Page 1 of 3

Basic Tune, K / Lat. for 550cc Can't get it smooth!

Posted: Mon Jul 20, 2009 9:48 am
by gsxryan
I got the car to idle now, but I'm trying to get the Kvalue and Latency set. I'm getting pissed at it because everything i try it's not getting smooth.

I've read through your quick start guides and it makes it sound easy, but Mine doesn't feel to be going as nicely.

o2 is unplugged:

K Value: i drove around about a 1/2mile to get 2500-3000 RPM set to 14.7 using the K value (now set to 137)

Latency, this is the pain in the ass part. I have messed with it at idle for prob about 2 hours now, going up and down trying to get a more smooth idle near 14.7. I adjust it high and it get super rich and will sputter under throttle. Adjust it the other way and it leans to much on accel and will backfire though the manifold. Any place in between still has a very spastic AFR range.. It'll stick at 11.0 and then move back to like 15.0 and repeat again and again. So I feel like i'm shooting in the dark.

Is there something special i should do with the TPS or IACV? It feels like the car is fighting and adjusting for something.

I'm getting one ECU Error CONSULT_DTC_EGT_SENSOR

I currently have the MAF extended from the compressor, but do no have a filter on it.

SR20DET Avenir
Deatshwerks 550cc JECS injectors (these are ported out 259cc)
Stock S14 MAF

Posted: Mon Jul 20, 2009 4:17 pm
by PL
Hi Ryan,

What is the part number of your ECU? Changing just the injectors should be pretty painless.

Firstly, have you ONLY fitted the bigger injectors and S14 AFM or were there other things? I always need to ask this cos it’s often other little contributing things that can cause problems – resulting in trying to tune around existing problems. Which can be very frustrating. Particularly when trying to learn how the whole process works. Incorrectly set up BOV’s and adjustable FPR’s are classic problem makers. Also, is it a known good AFM? Have you cleaned the sensing element?

I’d expect with drilled out standard injectors that you shouldn’t have to adjust latency. The pintle, coil and seat haven’t changed after all so I can see no reason why the latency should change. The golden rule is to leave latency alone unless you absolutely can’t get things right any other way. I’ve found that changing latency can make tuning very confusing if you do it up front.

Make sure you turn O2 Sensing off in the Feedback Flags section to make sure it’s not trying to chase closed loop.

TPS and IACV shouldn’t need to be touched. Just make sure TPS is adjusted to 0.5V and that the TPS closed indicator is coming on in NIStune when the throttle is closed (look in the Gauges Panel).

Just on a side note : it always works best to increase injector size and go to bigger AFM at the same time because they tend to offset each other - this keeps the K value closer to standard. Which means that other parameters get thrown out a whole lot less. Particularly load scaling in your fuel and IGN maps. Because K affects everything. The further K gets from std the more tuning work you have to do.

eg: Avenir std K = 33,107

If we fitted 550’s and Z32 AFM this becomes 33,107 x 370/550 (for the inj) x 500/290 (for the AFM) = 38,400. Which as about a 15% difference.

If we fitted just 550’s then it becomes 33,107 x 370/550 = 22,271. Which is a 32% difference.

Just fitting Z32 AFM is the worst though 33,107 x 500/290 = 57080. Which is a 72% difference!

PL

Posted: Tue Jul 21, 2009 12:34 pm
by gsxryan
It's a g20 with avenir motor, stock g20 MAF (I'm using the s14 280hp MAF setting in nistune, this may be an error, I may have to use the 250hp DE setting on there..)

The ECU number is 0J600 Manual

o2 feedback flag is off.

The 2 other areas in question:
I have the valve cover ventilated (usually plugs into the intake)
BOV is vented. Not routed into intake yet.

The car is driving around with what I've worked on today, but is not very smooth.
At cruising 1500 rpm runs at a 12.0-13.0
At cruising 2500 rpm runs at 14.7 ish
Light acceleration leans out greatly, and it stutters if you give too much gas.
Idle still seems to start out 14.7 and then will go to 12.0 and stay there.

Are you sure the time to open and close the injector will be the same as a 259cc injector compared to a bored 550cc?

My TPS is currently .40v, I will adjust this but this is within the FSM spec.

Posted: Tue Jul 21, 2009 1:10 pm
by PL
OK, BOV vented will cause a rich patch on gearchange. But you probably knew that. Otherwise it should be fine. MAKE SURE IT'S NOT OPEN AT CRUISE/IDLE!

Ventilated rocker cover to atmosphere should be OK.

I'm thinking your problem is the AFM. Your AFM MUST match what you have selected in NIStune or really weird things can happen.

Open the "MAF Translation" table (AKA VQ Map) and check what figures are in there. I just opened a 0J200 G20 map and the first 10 entries were "28". So you need to find an airflow meter with a VQ map that looks the same....

After much searching I found that the B14 seems to use the same VQ map (which makes sense - the two models seem to be the same car - I wouldn't know, we don't get either of them here).

To check this yourself just load the address file for the B14 and then load up a B14 MAP (I used 0J610). Open the VQ Map and you'll see the first 10 entries are "28". Bingo!

So you'll need to change your AFM in NIStune (Operations, Change Mass Airflow Meter) to B14.

Hopefully that'll sort it.

PL

Posted: Tue Jul 21, 2009 5:57 pm
by gsxryan
I've made sure the BOV is nice and shut, in fact it's been surging a bit. But when in it does BO it gets rich as suspected. Guess that'll have to do until i route it into the intake.

When i said i was using s14, i really meant b14. All these acronyms! lol

I just opened my OJ600 binfile and the VQ map is also 28, 28, 28 (it corresponds to the B13 MAFfile)

According to SR20 forums my g20 MAF is good to 250hp. Which brings a discrepancy in choosing with VQ map to use. I pulled up 3 possibilities that it could be. The B14 VQ file seems different.

280hp B14 - 6 22 45 75 112 (currently using)
280hp B13 - 28 28 28 28 28 (default with BINfile)
250hp S14, B14, N14 DE - 246 246 246 246 246

The ID on my MAF is 22680-53J01 AFH50-06B 6411 (google research comes up with BlueBird MAF or B14 MAF)

I'll try to play with these some unless you have some data that ids which one i need to use for certain

I've got 10 miles trying to tune the car now! No leaks, No smoke, just tuning left! Knock on wood...

Posted: Tue Jul 21, 2009 7:20 pm
by gsxryan
Have been digging into the forum for info, found a lot of helpful things that the nistune documentation didn't cover.

One question I have about latency values that I found. Is it still requred to convert ms to microseconds?
Those numbers you mention are in milliseconds. Nistune entry is in microseconds (it also

shows hex conversion in this window)

0.75ms = 750uS * 2 = 1500uS
0.93ms = 930uS * 2= 1860uS

Posted: Wed Jul 22, 2009 1:23 am
by Matt
dont need to multiply by 2 anymore

Posted: Wed Jul 22, 2009 11:54 am
by gsxryan
I feel like I have it VERY close now,

I'm using the DE binfile with the S14 Fuel Map for DET

RAW Value is zero in these areas. o2 unplugged.
crusing 4th gear - 3k RPM - 14.7 STABLE
Cruising below 3K is ~ 12.0
Idling gets around 11.0

This is a latency issue? Meaning i need to lower uS value?
Or should i just plug in the o2 and get that to work..

The car leans out while hitting another gear / acceling from a stop. I have to be very careful when doing this, not to open the TB too much because it's so lean.

Use the G20 0Jxx address file to read.

Posted: Wed Jul 22, 2009 12:21 pm
by gsxryan
Also, Should i have the MAF on the intake side of the turbo (before the filter) - currently this way
OR
Should i put the MAF in the coldside piping closest to the TB?

Posted: Wed Jul 22, 2009 12:35 pm
by PL
I'd keep your AFM pre-turbo. I'm not a fan of blo-thru AFM's as it's too easy to have them get contaminated with oil from turbo/breather. Either way will work though.

Looks like you should be using the B14 VQ map, not B13.

Use "Filtered values" in the fuel map and adjust them for around 0 in the closed loop area. At present you have 64 in them. This will cause your richness at low load.

You'll probably then need to increase K to get high load AFR's correct. I'd expect to be seeing full load figures in the fuel map between 20 and 50. So chop the top off that fuel map.

Pull a bunch of IGN timing out too. You want values of around 5 degrees in the midrange (full load) and ramping up to 10 degrees at higher RPM.

PL

Posted: Wed Jul 22, 2009 2:28 pm
by gsxryan
I'll keep the intake pre-turbo, don't want the hasstle of setting up a blowthrough setup anyway.

The B14 MAF does seem to play best with the car.

I adjusted "Filtered values" in the fuel map for 0 in the closed loop area.
The IDLE is settling around 14.7 now, but light cruising under 3k is still around 12.0. Cruising above 3k is around 14.7.

If i understand correctly, the map is defined like this?
RAW VALUES
0-127 Open loop control : 14.7 - RICH
128-255 Closed loop o2 : 192 - 14.7 Lean - Rich
I don't quite understand how without the o2, on closed loop control targeted for leaner than 14.7 is getting Actual 14.7.

I don't fully understand what you're wanting me to do with this one:
You'll probably then need to increase K to get high load AFR's correct. I'd expect to be seeing full load figures in the fuel map between 20 and 50. So chop the top off that fuel map.
I'm Not seeing the AFRs drop off at all while rolling on the throttle, in fact it seems to get leaner. I've only hit like 1/2 psi while lightly rolling on, but it's not getting richer so I'm afraid to push anymore.


I pulled a decent amount in the IGN timing map too. I simply imported the S14 DET timing map as well for this one. Is it still very far off?

Posted: Wed Jul 22, 2009 5:28 pm
by PL
Hmmm. Now I'M getting confused!

Normally the difference (raw) between open and closed loop operation is 192. Which is simply done by setting the two highest bits of the 8 bit binary word high - so 1100 0000 would be closed loop (which would show up in Filtered view as 0). And 0000 0000 would be open loop.

But there's something weird going on here. Your ECU is not like the S13/S15/S15 and N14 ECU's that I'm used to. It looks like it uses only the last bit to indicate close loop. Which is 128 in raw view.

But if you adjust one cell to 0 in raw view and then hit the "O" key you'll see it toggle between 0 and 192. Not 128. So NIStune is trying to work with 192 and it looks like it should be 128. Matt - hellllp!!

I'd go back to Raw view and just make them 0 again at low load. Sorry to mess you around. These things are not my usual playthings.

Is it accessing the right side of the maps when you boost it? It should go to approx half way across on 0psi boost. Once it comes on boost your numbers in the maps should increase rapidly and it should richen up.

I'd be pulling A LOT more timing out if it's a DE+T that is still running DE compression.

PL

Posted: Thu Jul 23, 2009 1:23 pm
by iaredam
On my U13 it also uses 128 for 14.7, the O key doesn't do anything.

Posted: Thu Jul 23, 2009 2:15 pm
by Matt
It wont on G20/B14 SR20DE or S14/U13 KA24DE ECUs

You will need to set your fuel map to 128 instead of 0 raw value for adjusting your K constant to get low-mid range AFRs as expected 14.7:1 AFR and then adjust your latency to get idle about right

After that set your fuel map to what you want (default is fine, you will notice it will be flat 128 for 14.7:1 around the bottom corner anyway)

Then use the VE table to adjust your trimming to achieve these AFRs

As pete mentioned, make sure the MAF you are using is the same as the VQ table otherwise it will throw inconsistencies and drive you crazy when you try and get your K constant and latency correct when initially setting up

Posted: Thu Jul 23, 2009 3:24 pm
by gsxryan
While narrowing down the K / lat Do you recommend leveling out the VE map to 128?
When I start adding fuel, will I exclusively add values over 128 to add fuel in the RAW table? Meaning, Will i need any reason to ever use any RAW values under 128?

So, when adding fuel I should focus on theoretical AFRs on that filtered view and adjust VE until the actual AFR reaches the theoretical. Will moving VE values lower than 128 result in richer actual AFRs?

I still don't know absolutely sure that VQ table is certainly for my MAF, but it seems to play nicest with the car so far.

Thank you guys so much for the help so far!