Idle surge and question on ignition maps values

[tom93vr4]

Sort of two parts to this post (first info, then a question), but to start I am in the process of converting my old SAFC & VPC based VR4 to the Jesters ECU with Chrome V2. If anyone wonders (have not seen many others using the HKS VPC), the VPC is a speed density 'converter' so I am attempting to keep that part operational.

So first part is idle surge - I've seen several posts on this subject so this is for information in case anyone else might find it useful or have comments. Anyway, I have been plagued by an idle surge. Usually rpm just continuously oscillates, often times it's so bad the engine eventually dies at a 'low' spot, but sometimes it might stabilize. Attempted BISS adjustment had no effect.

To make a long story short, I notice that when surging, the engine would operate in 4 cells of the ignition map; 750 & 1000 rpm and 30 & 40 load. I noticed none of these 4 cells had the same timing advance; they are 15, 18, 20, & 24. I changed the 18 to 15, and 24 to 20 (load cell 40 at 750 & 1000 rpm), so now there are only 2 different values.

Guess what? No more idle surge. I still sometimes have an undershoot/overshoot with 'push the clutch in', but it's always stabilized within 3 cycles or so. Maybe this is still related, maybe not - further testing needed. Overall, much much better.

So whats happening? My theory: We know the ECU controls idle speed by adjusting ignition timing - so it looks like the ECU gets confused when map timing values are changing while it's trying to stabilize idle speed. Sometimes the map value will inhibit or reduce the attempted speed correction, other times it may increase or exaggerate it - result is unstable or oscillation of engine speed. It is also possible that by some combination of conditions the engine will settle down into a small enough group of cells (like load 30 & 750 rpm) that it will stay there and thus be stable.

So, now for the second part of this post. While looking into the above idle surge issue I started looking at other parts of the ignition map and noticed what to me are "unusual" values. In the upper load cells, say above 120 or 140, there are substantial timing drops as rpm is increasing. At 1500 timing is 8 deg then at 2000 drops to 4, climbs to 13 @ 3500 then drops to 10 @ 4000. There is another small 1 deg drop from 5000 to 5500. Not saying this is bad, and I have not noticed any strange behavior related to it (not used the car up here much yet either, so ??), it's just, well, unusual. I admit I don't have much experience in timing maps, but the few DSM's & EVO's I have seen don't have this wavy look as rpm's increase in upper load cells.

So, anybody else noticed this before or have any insight on this? Comments, questions, whatever?

[Chris@Rvengeperformance]

Usually the idle surge is caused by a lean area near idle. Adding fuel there probably would've correct it. Tons of cars run stock timing in that area and do not have idle surge.

Areas where the factory reduced timing were either knock prone or caused emissions issues. You can feel free to smooth out certain areas, but you have to imagine all the various conditions the factory had to test in. Like doing WOT at 2000 in 6th gear.

Less timing around peak torque is not unusual or bad. The factory map runs a lot more timing after peak torque because the stock turbos fall off so bad and the factory exhaust housings are a restriction.

The stock timing map actually has great driveability IMO. You don't feel the large changes in it because as you past through those cells the ecu is interpolating between the two rows. It isn't like one moment you are doing 13 degrees of timing and the next ignition window its 19.

[tom93vr4]

Thanks for the input. Regarding the surge, in my case it did not really look like a lean condition. Feedback and Low fuel trims were in the OK range, and AFR was sometimes a little rich, sometimes a little lean, but it did not seem to affect the surge. Sometimes, during a 'surge event' the ecu would run afr from rich (14 or so) thru 14.7 to lean of 16ish and then back down to high 13's (these are all averages - instantaneous afr varies with rpm) without having any affect on surge magnitude. I also think in my case it may likely be related to the VPC - perhaps it's response speed (or something) is different than the stock Karman MAF or the more common speed density setup people use (don't remember what that is arc or maft?).

Might play with smoothing the upper load timing sometime in the future but not now. Like you say, I have not noticed any driveability issues in this area either. Who knows, maybe our engines are more knock prone at 2000, 4000, & 5000 rpm under certain conditions - mystery to me.