O2 feedback at idle and general O2 questions

[skylinegtrhr]

Can You borrow original JECS Bosch 740cc injectors to eliminate Yours current bored-out injectors as issue?

[Torque]

Hi,

If you raise TTPMIN you force the ECU to inject more fuel than the actual
computing model would dictate (basically MAF values + injector CC)

I am not entirely sure why TTPMIN is actually part of the ECU design, but possibly they used that value
to keep the engine alive (in certain situations)

So if you have to raise TTPMIN to get away from your popping you're covering up some other issue with your engine.

Keep in mind that huge injectors make it more difficult for the ECU to get idle right.

For example my log shows an injector opening time of 1.5 mS at 1500 rpm. (555cc)
With your injectors being almost twice that size you should be quite below 1mS,


I will double check my statement about closed loop only when in gear.

[Matt]

but possibly they used that value to keep the engine alive (in certain situations)

This is correct. TTPmin is to resolve issues with MAF measurement undershoot, whilst TTPmax for MAF overshoot (for example on deceleration or overshoot on acceleration)

[foliage]

So if you have to raise TTPMIN to get away from your popping you're covering up some other issue with your engine..

Yes you are absolutely right and that is essentially what I am doing, basically I've taken your advice and putting 3s in the TPMIN map as that is what it truely should be if the engine was working as expected, this way the O2 will learn the rest of the map properly, I've then put 8s in at the idle rpm only, as the car is hardly ever at idle the O2 will learn the amount of fuel to put in at cruise etc then when it goes to idle it will hit the tpmin floor and start overinjecting, the O2 will slowly relearn this and it will get lean again but it will run rich for a few minutes at least, as I'm never at idle for much longer than this it means I am rich at idle and I get no popping without completely wrecking the way the O2 circuit works by raising the floor all over the place.

Putting 3s in made the car run like shit for about an hour until the O2 long term learn came down and now it is the same as before.

It is a dodgy fix but I do not know why I need my idle to be so rich for it to not pop so it will do for the time being. Can anyone think of any real mechanical issues that could cause this misfire unless rich? I really do want to try another ECU to completely rule out an ECU issue, however it could just be a bad spray pattern and going to a fully sequential ECU may just cover up the injectors being the issue, so it would not be clear cut as I'm not aware of any other batch fired ECUs.

What would you guys do from here on to figure out the issue? I saw someone said they had some 740cc injectors they could lend to me, thanks for the offer but I don't really feel like pulling them out again unless I have to, if I did I would probably just put some nismo 550ccs in there as they can be had quite cheap these days and even if it is a problem some where else it would surely be an improvement as bigger injectors just magnify problems like these.

[Torque]

Happy New Year ..

So obviously the engine is running too lean at idle and elevated rpm.
Could it be that there is a small vacuum leak?


Is your K factor spot on?
Have you played with the Volt Latency Change / latency parameter?

What happens if you disable O2 sensing all together?


Also does your engine tend to stall when revving the engine up and then closing the throttle?


I would recommend getting a wide band, it's always handy to have and not too expensive.
It's a must have if you want to do more DIY.

[Torque]

Hello,

I double checked and O2 sensing is not dependent on a gear being selected.
(so I recalled that wrong)

However O2 sensing is only on when the throttle is opened, and not at idle (with the TPS closed).
(at least that's on my Z32 ECU)

[foliage]

Happy New Year ..

So obviously the engine is running too lean at idle and elevated rpm.
Could it be that there is a small vacuum leak?


Is your K factor spot on?
Have you played with the Volt Latency Change / latency parameter?

What happens if you disable O2 sensing all together?


Also does your engine tend to stall when revving the engine up and then closing the throttle?


I would recommend getting a wide band, it's always handy to have and not too expensive.
It's a must have if you want to do more DIY.

We smoke tested the intake as I did have a boost leak at some point however it did not turn up any other leaks, I guess a new leak since then is possible

What makes a k factor spot on? I played with it to see if it improved idle, obviously reverted it as it would have wrecked the rest of the tune.

What does the volt latency parameter affect? Not sure I even remember seeing it anywhere.

Got lucky with having nice return to idle, Pete commented that this is often a problem with some engine /ecu combinations.

Other than removing o2 of fuel cells how do you disable it completely?

If o2 is not active at idle then something else is actively tuning the fuel at idle, what else could remove fuel?

[Eric]

I once partly disassembled a RB20DET base rom and from what I remember it had 2 additional AFR related tables active up to ~3000 rpm.
One was TPS related, the other one I don't remember...

One quick way to find out if it is the O2-sensor causing your troubles, is to disconnect it.
Then try to get your idle smooth (and stays that way) and then reconnect the O2-sensor.
if it gets messed up again, just leave the O2-sensors disconnected (presuming you have a wideband sensor to monitor your mixture)

There's also a big chance that the ECU still has a lot of tables that are for the RB20 and not really suitable for the RB25.
That's the risk of using a 'hybrid' tune/ecu.
There are just so many unknown parameters and tables and it would need you to completely disassemble the ROM to figure this out for the most part (which takes a LOT of time) and ten you'll have to find the corresponding tables/values in the RB25 ROM.


> What makes a k factor spot on? I played with it to see if it improved idle, obviously reverted it as it would have wrecked the rest of the tune.

it shouldn't if you have tuned it correctly.
K is multiplier, so it should affect the complete map and not just a small section of it.

[foliage]

That's what I thought it just scales everything equally

[Matt]

K adjustments also affects TP which is used by other tables (and comparison parameters)

[Torque]

I once partly disassembled a RB20DET base rom and from what I remember it had 2 additional AFR related tables active up to ~3000 rpm.
One was TPS related, the other one I don't remember...

One quick way to find out if it is the O2-sensor causing your troubles, is to disconnect it.
Then try to get your idle smooth (and stays that way) and then reconnect the O2-sensor.
if it gets messed up again, just leave the O2-sensors disconnected (presuming you have a wideband sensor to monitor your mixture)
.

Instead of disconnecting the O2 sensor isn't it enough to set the feedback temperature to a very high value?
(Just wondering ..)

[RomChip200]

On Z32, O2 sensors are monitored at idle when TPS is closed, neutral switch has not impact.
Personnally I'm in closed loop at idle and in open loop the rest of the time (I modded the bin for that)
I achieve to make my Z32 flex fuel in this way: I monitor richness at idle and modify K on the fly, with of course some hysteresis and algorithm from my mind.

[foliage]

With a simple small band o2 sensor? Sounds a bit dodgy, I guess if you set simple upper and lower bounds of the main max k and you knew what k was needed to achieve the right afr.

What about ignition timing though?

[RomChip200]

Narrowband sensors are indeed very accurate to give you the stoechioemetric afr, and this is what you want to achieve when being in idle.
Ignition timing is also weighted accordingly even if I don't really boost when running unleaded (ethanol is safer), my timing is pretty advanced for ethanol.

[foliage]

So you've modified the ROM to scale your K factor and some how add timing to the main map?

That is pretty impressive, with ethanol I thought you wouldn't add extra timing at cruise/idle etc, have you used a seperate trim table to get this working?

I think you should post up what you have done and a guide, there would be LOTS of people interested in this sort of thing.