaling not just scalers but also maps

Post by Matt » Mon Jun 18, 2018 11:58 am

One thing which is missed from the above, especially regarding the 1000CC injectors is that the fuel values need adjusting by the same amount that TIM is adjusted for. For example:

370CC to 1000CC is about 2.7x the amount that TIM needs to be reduced by, but we also find that we have to flatten the fuel map by the same amount for it to keep the same AFRs (so values around 60 in the fuel map would now be about 22 in the top end)

NeWT
Posts: 32
Joined: Tue Oct 07, 2014 8:18 pm

Re: Auto rescaling not just scalers but also maps

Post by NeWT » Mon Jun 18, 2018 11:27 pm

If you agree that you need to adjust TP for boost, you not only need to rescale TP for fuel and timing, but then also find that your VCT, O2 feedback, Knock feedback, throttle enrichment etc are all operating from the previously configured TP scales, then you need to adjust TP for all of those too.
Are you sure that I need to readjust all TP related parameters if I replaced MAF and K? And if I rescale scalers? I thought "no", because:
Lets talk about real physical airflow in kg/min. And let this airflow be X kg/min at original R34 config at max 160TP, for example. Then I'm changing MAF to Z32 and K for (example) +50%. Then my max TP will be 160 + 50%= 240TP. But as I done this, the airflow of X [160TP pink MAF] will be equal to X [160TP Z32 MAF]. Am I right?
And if we will ajust scalers. I thought that TP scalers are only for fuel/ign/purge/VE maps, and not for other TP related parameters?
Then I think that if I look, for example, to VTC TP enable (=56 at stock ER34), then it will be right to leave it as is cause real airflow at TP56 [pink MAF] = TP56 [Z32 MAF].
Or no?

NeWT
Posts: 32
Joined: Tue Oct 07, 2014 8:18 pm

Re: Auto rescaling not just scalers but also maps

Post by NeWT » Mon Jun 18, 2018 11:42 pm

And after this, when I will want to reduce K and reduce TID to fit my "very strong in HP" car to max TP of ~250 (or smaller, just not to allow TP limited by 255), the K and TP will sail away, and only now I need to look at other TP related parameters? And I think, for example about VTC, it must be:
(56 (stock)) / ((K after MAF replaced) / (K after K reduced with TID reduced)).
So TP in VTC will be smaller then stock now.
It's my opinion. What do you think, colleagues?

Matt
Site Admin
Posts: 8222
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Re: Auto rescaling not just scalers but also maps

Post by Matt » Tue Jun 19, 2018 12:21 am

Lets talk about real physical airflow in kg/min. And let this airflow be X kg/min at original R34 config at max 160TP, for example. Then I'm changing MAF to Z32 and K for (example) +50%. Then my max TP will be 160 + 50%= 240TP. But as I done this, the airflow of X [160TP pink MAF] will be equal to X [160TP Z32 MAF]. Am I right?
Resize R34 MAF to Z32 MAF would be from K=348 to K=600 so that is a 1.72x multiplier. Our maximum TP (at max Z32 load) would be from TP=160 (say with R34 at 12psi of boost) to TP=275 (more than the map can physically index, so that is limited to 255)

The maximum HP capacity of the MAF figures changes from ~290HP to ~500HP

Current scales:
24 32 40 48 56 64 72 80 88 96 104 112 120 128 144 [160]

New scales
41 55 68 82 96 110 124 137 [151] 165 179 193 206 220 248 255

At 12psi where our MAF reads previously at 160 at full boost, looking at the above table we should now be maybe half way here. However VQ map is exponential, so actual position may be less with the Z32 MAF
And if we will ajust scalers. I thought that TP scalers are only for fuel/ign/purge/VE maps, and not for other TP related parameters?
VE is not TP based. Purge is not either. Other TP related parameters (for warmup, O2, feedback, accel enrichment) are available on per ECU basis (based on that I or others have discovered so far from diassembly)

Some background (based on actual tuning, not theoretical), we found massive issues with using a HPX MAF (3" 640rwhp) calibrated on my R34 prior to FP, whilst we could get K constant set correctly during warm dyno runs, the K constant resulted in higher TP and therefore high vlaue TP scales, that all ended up affecting O2 feedback and warmup.

My R34 with such a high TP ran so crap, that I have up using HPX back before working on feature pack. I did not code TIM for the injectors (no one was really running 1000CC five years ago) but for the HPX MAF support. Once I could pull K constant (and TP) lower, then the ECU started 'behaving'
Then I think that if I look, for example, to VTC TP enable (=56 at stock ER34), then it will be right to leave it as is cause real airflow at TP56 [pink MAF] = TP56 [Z32 MAF].
Off boost the TP will kick in around the same area perhaps on both cars. Would need to compare side by side against boost measurement vs TP map refernece to show if it is still activating VCT, and sticking in closed loop in the same areas than prior to tuning

mtnickel
Posts: 26
Joined: Sat May 17, 2014 8:57 am

Re: Auto rescaling not just scalers but also maps

Post by mtnickel » Thu May 09, 2019 11:26 am

To continue this discussion:
- The Limitation on the R34 not allowing TP higher than 255 may be the first legitimate reason I've seen for not performing a rescaling the way we are suggesting.
- The limitation of the S13 not being able to set a K constant high enough would also be a limitation.
It should be noted however that your proposed setup would result in TP SOMEWHAT being inaccurately referenced in other maps the ECU uses.

One thing that I think needs to be nailed down is the function of the MAF lookup table and it's relation to K and TP.

Do we agree that the MAF lookup table basically converts the non-linear voltage input to a data scale representing a percentage mass of air (g/sec)?
I look at it as a translation unit for the MAF we are using. Though the voltage is non-linear, the table lookup should be linear to airflow. meaning a 0 lookup value = 0g/sec and the maximum lookup value = max g/sec = 65535, along with every other value in between. The Nistune table even labels it as such with "percentage" as the second line in the table.
For example:

Z32 maf:
Lookup table
2v = 3111 = 4.7%
2.96v = 11189 = 17%
4v = 29324 = = 29234/65535 = 44.7% of full flow = 223.5hp
4.21v = 34864 = 53.2% = 266hp
5.12v = 65535 = 500hp

compared to a S14 Lookup table:
s14 maf:
2v = 4108 = 5.55%
2.96v = 12126 = 18.5%
4v = 30525 = 46.6% =
5.12v= 65535 = 100% = 266hp.

So now the ECU after looking up the Maf voltage, knows what PERCENTAGE of volume of air is coming into the engine (relative to the total capacity of the MAF).

This is where the K constant comes in. The ECU knows whether we are 10% or 100% of airflow, but it doesn't know how that corresponds to load on the engine and actual airflow. The K constant further converts that percentage flow to actual airflow.

Lets say we're on a stock s14 map.
K=33017
the Lookup table Multiplied by the K constant should equal the same thing regardless of what MAF we're running.
S14 maf: max HP = 33017*65535 = 2.16^10

If we switched to a Z32 MAF, the K would have to go up by 500/266 = 1.87x = 62062

Then lets say the engine is operating at the exact same HP level
266hp.
This will read about 4.21v on the Z32 maf. The lookup table will read 34864. multiplied by K = 2.16^10.

Though this is just for maximum HP. The same applies to all the other critical maps that rely on load (g/sec) or the TP (which is simply derived from RPM and load and injector size).
________________________________________________
Do we agree up to this point? If we can't agree that you should be able to change the MAF combined with K and run the engine exactly as is without changing anything, then we should settle that first.
________________________________________________
Some background (based on actual tuning, not theoretical), we found massive issues with using a HPX MAF (3" 640rwhp) calibrated on my R34 prior to FP, whilst we could get K constant set correctly during warm dyno runs, the K constant resulted in higher TP and therefore high vlaue TP scales, that all ended up affecting O2 feedback and warmup.
No one is arguing that the FP and TIM isn't needed. On the contrary, it is always needed for injector changes (or as I've now found out, in cases where K constant or TP value will "max out" due to binary reasons.
You also mention here that changing the MAF resulted in higher TP Scales. I think this is also an area you misunderstand us. When we chose to change maf, we are not advocating rescaling the TP index at the same time. You can only adjust TP index if you tune or rescale the entire fuel and igntion maps.

Was this MAF swap combined with an injector swap as well? Or were you hitting much higher TP levels on stock injectors?

If you left the TP scales stock, swapped in the HPX maf, and decreased the K constant (probably a lot), it should track the exact same cells as it did before the MAF swap (on stock injectors). If VTC kicks in at TP=24, then it should do that at that TP regardless of the MAF. If you are now wanting to increase Boost, you will be accessing TP values that are not even in the Fuel or ignition map. This is when you'd have the adjust index values to make room, and at the same time appropriately "compress" the stock map into fewer columns. This would require properly setting O2 flags for cells and such. But creating more room in the fuel maps for higher boost has no effect on the ECU knowing that TP =24 is the time to kick on VTC.
Current scales:
24 32 40 48 56 64 72 80 88 96 104 112 120 128 144 [160]

New scales
41 55 68 82 96 110 124 137 [151] 165 179 193 206 220 248 255

At 12psi where our MAF reads previously at 160 at full boost, looking at the above table we should now be maybe half way here. However VQ map is exponential, so actual position may be less with the Z32 MAF
1) The above scale would be bad
a) because you now have no lookup values near 24 in your new scale. These would be the low load values. While the new scale maximum can change, the base should stay the same. driveability would be bad.
b) The fact that the map traces near the middle should not effect driveability. if 160 was in fact your maximum load you'd expect, It would be a very poor use of the map, as you have lower resolution and would never use any of the area on the right of the map. We still would agree that the maximum TP load you need should correspond to the horsepower range you're expecting to run the engine in. But we believe it should be set by adjusting the index, as opposed to the K constant. And if you do adjust the Index, you better move the map data to proper corresponding columns.

One area I may have overlooked is if the INDEX values are used in other maps. Since they are specifically specified for Fuel and Igntion tables, I assume they are not used elsewhere. I believe elsewhere, the ECU uses the current reading TP value, not the index position or column number.

Thank you for highlighting the facts of sometimes hitting binary limit on K constant or Load value.

mtnickel
Posts: 26
Joined: Sat May 17, 2014 8:57 am

Re: Auto rescaling not just scalers but also maps

Post by mtnickel » Thu May 09, 2019 11:34 am

My R34 with such a high TP ran so crap, that I have up using HPX back before working on feature pack. I did not code TIM for the injectors (no one was really running 1000CC five years ago) but for the HPX MAF support. Once I could pull K constant (and TP) lower, then the ECU started 'behaving'
Were you on 1000CC injectors?
What did you do to "expand the TP" higher. Did you compress and adjust the whole fuel map to the lower side, along with deactivating O2 closed loop on cells that shouldn't have it.

Matt
Site Admin
Posts: 8222
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Re: Auto rescaling not just scalers but also maps

Post by Matt » Fri May 17, 2019 7:39 pm

I cannot remember with 1000CC injectors but would be 512 * 380/1000 to start with

With HPX MAF I needed to keep lowering K constant and keep TP around 208 (default timing max) and then adjust the mid-top end of the fuel map and timing map back a little

Post Reply