Matt87GTA

OK, so I have an 87 GTA, repinned to 730, running Super AUJP v3 WB with the LC1. Engine is a Miniram'd 383, ZZX cam (240/240, 595/595, 110 LSA), 10.7:1 compression, 42lb SVO injectors @ about 48 psi. Everything is hooked up properly and logging data and emulating properly. Data tracing is working, everything is fine. But here is the problem...

Injector constant is set to 42 and the cylinder volume is set for a 383. At idle I see about 60-70 Kpa. The NB O2 (a three wire) has good crosscounts but reads high and the BLM/INT go to the floor (80-90) while the LC1 reads LEAN, as in 17-20:1 and the exhaust smells strong and burns the eyes idling in the garage (door open of course). So is it really lean???

I played with VE a little bit it really had little affect. I decided to try some drastic things just to see what was going on and put 30 in for the injector constant and the idle cleaned up a lot, the LC1 read in the low 15s, high 14s to 1 but the NB O2 maxes out and stays there and the BLMs remain on the floor. Putting the injectors back to 42 and pulling fuel brings the BLMs up but it goes even LEANER on the LC1...

I have the LC1 simulated NB O2 output running into an old Autometer AF gauge that I had and it correlates with the LC1 readings in the datastream and not the NB O2.

Should I just ditch the NB all together and use the simulated NB output from the LC1 since that does actually correlate to what the LC1 is seeing? OR should I just go to open loop???

84KYSS

With that much cam I would read the stickies about O2 Constants. Hopefully that will help. Good luck.

JP86SS

iTrader: (1)

Thats because the signal is derived from the same sensor. They should match.
Curious what the PW is at idle on the 42's compared to the 30 setting.
Shift your idle o2's for sure to help.
Check that your TPS is zero'd out or you could be getting AE at idle.
I had to raise my threshold of AE MAP to 1.6 to keep from getting pulses of fuel mixed in. I didn;t even see it until I started logging with that new bit in the datastream.
When you were adjusting the VE table you may not have been getting the results you thought.
Double check your ADS file to be sure the MAP value you are tracing with is correct for the VE table. Z69' revealed that he had discovered most (if not all) of the definitions do not have a 10 KPA offset in them.
It is needed when tracing the VE table but not when logging MAP value alone. You may have been adjusting the block next door.
As a test you could change a value and see if there is a difference while staying in the same traced block.

junkcltr

iTrader: (1)

I run open loop under 1000 RPM (at idle). I added a couple of lines of code to the AUJP and made a BIN constant that sets the max. RPM to run open loop. It makes tuning around idle much easier.

I think the latest SAUJP_V4 has this open loop idle stuff in it too.

Matt87GTA

I'm in the process... I had searched and read up on some of the error threshold stuff but missed the constants stuff... THANKS! This sounds promising.

Matt87GTA

Indeed, but my question is how wise you think running the ECM off that O2 input would be?

As for the PW, it was around 1.3-1.4 at 42lb (LC1 reading 18-20:1) and 2.1-2.2 with the 30lb constant (LC1 reading 14-15.5:1).

Looking into that right now.

TPS is set to .57v at idle and 4.5v at WOT. Which constant in TP RT has the AE MAP threshold?

What should I double check? When I play back a log it will read 70-ish MAP and the four blocks at 60-70 MAP and 900-1000 RPM will be highlighted... Seems correct to me?

Matt87GTA

Well I went in and set the O2 idle constants to:

Idle O2 Rich/Lean slow Zero Error Upper: .46
Idle O2 Rich/Lean slow Zero Error Lower: .50
Idle O2 Rich/Lean Threshold Fast: .5
Idle O2 Rich/Lean Threshold Fast Hyst: .05

That had very little effect on anything, still way lean on the LC1, BLM/INT go to the floor...

I locked it into open loop and starting rescaling the VE table and I am slowly getting the AFR on the LC1 down to the low 15s at this point... The datatracing is a Godsend! I bumped the injector constant down to 40 as well to get me closer without running out of VE resolution (hopefully).

Still thinking about using the LC1 NB output to run the ECM once the VE tables are dialed in and letting it go to closed loop.

I'm still wondering why the NB and WB are nowhere near close to correlating...

89gta383

iTrader: (5)

I had the same problems with the LC1. It would show 18:1 at idle and then driving around once I took my foot off the gas it would go from 14.5 to 20:1 then settle back down to 17-18:1 but would error out right when I took my foot off the gas and then come back to life.

I stopped using it for idle tuning and use the blms now. I go to closed loop fine and it runs fine in part-throttle driving. It takes awhile to get the ve table right when buring chips though.

My cam doesn't have as much overlap as yours, but it idles at 70-75 kpa. 230-230 .600 .600 107lsa.

When I get home I'll send you my .bin file.

Matt87GTA

Thanks, I appreciate the .bin to review.

I guess I am hesitant to ditch the LC1 since, from what the Innovate people say, it is infinitely more accurate than a NB O2 which drives the BLMs. And my nose seems to verify that as it is much less "stinky" and idles smoother when the LC1 reads closer to 14.7:1 at idle even though the BLMs are on the floor. Locking open loop was the only way to verify this.

So who do you trust???

I spoke with one of the local guys that tunes a lot of LT1/LS1s and he actually ran into this exact same issue with an LM-1 on a 95 LT1 383 with the ZZ409 cam from TPIS in it... BLMs on the floor, LM-1 reading way lean. Successive attempts at rescaling VE to bring BLMs up resulted in even leaner readings on the LM-1...

mike1986fyrbird

What are the narrow band MV's sitting/doing at idle?

The LC1 usually reads lean because of camshaft overlap at idle. The extra oxygen from the cam overlap goes into the exhaust and fools the 02 sensor into thinking it is lean when it isn't. If it was actualy 20:1 afr the car wouldn't idle at all.

What is your VE set at where you idle??
Idle spark advance?

What do both of them read right before the car goes into closed loop idle?

Matt87GTA

I'm at work and will have to get back to you on the actual mv of the NB O2, but I recall it being about 800mv when the LC1 reads in the low 15:1 AFR range.

The VE isn't really "set" at all right now... I have been moving it up to try to get the LC1 to read a decent AFR. I began this tuning session with the VE tables that came in the Super AUJP V3 though and just changed the cylinder volume and injector constants to see where that put me first off.

EDIT: I believe the idle values (60-70Kpa vs. 900-1000RPM) are around 80 with the 40lb injector constant value (42lb SVO injectors) and the cylinder volume set for a 383.

I have to look again when I get home but I believe I had the idle spark area at about 25*.

I agree in that a 20:1 mixture would not have enough fuel in it to keep the car running worth a crap... And I understand that the large amounts of oxygen in the exhaust will tend to throw off the sensor into thinking it is lean when it really isn't.... What I don't understand is how the NB O2 and the LC-1 can be sooooo far off in the opposite directions if they are both seeing the same exhaust conditions....

85MikeTPI

iTrader: (2)

As an aside to your NB-vs-WB, have you done the PW mod
to your BIN?

https://www.thirdgen.org/forums/diy-...limit-fix.html

While the 383cid doesn't usually hit the bug, the 42# injectors would
probably knock it over. Easy way to check is to set the tables
for a 350 with 24# injectors, and see how the sensors react.
If more normal, then apply the patch to the bin and retry with
your actual values..

(I say this because I'm running 36# injectors on my 427cid motor... YMMV)

mike

mike1986fyrbird

If your VE Table values are at 80 where you idle that is WAY too high.
I had to cut that area in half pretty much from the stock values.
My car idles around 800-900/60-70kpa and is at 37 for VE.

You should set injector constant to what the injectors actually flow at the fuel pressure you are running. Not a big deal.

Volumetric efficiency; stock cams are more efficient at idle then big cams which are more efficient at higher rpms. The ve value in the idle range for your cam should be signifigantly lower then the stock came idle range (35kpa?) which has a VE value of about 48.

Matt87GTA

I know! I completely agree and understand that a large cam with lots of overlap will be less efficient until you build some RPM. It seemed precisely backwards to be doing what I was doing but I thought the LC1 would be the sensor to trust over the NB O2... It seems as if that is not the case.
----------
I am using a Super AUJP V3 (with WB patch) and from what I read, that has the PW fix in it. Is that not true?

JP86SS

iTrader: (1)

That is true, the PW fix is in that version.
The thing you may be overlooking is to not use only 1 item as an indication of what is going on.
Every tuning device has drawbacks that if taken as the only feedback can lead you to incorrect conclusions. If the WB doesn't indicate correctly at idle then don't use it as the basis for making adjustments (yet).
Stinky idle, rough, whatever... Get it to not be rich at idle, add some spark or remove, and smooth it out, then see what the WB trend looks like. The actual reading might not be what you THINK you want, but it can guide you.

JP86SS

iTrader: (1)

Yes, TPS AE MAP threshold.
On the other,
Just saying that if you make an adjustment, be sure to see the result in your data. No result could indicate the wrong cell was adjusted (10 Kpa higher cell should have been adjusted)

Matt87GTA

Thanks! I know all of that, but it is easy to forget it in the "heat of the moment" ...lol.

So I basically did what you just said. The LC-1 is just not correct, period. It can't be. The thing is PIG RICH when the LC-1 thinks it is in the low 15:1. I can smell it, the PW reflects that, everything is indicating that it is actually much closer to what the NB is telling me.

So... I scrapped the whole open loop idle and tuning with the LC-1 idea.

I reenabled the BLM learn (lowered the temp to stock again), rescaled the VE according to what actually would make sense for this engine (thanks Mike1986fyrebird for the reality check!) - as in lower the idle and lower RPM VE and raised the higher RPM VE to at least start to sort of match what the actual efficiency of this engine really is. I set the injector constant to 45lb and I am going to leave it there as that is what they really are at the pressure I am running.

The results were very promising. The BLMs were at 120-128 at idle, the exhaust was nowhere near as strong smelling as it was before, and it was a smoother idle (as smooth as this cam will idle that is). PW was at about 1.5ms, timing at about 25-27* .

I will be heading out to the garage to play with it some more today...

Matt87GTA

I checked that table and it is currently unadjusted from what the S_AUJP_v3 has it at which is 3.13... So you are saying you moved it UP to 1.60 on yours or you moved it up 1.6 from 3.13?

JP86SS

iTrader: (1)

Sorry, I meant "Min TPS for AE"
Addy 852E in case the name is different in your definition.
Now that your idle is clearing, does the reading on the LC-1 indicate differently?

Matt87GTA

None of the AE tables or constants match that address in the .xdf I am using... Should that be defined already in the 8d_v03_wb? I got the ecu from Moates and then converted into an xdf...

The LC-1 is still reading quite high. Idle BLM is very close to 128 now all the time and the LC-1 is reading 18-20:1 still. But the engine is much happier at this air fuel ratio... Part throttle is coming around but still a bit lean in some areas so I can't really say if the LC-1 is useful at part throttle light load stuff. But as I let the clutch out and get rolling in first gear the BLMs will be 130-140 at a MAP of 50-60 but the LC-1 will read 18-20:1 there as well with an occasional dip into the 17:1 areas.

Is there a tool that one can enter a given engine volume, and injector size and get a rough estimate of what the injector pulse width should be? I suppose that is all so relative to the specifics of the engine like the cam and cylinder head/combustion chamber design that it would be difficult to really lock down... But a ball park figure would probably be possible I would guess.

I did a little more VE scaling at idle and with the BLMs very happy the PW bounces from 1.30-1.39 - which is much, much lower than what my old calibration was (1.8) and even lower than what the LC-1 would have me running at (2.2!)...

JP86SS

iTrader: (1)

The Address of the calibration variable would display as "52E" in TP.
This is the "minimum Differential TPS% to Enter AE"
Not what I thought originally.
Can be added, use conversion of "(X/256)*100"
use 1 signifcant digit to display 00.0% TPS.
This will not help with idle problems, I must have been thinking of something else.

I'm wondering if the free air calibration is correct on your LC-1. In the morning when all is cold, try removing the sensor from the exhaust pipe and doing the calibration that way. Lack of a good "air" sample might make the working reading high. Never experienced that but theoretically could happen.

884+3

"The LC-1 is still reading quite high. Idle BLM is very close to 128 now all the time and the LC-1 is reading 18-20:1 still "
Sorry if this is old news to you but have you tried changing the speed settings on the LC-1? According to the owners manual there are 3 or 4 speed settings that will give you an averaged afr instead of displaying every puff of air in the exhaust stream. I never got around to testing the settings on mine before the bosch sensor quit on me.. Maybe they will help you get closer readings to the nb sensor.

Matt87GTA

No problem on the AE stuff JP... I haven't messed much with AE just yet so I don't know if I will add that table.

I didn't get a chance to work on it much the last few days, but I will tackle this tomorrow probably... I am going to try a different ground for the LC-1 and recalibrate it like you suggested in totally virgin air.

884+3 - I haven't heard anything about the speed settings on the LC-1 but I will check it out. Mine remains pretty lean all the time when the BLMs are pretty decent though so I don't know if it is a sampling rate issue.

I also have an LM-1 so one of these days I will put another bung in the exhaust or get a tailpipe clamp and run that against the LC-1 and the stock NB O2 and see how that looks.