People usually call it hunting or surging. At startup, mine will vary up/down about 1000-1500 rpms for 15-30 seconds. After that, perfect.

Really, it's not THAT bad or particularly noticeable...except to the perfectionists.

I'm very happy with my overall tune and have this single issue left. Doesn't seem likely there's a 15-sec vacuum leak at startup. Tried 2 IACs, flat timing in the 48, 64, and 96 "load" cells, and different timing values.

Maybe it's just "natural" for a cammed motor? Maybe I need a bit more fuel at startup? (Hate to do that though...who WANTS to burn more fuel?)

Almost everything was replaced with the 383 build.

 

If you have a data log that would help a lot.

 

I do. Whaddya looking for? IMO, not much to gleen from open-loop start-up. Target AFR is steady.

 

I remember looking at your spark table in an earlier thread (pe maf) and thought to myself that it would cause idle fluctuation. The values were different around idle but it sounds like you tried that. Mine will hunt pretty bad if the low rpm/load cells are not even.

Another thought since it does it while warming up would be start up fuel.

 

Without data and tune information to look at, everyone only knows what you say you see. You don't see what is causing the trouble, so you aren't going to say it.

A data log and tune file are worth a thousand forum posts

 

I used to have the lower cells "flat" but decided to go back and try my (self-proclaimed) smoothest bin, with smoothest timing curve. I like it because the car feels/behaves more like a factory car than seething dragster. Power feels (and probably is) more linear too.

I noticed the "stock" AYPY or ARAP bins have a pronounced difference/drop into the lower cells. If that's really indicative of the stock bin, then I wonder why it did not swing wildly from the factory?

Keep in mind, what I've got going on is subtle. I can't post logs because (as Andy knows) TunerPro won't let you do that. They're playable in sort-of-a "real-time" player -- as if you're driving the car around while viewing the logs. All the relevant parms/sensor feeds are available, where you pick any six to view simultaneously.

Not withholding anything intentionally.

Gotta wonder if anyone has an L98 that has a cold-start idle that's "rock-steady"?

 

I got some other log viewing tools. I'm not real big on the playback thing.

I had my brothers 89 with a 422 idling solid on the stock ECU several years back before he switched to a standalone.

 

Probably aren't too many things that would cause it...on an otherwise well-functioning bin.

One thing I notice with my other car(s) is how rich it smells on cold-start days. The vette doesn't smell as rich.

I've considered things like: Increasing crank fuel PW (vs. Temp), Rick/Lean Offset vs. Coolant, Open Loop AFR for temps below 90, Adding more start-up fuel with a slower delay, or decrease/increase spark correction in the sub-90-deg cells.

Not really that many choices....most of which I have tested in the past. Mostly, just programming for a higher cold rpm helped in the past...but I've lowered it again...to 1100 for the startup temps I'm talking about.

Actually, this is slower than my other "factory" cars but faster than the stock AYPY bin. (I have a stick)

BTW...My cam is similar to the LPE 219....with a couple degrees less duration and .015 less lift.

 

Play with your IAC speed, your IAC may be responding too fast or slow causing a surge, try to decrease the IAC stepper motor speed from the stock 16 rev's sec to less and see how it responds, then try more and see how it responds. It could also be too lean or rich or timing changes. Look at cold engine timing adders and AFR with a wideband.

 

Are you talking about the slope parm? Seems like (on the extended $6E mask) that I'd found a way to increase the IAC response time. It helped settle the fast (moving idle) down to it's slower (stationary) idle -- in a shorter period of time.

IIRC, that was measured in rpms/s. And, I increased it from 100rpms/s to 150rpms/s.

I don't remember any parm like you're describing above. But, I'll look again (especially at the extended mask).

 

rotten, poor fitting IAC emission hoses Also have you checked TPS position?

 

Timing tables are the real thing I want to (and should ask about).

As Andy pointed out, my spark table isn't (wasn't) flat across the 800rpm, 1000rpm, 1200rpm cells in the low-load columns. That's where it idles. Since I hadn't look at a tune for awhile, I forgot about the correction table (by coolant temp). It wasn't flat either.

I've made them both flat now and will install the new profile soon.

In the stock ARAP and AYPY (auto/stick) bins for an 89, timing rolls off to lower values in the lower load cells. I'm assuming there's a reason for that. (In another thread, I asked questions hoping to figure out the exact stock spark profile.)

I think it's interesting/informative to see regional table techniques. IOW, what did GM do in lower, mid, high, left, right, etc.. .values of each table. (spark or not). Most of that is straight forward. And, really not at issue now....except if timing was lowered (retarded) to keep the engine cooler at stop lights. (Or maybe to help parking lot trolling with a stick....or any other reason.)

Personally, I'd found higher-than-stock timing values seemed to work best for performance. (This is ONLY in the low, idle cells). In all other cells, less timing was the right solution -- after converting to much more efficient cylinder heads.

More timing in the low-load areas works similar to a bump in rpm speed. It makes the car easier to launch from an idle. It also makes less throttle necessary for the launches. (I have a ZF stick). Less throttle (revving) from a stop, makes driving seem more refined...not like driving a beast. Actually the same is true for my newer "throttled" back spark table. High/mid peformance is the same or better...while lower performance is a hair less snappy. Motor vibration is much less pronounced.

We've theorized the later is actually a product of the headers and power production in the lower rpms -- near resonance of the chassis. I'll also add that the side-pipe system I'm running more directly "connects" the motor to the frame. There are mounting points that are isolated with rubber bushings, but there ARE more contact points.

In lower rpms, it's almost like I was running solid motor mounts. These extra contact points -- combined with the secondard header pulse of Y-config headers -- create a bit more vibration. A bit more low-frequency "tuning-fork-like" rumble throughout my unique exhaust system.

The issue is esoteric...to say the least. It also has to do with my desire for maximum cooling at idle. I want the engine to create the least amount of heat and vibration at idle. I think less timing can accomplish that -- while more timing makes it easier to drive. That's why I'm "playing" with timing again.

I will go back to flat lower timing values and see what happens. Still, I'm left wondering why the General didn't code the low cells flat? Like the PE spark question I asked earlier this year, I'm curious if/what I'm missing? Doesn't mean I don't have a tune I won't be happy with...it's probably best described as I'm the "consummate" tinker-er....always trying to find something better.

(Did I mention I keep thinking about a FIRST or an HSR? )

 

I've seen the surge at idle. In my case it was a small vacuum leak. I also saw a bog at tip-in.

It seems the ECM overshoots the target RPM and then corrects and then overshoots on the opposite side. I think it's adding and removing fuel but going too far. I think you would see IAC counts cycling high and low if this were the case. The injector pulse widths would also be doing the same.

I think maybe an O2 sensor (especially a wideband sensor) that was "preheated" to operating temperature before starting the engine might help zero in on what is happening if it's a mixture problem.

 

The timing table is usually the culprit to idle hunting and surging.

But without data it is hard to do more than grope in the dark.

 

I think vacuum issues are more often the culprit...especially when present after warm-up.

I think there can also be issues from too much timing on a "hot" cam. IOW, surging can be minimized with less timing (unless I'm mistaken). This point should also have been included in the "balance" of items being looked at from my prior post.

When cam profile is part (or all) of the problem, I'm betting a log won't help.

Even if I posted timing tables, I'm not sure that would help either...unless someone knows where/why timing is shaped in the lowest of load columns (e.g., from the factory). I've seen bins from tuners like PCM4Less that are high/flat to the bottom row. I gotta wonder if that runs hotter. Or has trouble "getting down" to the correct idle speed.

Idle "logic" has some very interesting considerations IMO.

 

If your car was working fine and starts idling high or hunting it is often a vacuum issue, I am assuming all else is working correctly and you are trying to tune it, then the timing table is usually the cause of hunting and surging.

When working with a modified engine, what happens is the engine starts going through regions of the spark table that were not meant for idle, adds timing, the idle goes weak and it falls back in RPM again. If you watch in the data you will commonly see it follow a pattern of spark advance going up and down causing the surging. That is why you will commonly see tunes with the timing flattened out to keep it consistent in all idle areas.

You can also set up timing dips to hold the idle from going up into the higher RPM light load areas, thus giving you a much stronger idle.
Here is an example of a spark table on a setup that has no IAC:


Notice the hole in the spark table at 1100 RPM, that basically boxes the idle in and holds the engine from wanting to go over that RPM at idle even if it has the throttle blade idle set higher than it would normally be set, thus gives a very solid idle even without an IAC, yet prevents surging.
This table is an extreme example as it is from a solid roller engine with no IAC and cam duration of 252/263, you wouldn't normally need to do something this extreme, but it serves as a good example of how you can manipulate the idle behaviors with the spark table.

 

Interesting. Having a "low" hole isn't something I would have considered..since the move (out of a low cell) to a higher surrounding cell would create a more pronounced jump in timing. And, presumably, a more pronounced change in engine speed/response.

If anything, I'd think having: more timing in [rpm] cells below, more timing is [load] cells below, less timing in [rpm] cells above, and less timing in [load] cells above the desired load/rpm cell [for idle] would create the effect you described. The result would be "flat" response [immediately] above idle with an artificial bump in timing in lower rpms/loads to boost idle back up to the desired plateau.

Another idea or two comes to mind, which I'll come back and post later....

 

vacuum hoses near and around the brake booster, and to the cruise servo, has little cracks on the ends, the aggregate of them will produce leak albeit small, cut off the ends and your braking power goes up.

 

Banged the drum of equal lower rpm cells for quite awhile (see date of my prior post). Finally, took the queue from the ARAP/AYPY bin and created cells where the "48" load column had lower timing than the "64" cells. Now it's pretty darn stable.

Everything I tried (before) involved moving cells in both those columns to the same lower/higher value. I also made sure the Startup timing table (vs coolant temp) was the same in both of these columns.

Don't know why they need to be different...but it works.

 

Gregg,

Your def file should have a Closed Throttle SA. Are you running Closed Loop? Flattening both Fuel and Spark tables in the idle region really helps.
Smoothing in general makes a much better driving car.