Page 3 of 5

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Mon Jun 16, 2014 10:38 pm
by Matt
Yes, like Unreal, I am running a KA (98 MY) with a 91 ECU. I have no VE map. My K went from 39720 to 21200.
We have found these type of changes to TP can affect the functionality of other features in the ECU. What I found interesting was between S15 AT (370CC) and S15 MT (480CC) that K constant was changed, as well as all other TP reference tables and constants in the ECU!!

So Nissan/JECS rescaled everything around K constant changes. There was no other injection factor changes. One option here is to identify all the changed tables reliant on TP and perform a similar adjustment. Quite a bit of work, but would match the Nissan/JECS efforts with this particular vehicle which had same code bases, but different map definitions
Your injector resizing and MAF adjustment functions seem to be pretty spot on.
The MAF resizing has been improved by trying a bunch of different MAFs on my R34 on the dyno and measuring the amount of K change required between each. Seems to improved a lot compared to the previous ROM Editor figures!
I had to increase K only slightly to get where I wanted to be. Like unreal, I didn't have to touch my fuel maps that much at all. I did generally decrease all values by a percentage, normalizing them towards 0 or 192. I did this because each "step" of increase or decrease away from 192 is a larger step than it was when K was at 39720. I have fine tuned WOT since then.
MAF upgrades to Z32 MAF can help here since it will increase your K constant to offset the change done to injector sizing. There may be other things we can do here to resolve this. One thing I am adding is an final injection multiplier adjustment to remove the need for modifying K
The thing that bugs me is that this isn't the way it should work. I'm forcing the ECU to think the whole fuel map is closed loop and then telling the ECU to disregard O2 feedback and not apply trims.
I'm working on this one at the moment. Similar to my R34 cold start issues, and not able to use a HPX-N2 due to K constant increasing from 300 to 1300 (runs fine when warm, but not when cold).

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Tue Jun 17, 2014 10:16 am
by PL
I love it hoow you glossed over this bit Matt:

"One thing I am adding is a final injection multiplier adjustment to remove the need for modifying K"

I think this is gonna be a big thing. To be able to leave K untouched should make tuning a lot easier. Particularly on combinations that would normally throw K out a long way.

I hope I'm right.

PL

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Thu Jun 19, 2014 6:36 pm
by Matt
S14 KA24DE has an injection adjustment parameter of sorts. I will look over that in the next few days along with VE map disassembly. The parameter for this item is already in the software

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Fri Jun 20, 2014 3:37 am
by unreal-
Matt wrote:S14 KA24DE has an injection adjustment parameter of sorts. I will look over that in the next few days along with VE map disassembly. The parameter for this item is already in the software
Image

I feel like a kid at Christmas! :)

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Thu Jun 26, 2014 5:28 am
by unreal-
Matt

I was looking in Nistune today in expert mode (seeing what options it has) for my S14 KA24DE and i noticed the final injection constant hiding in the software like you said. If this value is changed, will it affect the tune as it sits now? Or do you have to do some additional fixes?

Image

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Thu Jun 26, 2014 12:35 pm
by Matt
I know it affects the tune on the Z32 16 bit ECUs but have not looked at it much on other vehicles. The parameter which I thought I had on ER34 does not do anything. I'm still to look at the S14 KA24DE but been swamped with support work (and subsequent release) this week to get to the S14 KA. Still going over the P11 SR20DE ECU at the moment and then will move to that next

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Sat Aug 30, 2014 3:50 pm
by unreal-
any updates on this?

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Mon Sep 01, 2014 11:28 am
by Matt
Was upto looking at this same parameter on my R34 and finished FP1 code for S14 KA24DE. Once my hand has healed can continue on this one

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Sat Sep 13, 2014 9:31 am
by unreal-
Matt wrote:Was upto looking at this same parameter on my R34 and finished FP1 code for S14 KA24DE. Once my hand has healed can continue on this one
Sounds good :) Ive been hesitant to try this on my own, but am hoping this value actually DOES something; It would be nice to finally be able to use a O2 sensor up front. My MPGs would probably be a LOT better.

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Sat Sep 13, 2014 10:50 pm
by Matt
I'm looking also through all the TP releated parameters (when you modify K constant, it affects TP load scaling). So I've been working out which tables use it so we can rescale those also (comparing S15 370CC AT and 480CC tunes which have same maps but different K and TP scaling)

Will keep you posted, just finishing up critical feature pack updates at the moment

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Fri Oct 17, 2014 7:28 am
by unreal-
any updates on this?

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Mon Oct 20, 2014 10:32 pm
by Matt
I haven't had much luck with changes to my R34 code. The added code does not change the AFRs like with earlier ECUs so I'm back to going through the code. Looking through the Z32 1996, A32, S14KA 1995-96 and B14 code is very similar so once I crack one I will get the rest.

At the moment I've got a few other things piled up on my workload and then I'll get back onto this

What I have found from our dyno sessions is that we modify K constant still for the MAF (based on additional load) to avoid changing any of the TP scales in the ECU and then offset the total injection adjustment to get the fueling back to normal. The reason why we don't change the TP scales is that there is more than those for fuel and timing (ie for o2 sensing, accel enrichment, knock etc)

I'm going to document this further but it goes like if your MAF reads 30% extra then normally your TP would read 30% past the edge of your maps. So you adjust K constant to move your TP back into line with the original scales. Once you do this that also affects your injection, where we then adjust our total injection time parameter to suit. Also when you adjust for injectors, instead of modifying K you only adjust the total injection time. Works a treat, even for larger MAFs like the HPX which measure stupid amounts of airflow (estimated over 500rwkw)

So two parameters are modified, K constant for additional airflow compensation, and then total injection multiplier to adjust for total injection time. This appears what is doing on in the 1993-1995 Z32 16 bit models where two parameters are changed in conjunction with each other. Should be the same for other models then. I'll try it out on S14 KA again tomorrow and let you know

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Tue Oct 21, 2014 12:03 pm
by Matt

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Wed Oct 22, 2014 4:19 pm
by STATUS
s14 is pretty common to do this, must be a tp scalar for the 02 feed back. it will hold 35psi in closed loop till 3200 before it releases lol

Re: S14 SR20DET holding on to closed loop??!?!?!

Posted: Wed Oct 22, 2014 5:46 pm
by Matt
There would be. This is why I added the injection adjustment for resizing injectors. You still use K constant for rescaling maf (since the load changes)

You want to leave the TP scales for fuel and timing (and everything else) alone and just modify for increased fuel where it doesnt affect load so I added 'total injection multiplier' to the ECU code