bd0nalds0n

Well, I spent a little time farting around with the car this Saturday afternoon. I went ahead and tuned it while hot, rather than during the initial warmup. I didn't do a lot of sharkplotting, just a shark re-tune.

The areas I've found most challenging to get right are in the 1800 - 2400 RPM columns and around 70 - 100 on the load rows. Under different circumstances, these are the cells visited at the shift point between 1-2 gear, during mild city driving, and as I roll on the throttle just after the clutch is out, there's a momentary lean spike that creates a jerk.

I'm not sure that the sharkplotter is the best way to tune these cells for smooth initial acceleration, since most of the samples it would take aren't actually in the shifts between 1-2 gear (i.e., acceleration when not much time is spent in a cell but you go across cells/rows pretty quickly), but rather during cruising in higher gears where you're moving between cells more slowly. A lot more samples get taken during the latter conditions than the former. My preliminary and unscientific answer is to tune these cells manually until I get the drivability I'm looking for during open loop (richen up the cells where I'm seeing the spike in AFR), then let the LH figure/smooth it out further during closed loop.

I actually needed to move some of these cells past 127, so I resized the injectors a little bit smaller on the ST (making all the other cells richer) then retuned everything again.

* * *

I had a question for Andrew who noted a lot of false knocks at one point: Andrew, were you by any chance using an inverter to power your laptop when the false knocks were logged?

I used an inverter today to keep my laptop going at maximum brightness (all the other times prior I had used the laptop on regular laptop battery power, but I don't have very good battery life).

Maybe it's in the manual, or maybe it's common knowledge to everyone but me, but the sharktuner didn't seem to like the inverter. It would log knocks at various times, contemporaneous with the fan on the inverter speeding up or slowing down. I had the inverter wired directly to the battery, but figured that it was doing something the ST didn't like, so I disconnected it. I didn't think it would be that big of a deal, I don't think the power supply on a laptop draws all that much current...?

Anyway, I went ahead and swapped back the NB o2 sensor and enabled closed loop and adaptation. We'll see how everything does tomorrow.

The car was idling a little bit high, which I thought might be a vacuum leak, but it's rock stable and spraying starting fluid around the intake didn't do a thing whereas when I've had vacuum leaks in the past, doing so causes a very noticable change. I've been fighting vacuum leaks since I installed the SC kit, so it wouldn't be all that surprising, but I'm not all that excited about taking everything apart--again--and resealing it.

* * *

I'm getting closer to a good spot regarding the cranking pulse witdth and cranking enrichment during starts. I'm using 2.85 warm cranking pulse width and my enrichment across the temp cells is:

Temp (F).......-24.....-12......5.....17......32.....46.....57.....73.....87.....102.....122.... .145.....167.....203.....206
Enrichment.....94......82.....58.....46......36.....28.....17.....14.... 13......13.........5........5........0........0........0



Still interested in seeing what other twin screw, 42lb injector running people are using for their cranking pulse width and cranking enrichment...

AO

Brian-

I was STing today and used a small inverter and had no issues, but I have the ST2, so I don;t know if that makes a difference. Before I got the ST2, I would use Theo's Diagnostic Tester on my old laptop and the same inverter. I got tons of knocks with that combination - we're talking like 1000 knock in a few seconds. For comparison, I was getting 10's today. (Got it down to jsut a random knock here and there today)

But this is great timing. We were discussing at the Frenzy this weekend, about how the SPer does it's calculations. My experience with the LH side is that when I use the SPer to calculate my cell values, I go very rich. I think this is becuase I'm running 42lb injectors, but need to set the STer at almost 19.0 to get a decent idle.

Does injector size somehow get pulled into the calcs? Maybe by looking at Injector pulse width? Just wondering.

bd0nalds0n

Hi Andrew,

Do you have an X pipe with 2 bungs, with the NB on one side and the WB on the other? If so, try tuning with the WB in the hole normally occupied by the NB. Doing so made a big difference for me (although it probably warrents exploration as to why there is a difference between sides).

Also, have you gotten the updated Sharkplotter software? If you're using a SuperMAF, earlier versions of the SP software used stock MAF load values, instead of the SuperMAF values, causing a mismatch between the load column/cells/values being output by the ST, and imported by the SP, adjusted, then copied back to the ST map. The latest version of the SP software allows you to check a SMAF box, so the load scales are the same between the two programs. That might be part of your problems, at least, it was part of mine.

As I understand it, injector size definitely gets pulled into the calculation. It may not be linear, but an injector 2x the size requires 1/2 the pulse width to deliver the same amount of fuel. Adjusting the injector size up or down fools the LH into delivering less (a higher indicated injector than actual will deliver less fuel than the LH expects and vice versa) or more fuel per cycle. So if you're bumping against -127 or 127 in your map, you can compensate by choosing a custom size. -127 means you need less fuel, so you'd choose an injector size larger than actual; 127 means you need more fuel, so you'd choose an injector size smaller than normal, then retune the rest of the map (resulting in reductions in all the cells that weren't maxed out).

The more I jigger with things, the more it feels like I'm flying by the seat of my pants and putting a band aid on a problem I should really be sorting out and fixing properly. But I just don't yet have enough experience to know for sure. PorKen's post about leaning out when heat soaked actually brightened my spirits, because what he described is almost exactly what I'm experiencing. I'm just trying to find a happy medium between the rich/lean at cool/hot temps during open loop, then hope/assume/test that the o2 adaptation is enough to compensate between the 2 extremes during normal driving. I suspect it will be. If not, then I know I need to go back and troubleshoot elsewhere. Experts can feel free to chime in.

Tony

Thanks for the info Brian...Im having a heck of a time getting mine to start now when its hot. Ive had my SMAF checked...bought a used LH...just got a new/refurbished LH...checked the tempII sensor.
Fuel pressure seems to hold between starts, so no leaky injectors as far as i can tell.
The car starts fine when cold.
Its gotten so annoying that to drive the car now i have a switch inline with my fuel pump to depower it when i crank the car to clear it.

Its FINALLY gotten cool enough around here to tinker in the garage again but the issue cropped up quickley. Gassed the car up..came back to start it and it wouldnt. Ever since that time ..oh..8 months ago now it doesnt liek hot/warm starts.

Sorry for going a bit OT here, but i gather as much info as i can when it relates to this.

bd0nalds0n

My warm start issues seemed to be more rich-related than lean-related. You might want to try reducing the cranking pulse width, then richening the colder setions of the cranking map to compensate, while leaving the hotter cells at zero. That would, I think, cause a leaner starting condition when warm.

I'm not sure I understand the relationship between warm cranking pulse width and the cranking map. If I'm at 2.85 on the warm cranking map, and I add 10 to the cell corresponding to the current temperature, have I taken the combined cranking pulse width to 2.95?

jcorenman

That makes no sense, why would it be that far off? Do you have a superMAF? If not are you perhaps max'ing the MAF output?

No, for SP just AFR (and O2-adjust if you are running closed-loop). SP doesn't care what the injector size is, it just makes a proportional change to the map number based on how far the AFR is from the target.
(The ST makes use of injector size to determine pulse-width as Brian correctly described).

Can you send some log files and maps that show what's going on?

Thanks,

jcorenman

There's two issues here: steady-state, and transients. You first need to get the map correct for steady-state running in those cells. It doesn't matter how you do it, i.e. use any gear, drag the brakes if needed, log data for SP or do it live with ST-- you need to run the car in those cells with a relatively steady throttle and get the basic steady-state AFR set correctly.

Then, if it goes lean when rolling on the throttle then take a look at the acceleration enrichment (ST's Fuel-parameter page). SP can't help there, and neither will fiddling with the base maps. If goes is lean with acceleration then add more fuel with the acceleration parameters.

This is going to be completely unpredictable, unfortunately. Some inverters generate a lot of "hash" (in the form of current-spikes) back onto their 12-volt inputs. Depending on where the inverter power is connected, then these current-spikes turn into voltage spikes on the CE panel and create noise signals where they don't belong. The knock-sensor inputs in particular are very small signals and susceptible to interference. Better inverters have better filtering, and are less prone to noise generation. The best option is to power an inverter directly from the battery (with an appropriate fuse), the battery is a stable fixed voltage-source and won't get bounced around by the inverter.

Brian, are you still seeing a large difference in AFR between left and right banks? If that is still around then I think you need to sort that out before digging too much deeper, it's going to confuse everything.

Going back to the running-lean-after-heat-soak issue:

I think I saw the same thing today with our GT. We had occasion to go for a long drive today, started out rich for the first hour or so, then slowly drifted lean (5-7% either side of stoich). I was getting the same results open or closed-loop, so it wasn't the O2 sensors shifting (unless they were both shifting). It wasn't altitude, it was rich when we started, near stoich at 5000 feet and lean when we were back at sea level. Sitting for an hour while we had lunch didn't change much, just somewhat less lean for a while. And definitely not related to coolant temp or whether we were going uphill with heavy throttle or cruising downhill. Aircon was off until late in the day.

However, after refueling on the way back it was back to being about 5% rich... When running closed-loop it was well within the O2-adjust range so for normal cruising the LH just takes care of it (for NBO2-equipped cars).

Which leads me back to fuel temperature. Hot fuel is less dense which leads to leaner mixture, but I don't know by how much. And I also don't know how warm the fuel was, next time I'll bring the IR gun or fix up a thermocouple on the fuel-feed line.

John Speake

Jim,
The reason Adnrew uses illogical injector size in the ST Parameters page compared top the 42# injectors he has fitted is because the S/C setup he is using only has a MAF in one side of the dual intake system. So he is measuring only half the air entering the engine.

A stock MAf is used, not a SMAF.

John Speake

Hi Brian
A note at the bottom of the cranking fuel map states that a change in value of 1 in that map results in a change of the cranking pulse of about 4%. So your change of 10 units would change a standard pulse width of 4,2mS to 5.9mS (for example).

I just measured a change of 10 units on the test jig with your settings quoted, and the change is from 2.85mS to 4.09mS So somewhat more than the 4% quoted.

Bill Ball

Brian reports having issues switching O2 adaptation off, tuning, then having things go awry when switching it back on. So have I.

I read a section of some part of the WSM or other Porsche document about the O2 adaptation that talked about the way it corrects consistent errors from stoich in the AFR over time that gave me the impression this is more than just going between open and closed loop when you switch the parameter off and on in ST. I can't find that document now. Can anyone find this document? On the fuel system monitoring page of ST you see an instantaneous correction, and you see a more stable correction value accumulate over time. I was having problems with this latter value rising to 25% over time and rarely used parts of the map going rich. It seemed to me the O2 adaptation was responding to errors from stoich as you were driving and saving some general correction that applied to the WHOLE map, including areas that hadn't been tuned or might have been tuned earlier.

So, with O2 adaptation disabled am I just running open loop OR is there more to this setting? The car tunes easier with it disabled, and that has become standard ST practice, but it seems to be more stable after tuning to leave it disabled in that I don't see whatever that cumulative O2 adaptation correction is and odd changes in some parts of the map outside of my daily driving around cells.

John Speake

Hi Bill,
The adaptation defeat was added mainly for tuning around idle, where the adaptation loop "fought" with ST Autotune. It may not be necessary to defeat it for tuning the higher load/rpm cells.

The "adaption" is actually two things, O2 loop adaptation and Idle stabilisier loop adaptation.

However being around 25% should not be a problem, the adaption hasn't reached the limits of its range.
The quick movement away from 0% is more likely to be due to the ISV correction than the O2 correction because you have just tuned the fuel map for stoich at idle and cruise.

Defeating adaptation in ST does not switch the O2 loop off with a cat equipped car. It only defeats the adapation .

Autotune when active effectively keeps the LH on open loop so it doesn't fight the autotune. I sets the O2 loop to mid range. When Autotune is stopped, then the O2 loop is re-activated.

dprantl

I just switched over to 42lb injectors this weekend and am having a similar experience. I currently have my cranking pulse width set at 2.80 and my cranking map is starting to look similar to the above. I also have the injector size set to 36lb in the Fuel Parameters tab.

Dan
'91 928GT S/C 475hp/460lb.ft

AO

Sorry I negelted to point that out that my car only meters a fraction of the air (kind of forget sometimes). Now that I seem to have a pretty good tune, I'll take some logs over the next few days and see what the SP come up with for suggested values.

Also, I just wanted to add to Bill's experience. If I tune with the O2 adaption enabled, I see to get more extreme cell values (127 & -127) than when I tune without it.

I left it disabled (and zeroed out) after yesterday's STing on the way back from Frenzy. I'll try re-enabling it now and report my results.

Better yet, I'll data log some now and some after I enable it.

Bill Ball

Ah, so just disabling O2 adaptation leaves the car closed loop. I was getting that confused with Brian's saying he was going back and forth between open and closed loop. For me, leaving O2 adaptation permanently disabled seems to be OK. My idle and cruise are good and WOT is no longer over-rich.

AO

Update on my issue: Ummm... er... well, I kind of didn't put it into non-cat mode. D'oh! Need to fix it and then report back .