bd0nalds0n

I'm not sure if it's my tune that's not yet sufficiently refined, or whether I have another issue to chase down and resolve.

When I'm coming just off-idle, especially between the 1-2 shift, and not immediately so but maybe 2 seconds after, under mild acceleration, I get a bit of a lurch/lunge/hiccup. It's a bit of non-smoothness in my otherwise attempts to roll on smoothly.

This can occur anywhere between 1200 and 2200 RPMs, so I don't think it's likely to be a particular cell on the sharktuner/plotter map, although it could be along a particular transition in the LOAD row, rather than RPM column.

I initially thought it was related to the injectors turning back on after decel, but I don't think it is.

My question is whether, once the throttle body is cracked open and air starts to flow through the TB, whether MAF aging can cause a short spike (or trough?) in voltage that might feed back into how the car behaves?

It could also be that my throttle cable is snagging on something, or the TB isn't opening completely smoothly, causing a jagged roll-on in throttle that manifests the same symptoms. It does seems to be throttle-position/activation related, I just don't yet know if it's mechanical, electrical, or fuel related. Almost like there's a bit of a dead spot, and a little bit more throttle triggers a moment of overcompensation--like a step rather than a slope.

I guess I'll tinker with the throttle cable adjustment to see if that changes the onset of the symptoms.

Any other ideas?

Bill Ball

Uhm, if the fuel tuning were off, you should see it in the AFR. I had some problems with bogging on partial throttle, and it was clearly a lean problem, AFRs of 17-19. That turned out to be due to miswiring the NB O2 emulation to the LH, so I was running open loop all the time. MAF response seemed OK but the AFR would go way lean. Why it did this from just being open loop is not clear to me. You'd think I could tune it regardless, but that's what happened. Once the O2 sensor was wired to the LH, that stopped.

An aged MAF that is otherwise not defective should just give lower voltages, interpreted as lower airflow, hence signaling less fuel needed, leaning the mixture, but Sharktuning would compensate for that. I would even expect the LH's built-in O2 adaptation to take care of that for idle and everyday driving. If you suspect some other problems causing a MAF spike, again, you should see this reflected in the MAF number and AFR.

This sounds like it could be a ignition problem.

Do you still have access to a Sharktuner? I would log the bad area for both fuel and ignition.

928autobahndreamer

Are you sure that this is an engine flat spot and not a transmission issue? If you have a sticky valve can there be a delay in engaging the next gear that can mimic this?

I have a bit of a 2-3 flare shift that can mimic this kind of symptom.

bd0nalds0n

Sorry, I need to update my signature. My auto was converted to a 5 speed. It's actually after I carefully shift from 1-2 and trying to roll on the throttle right after having let out the clutch very smoothly. If I'm driving a bit more spirited, it doesn't happen at all b/c I'm already further into the throttle by that point.

I really don't think it's an ignition problem and am leaning more towards mechanical--that either the TB or the throttle cable is dragging at some point and with a little more pressure "pops" a bit, causing the lurch. At least, that's what it feels like.

bd0nalds0n

Actually, I'm pretty sure it's a tuning issue. I've got some cells adjacent to one another that have big swings in values, which creates a pretty spiky map. I'm going to try to refine over the weekend.

jcorenman

Spikey map bad... But first try logging it with ST while provoking the problem, then take a look with SharkPlotter.

Select SP's "point" cursor mode ("+" button) and then select a point just before the shift, then step through successive points with the left/right arrow keys. SP will show you which cell you are in for each point, and the AFR, and the map value for that cell (click the "x" reference-map button). You should be able to relate that to the point where it stumbles.

Maps should be relatively smooth with changing load, i.e. smooth changes down each column. There can be significant changes with RPM, from cell to cell when moving across a row-- because of intake/exhaust resonance, etc.

John Speake

The SharkPlotter is your friend !

MAF ageing is equally applied across the whole range of airflows. So if it is -5% on calibration at idle flow, it will also be -5% at WOT flow. It will not give strange dips or peaks.

Hilton

When you converted to manual gearbox, did you change the coding plug for the LH to match the transmission?

BBX

Xlot, could you expand on this please?

bd0nalds0n

No I haven't changed the coding plug yet, so the car still thinks it's an automatic, driving around in park. I'm not sure what changing the coding plug would do exactly. I know the autos reduce timing at shifts, but I'm running a custom ignition map anyway. It's on the list of parts to get, I just haven't gotten around to it.

I changed some of the spikier cells and that smoothed things out. So we can rule out the map or a dragging throttle cable. It all goes back to the tune.

What I need to do is drive around and log a lot of samples in stop and go driving this weekend. I also thought about copying the map into Tuner Pro like Porken did and running a 3D smoothing program. For the time being, I'm eyeballing it and doing it manually in the cells/areas where I see the widest ranges between cells. It's worked pretty well so far.

Refine, refine, refine. But I'm really enjoying the fruits of the labor, so it's recently turned satisfying after months of delay and frustration.

Hilton

I thought the auto coding plug also configures the LH to compensate for the drain of the torque converter at idle and just off-idle rpm's?

Someone like John Speake could shed more light on this.

John Speake

The coding plug when set to auto gives an idle speed reduction of 50rpm from the normal 675rpm when the gear selector is put into gear from N or P. This is to reduce the "clunk" as the load is applied to transmission.

mj1pate

Bill (and others); This being the case, wouldn't degrading MAF performance be evident in spark plug condition? That is to say...if spark plug visual inspection suggests a good mixture, then can we resolve that the MAF is still operating as per spec?

jcorenman

Michael, the problem with that is that the LH is busy dialing the mixture back to stoich when it is in closed-loop mode, which is most of the time (assuming a car originally equipped with cats, with a NBO2 sensor). So once the LH has adapted the mixture will be about right-- except for WOT or higher RPM's and loads, when the LH goes open-loop. (The S4's factory tune is pretty rich in those areas, perhaps as a "safety" against aging MAF's).

When you hear reports of cars that have an idle problem after the LH has been disconnected (until it re-adapts), that's likely a MAF-calibration issue.

The other problem is that plugs are hard to read, at least I'm no good at it. The best guide is a WBO2 sensor/display.

Cheers, Jim

jcorenman

Brian, if you have a soldering iron then you can rewire the plug you have-- its just jumper-wires. The info is in the WSM.

The problem with simple smoothing-- e.g. copying to TunerPro-- is that it will change "good" cells as well as "bad" ones just to make them similar. It's much better, in my view, to get the correct numbers into the map-- if the numbers are right, then they shouldn't be spikey.

If you collect enough ST log data then SharkPlotter should get rid of most of the "spikieness" with its auto-adjust-- see my response to Fred in the "Sharkplotter" thread.

Remember that you can also ask SharkPlotter to adjust only a selected range of cells if you just want to "tweak" a certain area-- just select those cells first on the main plot (click the upper-left cell, then hold the shift-key down and click the lower-right corner cell of the range you want to adjust). Then select Tools menu, "Adjust fuel map". You can either specify a target AFR for that range or tell it to use its target-AFR map.

Cheers, Jim