Cold running issues...

[Corey872]

Kmanager 1.4.1.0
02 RSX w/ mods in signature

I seem to be having 3 issues occurring at the same time - don't know if they are related or not. I will try to get some datalogs which show each clearly, but though someone might have some insight until then. Questions are at the bottom.

1. I turned off closed loop and tuned the cold start characteristics until the engine ran great, then I enable closed loop. When I first start the car it runs/drives fine for the first ~15 seconds until the O2 sensor is warm and the car switches to closed loop. Short term fuel trim then swings wildly negative as the ECU tries to take out my cold enrichment and get the mix back to stoich. Running stoic with the cold engine then kills the drivability, though once the engine is warm (~5 minutes of gentle city driving at 32F and below) it is back to normal good driving / stoic mixture as controlled by the O2 sensor.


2. I always seem to have ~1/2 second lag between TPS going to zero and the fuel injectors stop firing. I have checked 'disable fuel overrun cutoff delay' and made sure the 'advanced' cutoff pressures are high enough. TPS is 0 and MAP is lower than the cutoff pressure, but the injectors are still firing. One thing I don't quite understand is say I shift under light throttle at 3400 rpm on the low cam. In the advanced tab, the cells for 1093 and 1406 rpm on the low cam are highlighted as are the cells for 2375 and 3000 on the high cam. I would think this would read solidly from the highest cells of 1875 low and 3000 high, and be using the value in 1875 low?


3. The fuel trim and the AFR always seem to be about 1/2 second out of phase. This is especially notable during shifting due to the condition above. The TPS snaps closed, but the injectors stay on, this creates a rich surge and STFT starts shifting negative to compensate. Just about the time STFT goes full negative, I'm opening the throttle which now creates a huge lean spike and takes the trim another 1/2 second to compensate.


Any input on the these conditions would be great. As for specific questions

If I have fuel enrichment for cold temps, but the car goes into closed loop, who wins...the enrichment or ECU driving to stoic? Right now, it seems the stoic does.

If the fuel overrun delay is disabled and the actual MAP is below the cutoff MAP, shouldn't the injector immediately switch off?

If I shift at 3500rpm on the low cam, which cell would be read...the 1875rpm low cam, or the 3000rpm high cam?

Does the ECU interpolate beyond these rpms or just use the value in the highest cell? It is a fairly long gap from 1875 to 3500 rpm.

Does the 'injector size' input in Kmanager have any bearing on the time the ECU takes to react to changes? Would the injector cut off quicker if I put in a 500cc or 2000cc size instead of the actual 1000cc I am running?

I would also note, this is not a fuel 'dribbling' after the injector is off - the datalog actually shows 1.5ms injector on time.

[Spunkster]

http://www.hondata.com/forum/viewtopic.php?t=3629

[Corey872]

As I said, I'm working to get clear, concise datalogs. Is there any chance I could get answers to the technical questions in the mean time?

Any input on the these conditions would be great. As for specific questions

If I have fuel enrichment for cold temps, but the car goes into closed loop, who wins...the enrichment or ECU driving to stoic? Right now, it seems the stoic does.

If the fuel overrun delay is disabled and the actual MAP is below the cutoff MAP, shouldn't the injector immediately switch off?

If I shift at 3500rpm on the low cam, which cell would be read...the 1875rpm low cam, or the 3000rpm high cam?

Does the ECU interpolate beyond these rpms or just use the value in the highest cell? It is a fairly long gap from 1875 to 3500 rpm.

Does the 'injector size' input in Kmanager have any bearing on the time the ECU takes to react to changes? Would the injector cut off quicker if I put in a 500cc or 2000cc size instead of the actual 1000cc I am running?

[Spunkster]

If you are on the low cam it will never read from the high cam maps.

The ECU always interpolates between the columns no matter the spacing.

The size of hte injector listed in the software is only used to adjusting the fuel trim, and has nothing to do with how the injector reacts.

YOur other questions you need to provide datalogs for us to see what you are referring to.

[Corey872]

- serial number of your K-Pro: 2227
- KManager version: 1.4.1.0
- the calibration: attached
- a datalog of a problem: attached
- what sort of vehicle you have: 2002 RSX Type S
- what sort of ECU you are using: Stock PRB ECU
- anything about the car which has been modified from stock which may affect the ECU (don't include what sort of rims you have nor stereo etc):

Engine:

Stock block, bored 0.5mm over, plateau honed
Mahle 12.5:1 pistons, Eagle 4.470" H beam rods
Reciprocating assemblies balanced to .05 gram
Stock bearings matched to .0001" clearance
Light head porting on intake
ARP head studs and con rod bolts

Intake/Fuel/Exhaust:

RBC Manifold, Custom TB adapter, stock throttle body
3/32" Teflon intake manifold gasket
ID 1000cc injectors, aluminum fuel rail
Walbro 255lph in-tank pump
Mallory 1:1 rising rate FPR @ 50 psi base pressure
Kpro engine management
NEW Bosch 9005 O2 sensor

Precision 5857 Billet turbo
.60AR E cover compressor
.70AR T3 Turbine, port matched to manifold
TiAl 38mm wastegate
Sidewinder manifold, light porting at exhaust runners and turbo inlet
3" exhaust, 3" custom exhaust with dual Magnaflow resonators
Stainless steel flex joints in down pipe and main exhaust pipe

Coatings:

CBX Powerkote - Ceramic heat reflecting coating on combustion chamber roof, piston crown
TLTD - Oil shedding / thermal transfer coating on piston bottom, con-rods, oil pan, timing chain cover
PKSX - Extreme pressure lube on wrist pin journals, cylinder walls, turbo bearings
DFL-1 Dry film lube piston skirts, Blow off Valve


- a specific description of the problem:


1. Injectors don't shut off when TPS is 0 and MAP cutoff pressure is met

2. Fuel Trim and AFR oscillate wildly and 'out of synch'


Datalogs:
fuel cutoff half second delay.kdl

This is typical of the fuel over-run. Car is warmed to operating temperature and cruising highway under light throttle at nearly 0 STFT. At around 12s in the datalog I let off the throttle. TPS falls to 0 at 12.144s. MAP falls below 45 kPa (the value in my 1875 rpm cell for cut-off) by 12.291s. Both conditions for cutoff are now met. The injector remains firing at 1.85ms and STFT ultimately drops to -17% trying to shut off fuel. But the injector doesn't shut off until 12.957s. 12.957s - 12.291s = 0.666s This is repeated several times in the datalog and each time there is 1/2 second+ delay in the injectors cutting off. 'Disable fuel over-run cutoff is checked in closed loop and the pressure in advanced overrun cutoff for low cam is 45 kPa at 1875 rpm. These should cause the fuel to cut off nearly immediately, no?

idle oscillations.kdl

Engine is warmed and idling, very nearly 0 STFT. I smoothly open the throttle to get a fast idle at 1500 rpm. The STFT is continually ~1 second 'behind' the AFR. The AFR goes lean but just about the time the STFT catches up, AFR is swinging the other way. This oscillation grows from a slight ripple to a wide swing.

[Corey872]

Here is a datalog of cold start-up with a water and air temp around 40F. Start and idle is good - O2 starts reading at ~23 seconds and by 35 seconds, the ECU goes into closed loop and starts trying to drive AFR to 13.2. This causes some trouble as the engine is still cold. By about 1:10s, the AFCMD has gone to 14.6 even though the water temp is barely 69 degrees. The water temp comp is interpolating between two cells of 14.8% and 6.2%, which by my chart should be around 9% enrichment around 70 degrees water temp?

There is a period around 1:50s where the STFT has gone to -22% to get the AFR up to 14.6 even though the coolant temp is not even 100F yet.

I do a couple of shifts around 1:20s and up through the gears again at 3:25s. In each instance, the fact the injectors don't shut off at 0 TPS causes a rich surge. The STFT swings wildly negative to try and compensate, but because of the lag time, the STFT is far negative just the time the throttle opens after the shift - which creates a huge lean spike and causes the engine to stumble. STFT takes another second to react, then things start to stabilize.

As the engine fully warms up around the 6:30s mark, steady state cruising again around 14.6 AFR and nearly zero on the STFT.


Question:

If I have a temperature compensation set to add fuel when the engine is cold, shouldn't the AFR comand show this? ie - If the temp comp is a value of 0%, then AFRCMD would be 14.6, but if I have a 10% fuel correction, should AFRCMD be 14.6 - 10% or around 13.1?

[Corey872]

Seems like I have been able to fix at least one issue. Just for the heck of it, I decided to load a cal with the "Disable...cut-off delay" unchecked. Now I get an instant cut-off when closing the throttle?!?

Maybe it is just me, but does it seem like that logic is backwards? The box says "Disable fuel over-run cutoff delay", so I assumed by checking it, I was essentially saying.."Yes, disable the delay and shut the fuel off immediately" By leaving it unchecked, I assumed the delay would still be enabled. Oh well, I guess as long as I remember for future reference.

In thinking about this large lag between AFR and STFT, I wonder if this may be due to the location of the oxygen sensor? It is near the stock location on the car, but with the sidewinder manifold and associated exhaust, that is probably 6 feet from the exhaust ports. Wonder if a location closer to the turbine outlet would allow for a quicker reaction time - especially at low rpm when the gas flow is slower?