Problem with LS7 low speed clutch-in engine stall

[hapnermw]

I've got a new engine idle problem (E38 ECM) I need some help with. This engine is running well. It starts and idles perfectly. It comes to a clutch-in stop from street speeds and drops to idle perfectly. It cruises and accelerates well. It will run in 6th at less then 1500 rpm with no surging.

The problem is when the car is creeping forward or backward and the clutch is disengaged, the rpms drop to zero and the engine stalls. The attached image shows a relevant portion of the attached log. I've also attached the current tune and an oem LS7 tune for reference. The image shows the car idling and then it is slowly backed up starting at frame 2398. Then the clutch is depressed and the car stops, the rpms trail off to zero as the engine dies at frame 2477. I had flashed the ECM several times just prior to this log so the LTFTs haven't had time to stabilize.

This engine is still being broken it. It has about 100 miles on it. It's what amounts to a race version of an LS7 with a standard deck/bore RHS block (427 ci), ARE 3 stage dry sump, Callies rotating assembly, CNC'ed Brodix heads with stock compression (11.3), Moly coated Titanium intake valves, Stainless exhaust valves, Nextek springs, Katech Torquer cam, CHE rockers, etc. The MAF is stock and the CAI uses the stock MAF table.

The tune is a lightly modified OEM LS7 tune. The only changes are a significant increase in idle air that is required to stabilize startup; idle rpm is slightly increased; the open and closed loop VVE is increased in rough proportion to the increased torque the engine makes over stock - a 10% overall VVE increase and an additional 5% in the 3200 - 5600 rpm band.

I thought this issue might have something to do with CFCO so I tested setting CFCO Min Engine RPM (B2424) to 0 to disable CFCO. This had no affect.

Does anyone have any thoughts about what might be causing this or what further diagnosis should be done?

14544

[swingtan]

Does this occur more so when hot or cold, or is there no difference?

It's possible that you still do not have sufficient idle air flow and are relying too heavily on the corrections.

It's certainly not CFCO, as the injectors are still firing until the engine stops.

I think over all there needs to be a bit more work on the tune to get it to run nicely. There are quite a few things that look a bit strange and for your setup, really don't suit.

More idle air, raise the minimum final timing, fix the idle corrections.....


Simon.

[hapnermw]

Thanks for the response ...

The issue seems to be temp independent. I'll try more idle air and see if this helps.

[hapnermw]

I added quite a bit more idle air (+50%). Most cases now result in partial stall that takes two oscillations to recover stable idle. There must be some other underlying issue.

I will try +20% final timing tomorrow.

I forgot to mention that P0106 is consistently thrown which is why it is set to NOMIL. I had used this cam on a stock LS7 and didn't have this issue; however, now I'm told that the second of the ECM's MAP tests is consistently failing causing the P0106 DTC. I'm told this happens with about 50% of cammed engines. I had thought this might be an indication of an intermittent MAP problem; however, the logged MAP values in look fine.

[swingtan]

Timing may not help, but your idle correction tabled are significantly off. It's probably over correcting and causing the hunt.

Simon

[hapnermw]

I added 12 degrees of advance in the 1280/0.28 area and a proportionally smaller amount going out in the base idle spark table and this along with a 60% increase over stock in cranking idle air seems to have resolved my issue.

Now that idle is stable, I have one other issue.

The engine was briefly on the dyne just to verify things were generally working after the install and a few hours of operation. The engine was running rich (off the richness scale). Subsequent to this I roughly increased VVE in proportion to the increase in the area under the torque curve over stock. The primary changes to this engine over stock is higher VE due to cam and higher flow CNCed heads. The fuel system is stock; the compression ratio and cylinder volume is stock. CAI uses the stock MAF table.

LTFT is currently 14% since the ECM was just flashed.

Since the engine was rich in OL prior to the VE increase, it seems that with the increase in VE, the ECM be adding even more OL fuel. For break-in, OL being rich doesn't seem to be a problem; and, I'm expecting this to be at least partially resolved as the engine learns the tune.

I would like to better understand what engine and ECM factors would be causing rich OL and what are typical strategies for stabilizing them to near stock OL AFR (i.e. I want a safe/conservative OL AFR - not a dyne queen :>). I understand the VVE wideband road tuning process but I don't have a wideband and no one in my area does road tuning (I'm not impressed with the general skill/results of the dyno tuners in my area). My intention is to get things roughly in balance with as near a stock tune as practical and then dyno to verify OL AFR is stable and safe before taking the car to the track.

[Highlander]

Best advice is to post a log file and the tune... sounds weird that you added to the VVE if the car was rich... this is assuming your are running in VE mode, otherwise the maf will take over at 4000rpms (stock) and the VVE will have no effect whatsoever on the fueling.

[hapnermw]

It's running in stock MAF mode with stock narrowband O2s.

[swingtan]

The MAF will work all the time, not just above 4000 (high speed air mode).

To be honest, just blindly adding / subtracting from the stock VVE is not a good method of getting it right. The stock VVE is only dialed in for below 4000 RPM and not a great job at that. Also, when tuning the VVE or MAF, turn off the LTFT's and clear the learnt data, it only makes things more difficult leaving them on. Once your happy with the fueling you can turn them back on if you really want, but I tend to leave them off.

Simon