So here's an interesting question for you 7730 $8D experts...

When I do a warm or hot restart, for about 30 seconds or so, the throttle crispness is just a little bit better than during normal driving... just enough to be noticeable. I know it's not a function of open or closed loop because I'm already idling in open loop permanently. It's almost like it's some different mode of operation.

My tune is already such that the throttle is super crisp, so if I had to characterize this I'd say it was like there's this extra 5-10% of "crispness" that's noticeable enough for about 30 seconds after a warm/hot that I'd like to somehow get to run like that all the time if possible.

I've gone all up and down tunerpro trying to find anything fueling or spark related that's also a function of time after start up, but to no avail.

Not sure if this phenomenon has always been there or if just my new 383 combo is making it more pronounced... I never really noticed it with my old ZZ4.

 

What values do you have for these items:



Just a stab in the dark, but because you did not mention anything about cold start crispness, it might be there is an AFR (fueling) change going from OL idle to CL when RPM exceeds your OL RPM threshold, and where the times above have been exceeded. Might try changing above times to 0 and see what happens. If crispness is not noticed as before, OL to CL transition may be the issue. Just a guess.

Also, before changing anything, might post an exported CSV log where you indicate a time in seconds when the crispness stops. Can determine that by using cell phone timer to get approximate seconds after startup where crispness stops.

Elky

 

Thanks Elky. I was looking at those as well, but since I idle in open loop, I didn't think they were really playing into it.

My closed loop kicks in at 1500 rpm and this extra throttle crispness is definitely noticeable below 1500.

1500 rpm seemed to be the best spot for the Miniram for seamless transition from open to closed loop.

The throttle crispness thing isn't apparent at cold start. Just on warm/hot restart.

When I get a chance, yeah I'll run the car as is, take data, then run it with those set to 0 and then take data.

As it stands, my settings are 80 sec @ cold, 30 sec @ warm, and 10 sec @ hot. I can say that the throttle crispness thing definitely lasts longer than 10 sec... so it seems to be continuing even after the timer ostensibly tells the ECM to go into closed loop (again notwithstanding that I don't get closed loop below 1500 rpm).

 

Ok, finally had a chance to take the Camaro out...
Files attached below.

baseline run is just normal driving after the engine fully warmed up

restart hot is a restart with the same bin as baseline run

restart hot with timer zeroed out is with the mods you described below.

I don't think the timer zeroed out really accomplished the objective though. Maybe you'll see something in the data.

 

Tks for files, but where are the throttle tap samples for crispness and then it disappearing? Maybe I don't fully understand the issue.

 

I was hoping you'd see something in the data as it mainly shows up on off idle transition. So the data mainly focused on that.

it's a little hard to explain.... its just a subtlety I notice while driving it.

 

Don't see anything. Given the logs are HOT restart and your hot OL to Cl time is 10 seconds, that can't be it because always in CL off idle (>1500). Only thing is WB AFR goes slightly leaner from OL to CL, but that's way past initial throttle response.

 

Ok, thanks for checking Elky. Like I said, the current throttle response is actually very good, so I'm actually very happy even if this question wasn't figured out.

 

I think I got it...

I added 5 deg to the idle spark timing. When I had gotten the 383, I had initially taken out some idle timing due to the wide lobe separation angle on the cam (didn't think I needed as much as with the old ZZ4).

I put it back in and I think I got that last bit of throttle crispness I was looking for.

 

Makes sense as long as no knocks.

 

Ok, so an update...

I was going over some things and found that my injector offset was off by (I kid you not) 32 microseconds in the voltage range I generally operate in, versus the datasheet on my injectors. The value at 14.4V should have had 793 microseconds and I had 761 microseconds due to an interpolation.

So I corrected it.

I was surprised when I found out that it now feels all the time exactly like the 30 seconds right after warm/hot start (what I was shooting for at the beginning of this thread). I thought the additional spark timing was it, but evidently not...

I was surprised because I played around with the AE for quite some time and couldn't duplicate it. But 32 microseconds on the injector offset does it? Doesn't the increase in offset simply mean more fuel? How would that method of more fuel be different than adding more AE fuel?

 

Certainly you are feeling something that is not possible to determine by looking at logs, but consider this when looking at the log re-start_hot.csv (see attached):

1)
It's remarkable that 32 microseconds of fuel can have any noticeable effect. Think about it. Your average idling PW before voltage compensation is between 2.37ms and 2.03ms depending upon which idle area is considered (more below). That's 2370 to 2030 microsec. 23 microsec in that range is a rounding error. And if off idle, it's even more so. Further, the change you made is not actually 32 microsec. Because your voltages average 13.6-13.7v, the value taken from the offset table is interpolated between the values for 12.8v and 14.4v. This means the 14.4 value is not solely being used.

2) There are 3 true idle areas at 0 MPH:
#1 at samples 125-162 (immediately after startup and after RPM settles down)
#2 at samples 350-407
#3 from sample 685 to the end

There appears to be a disconnect. Look at IAC steps. In #1, they average 37, but in #3 they average 12 which seems very low. They are in between these values at 17 in #2. Also, while perhaps not meaningful, RPM is < Commanded in #1, but > than commanded in #2 and #3. Over several samples, you would expect to see average RPM = commanded.

Is it possible that the higher IAC value in #1 lends itself to a better throttle response that in #2 and #3. Don't know, but it's strange the 3x step difference when idling just after startup .vs. later. Could IAC recalibration be in order?


Yes, hiher offset = more fuel, but again the 32 microsec (or the interpolated value) is negligible. AE is provided in ms and is MUCH more fuel.

HTH, Elky

 

Thanks Elky.

Yeah my IAC steps are low at idle because I have that idle-air-manifold that handles the majority of the idle air.

 

This minor change to the voltage offset had such a profound effect I decided to try idling in closed loop again. Amazingly it's just as nice as in open loop. I do have the threshold set higher than stock (about 700 mV), but it's idling really nicely.

Crazy....

 

I made another change to a parameter I found...

AE TPS-Minimum 12.5ms TPS Change Threshold (0x08=3.13%)

Factory it's 3.12%

I changed it to .39% (which I think is the minimum). I suppose the Miniram is sensitive enough to the throttle-tip in that despite how good my tip-in already is, I still noticed a bit more of an improvement.

I also have the AE TPS-Min 6.25ms TPS Chg to Invoke if NOT in AE TPS set to .39% as well.

3% may not seem like much before AE TPS kicks in, but on the Miniram, yeah you can tell the difference.

 

Question...

I'm looking at some time dependent parameters in the ANHT HAC vs Tunerpro parameters...

Do any of these cause the ECM to run in a different mode of operation? Or are they simply performed "in the background". The time durations are long enough to be within my observations about how long this phenomenon lasts...

L8255: FCB 146 ; IF COOL <= 69.5 C, (157F) DISABLE ERROR ; ARG = (DEG C + 40) * (256/192)
L8256: FCB 60 ; 30 SEC, IF RUN TIME < DISABLE 13
L8257: FCB 79 ; 0.035 VDC, IF o2 =< DIABLE
L8258: FCB 124 ; 0.549 VDC, IF o2 > DIABLE 13
L8259: FCB 13 ; 5% TPS, MIN FOR 13
L825A: FCB 30 ; 15 SEC MIN COND'S PERIOD

L8672: FCB 25 ; IF ENG RUN TIME < 25 SEC ENABLE HOT SA RETARD

 

So I'm still curious about this...

One thing I also notice on the hot/warm restart, blipping the throttle in neutral is just super-crisp. Not even a hint or a millisecond of hesitation.

But again, after about 30 seconds or so (even though I'm still idling in open loop), you start to notice a minor hesitation. What's different between the open loop right after start up and the open loop after ~30 seconds?

It's almost like after 30 seconds or so, the accelerator enrichment is being applied differently than right after start up, like some different mode of operation.

Is it in some sort self diagnostic mode that causes the AE to behave differently?

Over time, I've swapped ECMs, memcals, all of the different sensors, removing the Ostrich-II and running the chip by itself, etc, but nothing seems to make a difference.

Anyone have any thoughts?

 

All items except L8672 in this list are used while engine running to determine if error 13 (o2 sensor failure) should be indicted (if o2 mvs remain between 350 and 550 for some period of time).


L8672: Title is really a misnomer. Has nothing to with actual engine start. Is just "Hot Retard" where when all components of SA-TDC are determined and summed:
If coolant >= 113 AND engine run time >= 25 seconds, don't add any SA
else, add 0.35 degrees SA

 

Can't remember if loss of throttle crispness occurs only on cold start or cold and hot. But look at post 12 where IAC steps are discussed. You may be seeing the effects of what's called "warm kickdown". For the most part, after coolant reaches 113*F, IAC steps are immediately reduced by 10. Happens irrespective of OL or CL. You can see this in a log. Start cold engine in P/N, let it run and watch IAC steps drop 10 as coolant reaches 113*F.

The calibration item 0x62D=IAC Param, Steps Added for Warm Kickdown controls this. Default is 5 steps which is multiplied by 2 in the code to yield 10 steps. Might try setting to 0 steps. IAC steps will not be changed.

HTH, Elky

 

Thanks Elky, I'll try that as soon as I get a chance.

So if the throttle response is better with the added steps, then I should expect that after setting this to zero, I shouldn't be getting the better throttle response on the warm/hot restart?

Overall though.. I wish I could figure out how to make it run all the time like it does for that 30 seconds after the warm restart. You can tell it some how is running differently.

 

Sorry don't understand. Bottom line, with 0x62D set =5, you will have 10 less IAC steps anytime coolant is >= 113*F. You'll need to determine if =0 helps response.

 

Finally had the chance to try this out. Unfortunately didn't make much difference.

Once thing I'm noticing in the data is that either TP is not always reporting when AE TPS and AE MAP are active, or it's not activating all the time with throttle changes. Even when i stab the throttle. With the SAUJP XDF, I do have those flags enabled (report when in AE Delta TPS/MAP). So I'm not sure what's happening there...

 

Hmmm. "didn't make much difference"? So if had "some" difference as opposed to NO difference, there's a clue there.

If AE TPS and AE MAP are being reported even one time in a log, they will always be reported provided the criteria for reporting them is met. Will never have random reporting unless there is an electrical connection issue between the sensors and the ECM. If there is lack of AE reporting when stabbing the throttle, there is something major wrong.

If you are referring to setting the XDF flags beginning with "Light/LED", none of them should be set unless you are using the CEL or a LED to report data via ECM Pin F5. AE-TPS(0x035 b3) and AE-MAP(0x035 b2) are reported via the ADX from 0x035 using "XDF ALDL, Word# 06 Display Word-Reports Various Status Bits" set to 0035.

 

OOOOkay... I'm an idiot...

It was a timing issue all this time, not a fueling issue. It should have been a huge clue last year (post #9 above) when the added timing made a noticeable improvement...

I figured out what's going on.... though what a saga this was to get to this point though...

So what happened was I had swapped distributor housings a little while ago due to another issue (long story) and the housing I swapped to didn't have the lower bronze bushing in it (don't ask me why... lol... I don't know/remember; it was simply in my parts bin and I grabbed it). Any rate, I didn't notice it while setting the axial gap on the gear. But it had enough wobble to it that the cam gear ended up utterly destroying the distributor gear. Fortunately the cam gear came away completely unscathed as did the rest of the engine.

Dulcich and Freiberger on Roadkill Garage would be proud.

(Yes the gear was melonized. It was the factory L98 gear used with GM's roller cams, which came with the ZZ4 I used to have).

Any rate, I put in a new gear and the "throttle crispness" I've been looking for finally materialized.

I had always known my old gear had a little wear on it as it had been on the car since ~1999. But I never really thought it would have such a noticeable impact on engine performance so I never considered it as a possible cause.

I guess lessons learned... even a little bit of wear can cause sufficiently erratic spark timing on acceleration to show up as a very minor tip-in stumble.

Now that I know things are that sensitive to spark timing, I'm probably going to replace the distributor shaft and pickup coil as well. Both also date back to the ZZ4. Figure I'll just optimize everything and get the most I can get out of the ignition system.

Any rate, sorry for the wild goose chase Elky...

(Elky on the left, me on the right)

PS... I guess the only loose thread on this is why the warm startup seemed to compensate for that worn distributor gear. I had always thought it was something temperature related, but I guess not. Not sure it matters anymore though, now that it appears to be resolved.

 

Not to keep beating a dead horse...

I found this parameter with this description in Tunerpro... Is this adding spark timing for a minute or so?

SPARK ADV, Startup SA Decay Degrees .vs. Startup Coolant

Once Injection Delay Counter #1 =0 (Injection Delay Counts from Table 0x1C9), Delay Counter #2 is loaded with 20 injection counts from Scalar 0x1B8. Counter #2 is reduced 1 count every injector firing. When Delay Counter #2 =0, only then can Decay of Startup SA begin.
Decay results from a value being extracted from this Table based on Startup Coolant Temp. This SA Decay value is subtracted from beginning Startup SA until Beginning Startup SA is <=0. Once <= 0, Startup SA is no longer added to Total SA-TDC.
The entire Delay #1 and #2 and the Decay process can take 45-75 seconds.

 

Yes. There is usually afterstart fuel also added. The two are to help prevent stalling right after start up. Some call this choke SA/fuel, others call it afterstart SA/fuel. It is helpful.

RBob.

 

Interesting... thanks RBob.

 

Going back and doing some research on this... Car is running phenomenally, but my curiosity on this is sticking in my craw... lol. Still want to understand why it was different post-startup than during normal operation. The distributor gear issue I mentioned earlier couldn't have been the root cause of it.

Its interesting how the Corvettes and Camaros are programmed differently for the spark timing.

Get a load of how much timing GM put in at low to moderate maps. There's even variation between automatic (ANHT) and 6-speed (AXCN) on the startup advance. Nevertheless, for the first minute or so after start up on the Vettes, you could be temporarily getting up over 50 deg of advance! GM must have considered this safe to do.

I may try it on mine... increasing the low-map timing to the L98 Vette. It's even more aggressive than the '96 LT4 Vette map that I'm currently running and I do know the engine seems to like the added startup spark timing for that first minute or so. If so, the Miniram must like spark timing even more than I originally thought.

The L98 Vette's are essentially running 6-8° more during normal operation than I am. My setup uses S_AUJP startup spark advance with the LT4 timing map, so this first minute or so of post-startup is putting me into the range of the L98's normal timing map. That's gotta mean something.

Corvette ANHT - 0 startup advance after 92C.


Corvette AXCN - startup advance zero'd out after 56C



Camaro (AUJP) - zeroed out after 92C. Still, getting up into the mid-40's on total timing temporarily.

 

The biggest difference between the 87 - 91/92 Y-body and F-body L98 engines are the heads. Aluminum versus cast iron, with different combustion chamber shapes.

Note that even though the SA may add up to 50°+, it isn't what the engine is going to get. The distributor is limited to a max 42° of advance, then the 6° base, so at crank timing won't be over 48°. The code checks for the maximum advance that can be programmed into the distributor, and reduces as required.

RBob.

 

That's true... I forgot about the max spark advance parameter.

So if I'm understanding correctly (which is always a crap shoot... lol) then if you can never get over 48 at the crank, that still means that the ~46 I'm seeing in the main spark table is still being realized at the crank?

 

Yes, as long as nothing is subtracting SA, or any adders, it will be 46° at the crank.

RBob.

 

Thanks RBob

 

Finding some articles discussing timing at cruising rpms/loads.

Seems like 46-50 is not out of the question. I guess it's like Grumpy used to say... "give the engine what it wants".

https://www.motortrend.com/how-to/0604rc-ignition-timing/

https://www.motortrend.com/how-to/set-ignition-curves-create-optimal-performance/

 

Wow.... so in addition to the increased cruise timing, I added 5 deg to my idle timing. Picked up a 0.5" Hg of vacuum and it's noticeable smoother, and even smells better. Must have been some minor combustion inefficiency previously. Very nice improvement.