ITB & idle changing - Question about stability

[Yves]

I run an ITB crossram setup on my SBC. the trumpets are above the valve cover on the opposing bank and the TB's are paired. So 1+3, 5+7, 2+4, 6+8. The throttle linkage is between the paired TB's so in the middle of the valve cover approximately. There is a connection piece that connects the paired TB's of each bank and each common TB shaft has a lever that has a metal rod attaching it to a sort of cross on the top and in the middle of the crossram. This cross is connected to the throttle with a regular throttle cable.

I also have a vacuum chamber below the runners, with a drilled 5 mm hole in the runner connecting the runner to the vacuum chamber. The vacuum chamber is also connected to the IAC, so I have some way of regulating idle speed.

Setting base idle speed is done by screws on the lever that open the bank more or less.

One problem that I'm experiencing is that after sitting when warm, the throttle opening increases due to the lever become a bit longer (alu) and the rod expands less, so it pulls the throttle open a bit.

ITB's require balancing. This is especially difficult since the TB's are paired, so I can only balance one pair vs the other and so on.

Now here is the question : I have always had the throttle open approximately 1% and had the IAC deal with idle speed. This has worked so far, but it has one draw back and that is that not every throttle seems to flow the same amount making it difficult to maintain a stable idle speed and AFR.
I was thinking of shutting the throttle entirely and having the IAC take care of all of the idle air the engine needs.

Any thoughts ?

Thanks in advance.

[lutorm]

I'm in a similar situation, but I went the other way: I adjusted the throttle linkage to get even MAP with the throttles "completely" closed. I have a very even MAP when idling, but because a completely closed throttle is resting on the stop and not held by the linkage, you tend to unload the linkage a bit in the closed position. The result is uneven response when you crack the throttle, since there's some play that needs to be taken up. So now I have AFR excursions and uneven MAP between the banks in very low-throttle driving instead. I'm thinking it would be better to adjust it the have even MAP at slightly-above-idle throttle instead, because I think an uneven idle will be less noticeable than uneven off-idle.

I'd question a mechanical setup that can't synchronize every butterfly separately, since it's so ridiculously sensitive at idle.

[jsmcortina]

I concurr with the last comment. You need to be able to adjust each throttle and they need to be off the stop slightly. I'd run as little idle-air as possible and set your slowest idle speed on the throttles themselves. That should make them less sensitive.

James

[ertert]

I'm tuning idle on a vtwin with individual linked butterflies at the moment and I found a pretty good video with some tips for ITB's:

https://www.youtube.com/watch?v=H-zHSvZ7PSg

Its from motec but they don't talk about anything too product specific and focus on general idle tuning theory and methods so pretty transferable to our ecu.

Everything they talk about is covered in our manuals but I found the same concepts explained form a different perspective really helped improve my understanding.

Basically for ITB they suggest:

1) Use low ignition timing to reduce idle speed
2) Then open butterflies to lift idle speed

Basically you get more air in for the same idle speed. More total air in means the differences between butterflies matter less. Its not perfect and there are downsides to it that they discuss in the video but it could be worth trying if you get really stuck. I think the whole video is worth a watch but you can skip to the 25 minute mark for the ITB part.

[Yves]

Thanks for the comments so far. I will view that video. I agree it looks quite interesting.

I agree that it would have been better to have each throttle individually adjustable, but this is not the only system that works like that. For instance the Jenvey dcoe type throttle bodies (mine are also dcoe types but not so large) have a common shaft on a paired set. Air can be individually adjusted by way of an air screw that is not present on mine.

I was not aware of senstivity issues that would arise with setting the throttle in closed position and the following sensitive at very low throttle openings. Have to think that over.

[whittlebeast]

I am tuning a big block ITB setup right now. It appears that even expansion of the intake from heat is an issue. Most of the aggravation will be on tip-in.

Andy

[Yves]

I'm tuning idle on a vtwin with individual linked butterflies at the moment and I found a pretty good video with some tips for ITB's:

https://www.youtube.com/watch?v=H-zHSvZ7PSg

Its from motec but they don't talk about anything too product specific and focus on general idle tuning theory and methods so pretty transferable to our ecu.

Everything they talk about is covered in our manuals but I found the same concepts explained form a different perspective really helped improve my understanding.

Basically for ITB they suggest:

1) Use low ignition timing to reduce idle speed
2) Then open butterflies to lift idle speed

Basically you get more air in for the same idle speed. More total air in means the differences between butterflies matter less. Its not perfect and there are downsides to it that they discuss in the video but it could be worth trying if you get really stuck. I think the whole video is worth a watch but you can skip to the 25 minute mark for the ITB part.

Just went over the ITB part. Very interesting indeed.

I must say that I'm in the area where large camshaft meets ITB's, so the especially difficult part to get right. Cam is a 250° at .050" lift with an lsa of 106°(roller).
I had it as low as 12° of ign timing at idle, but the issue is you also loose a lot of throttle and you cannot just drop in a lot of idle all at once when you go of the throttle stops.

I have been wondering why I always saw airflow go up when ign timing was less. Now it hits, me, the IAC opens more and lets in more air. Ok lesson learned.
I also saw him mention something about overlap letting more air through into the exhaust which subsequently reads lean, but is not necessarily what the engine sees. I stayed out of the +1 lamda range because it seems very lean, but this might not be as problematic as I was thinking. So I might have had it too rich. And this engine certainly doesn't like to run rich.

[Yves]

I am tuning a big block ITB setup right now. It appears that even expansion of the intake from heat is an issue. Most of the aggravation will be on tip-in.

Andy

possibly. Hard to be sure, but in my case something is changing due to heat when starting up when hot.

[lutorm]

AFR issues at hot idle relative to cold idle can also be from the MAT/CLT curve. I've been tuning this curve at idle values by making it so the AFR doesn't change as the engine heats up. If the engine riches up as it heats up, that indicates the curve has too much MAT contribution (air entering has lower density than the ECU thinks) and if it leans out it has too much CLT (air entering has higher density than the ECU thinks).

If your idle pulse widths are small, it could also be your dead time is wrong.

Of course these should affect all cylinders equally, so if it's just one bank relative to the other, it seems unlikely this is factor.

[Yves]

I tested my dead time and the numbers used are correct.

I thought that the MAT values were calculated based on the gas laws. Why would the standard curve not be correct ? Not that I ever tried to change it.

[lutorm]

I mean the table that decides how much the intake air heats up above the MAT before getting into the cylinder at low flow rates, not the air density vs temperature curve.

[Yves]

I agree that air heats up on it's way into the cylinder. But this does impact the density of the air.
However, barring the warmup where AFR's are compensated by the WUE, I believe that at normal operating temperatures this should have been incorporated in the VE numbers in the table. Am I right ?

[lutorm]

I agree that air heats up on it's way into the cylinder. But this does impact the density of the air.
However, barring the warmup where AFR's are compensated by the WUE, I believe that at normal operating temperatures this should have been incorporated in the VE numbers in the table. Am I right ?

True, but only at a single temperature. If we assume for a moment that air at idle has the time to completely reach CLT during intake, and you tune the VE to get the correct AFR at a canonical operating temperature, the engine will run lean during warmup. You can compensate for this to some extent with the WUE, but that applies to all MAP/RPMs, so if you use that to set the idle AFR, the AFR during non-idle, when the air flow remains at MAT temps won't be correct.

Also, unless your engine keeps operating temperature perfectly, it won't work. I see CLTs from 80C (when the thermostat opens) to 103C (when the fan turns on) depending on operating conditions, and normally you'd have zero WUE for all those temps. That's still a 7% difference in air density.

[Yves]

Can you explain how you tuned that curve then ?

I also have the following question : in the video it was explained that more ign timing at idle does make the idle rpm go up. This only seems logical. Due to the engine creating more torque from the same intake charge, it needed less air to do so. Hence, why I saw my maf readings go down at the same time. Can we then conclude that the best idle ign timing would be where there is no increase in rpm and no decrease in MAF anymore ?
The whole video also has me wondering about their statement that on a large cam you would need a leaner mixture because the O2 is not reading the actual exhaust gas AFR but the air that is drawn through the engine because of the increased overlap. This points in my understanding to a situation where overscavenging might exist. How then do you tune the AFR. I mean a +1 lambda seems counterintuitive.

Also, with regards to the above comments, when letting idle depend solely on the iac with a baseline opening of the iac below which the iac cannot go, would it help to crack the throttle open just a tiny bit, like .1% or so to prevent the lag in reaction when you crack open the throttle when accelerating ?

[lutorm]

Can you explain how you tuned that curve then ?

I don't have a 100% good explanation, I'm still playing with it, but my approach is to attempt to tune the idle VE as soon as the WUE has gone away, when the CLT and MAT are as similar as possible. In this situation the MAT/CLT curve has no effect. Then, after the engine has warmed up, any change in the AFR from the temp you tuned it to must be due to the intake air not being the temp the ECU thinks it is, which can be fixed with the MAT/CLT curve. It should then be possible to adjust this curve to get the AFRs back to what they were when you tuned it.

Since you don't normally run without WUE until the engine is substantially warmer than any reasonable MAT, this won't reflect the full effect, so it's probably necessary to iterate on this a few times.

But to do this you need to be able to trust your idle AFR, so if you have an overlap cam like they talk about in that video then it probably won't work.

[Yves]

I did a test today with the throttle completely closed. My IAC was not able to provide the air that the engine needs to idle well. So I had to open them again to .6%.

[Yves]

I was balancing throttles today and finally got them to be almost equal (on the flow meter there is les than one stripe difference which according to its manual is within tolerance).

However, based on my individual ego correction (O2 for each cyl) I still see the ecu making corrections. I find this odd. Any ideas what can be going on ?

[lutorm]

I was balancing throttles today and finally got them to be almost equal (on the flow meter there is les than one stripe difference which according to its manual is within tolerance).

However, based on my individual ego correction (O2 for each cyl) I still see the ecu making corrections. I find this odd. Any ideas what can be going on ?

One possibility, that I assume you've already checked, is that the mechanical linkage isn't identical so they don't move identically.

Barring that, it could also be differences in airflow. Since they are synchronized at idle, they may still not flow identically when not at idle. Are all the intakes identical in geometry? I'd also assume slightly different compressions, valve lifts, etc, would cause small differences in VE between the cylinders that you could see if you measured them all. Maybe you can use trim tables to correct for this?

[Yves]

I checked this only at idle. Not sure about the balance when driving. The weather doesn't allow me to drive at this moment.

I think that it might have something to do with map variations per cylinder and the fueling strategy being ITB where in the lower load regions of the table it runs fully on map.
The map is highly volatile due to the camshaft involved.
I'm starting to wonder if AN fueling strategy might be better suited.

[lutorm]

I checked this only at idle. Not sure about the balance when driving. The weather doesn't allow me to drive at this moment.

I think that it might have something to do with map variations per cylinder and the fueling strategy being ITB where in the lower load regions of the table it runs fully on map.
The map is highly volatile due to the camshaft involved.
I'm starting to wonder if AN fueling strategy might be better suited.

If it's due to the map reading being bad, I'd expect the EGO correction to random walk. I guess you didn't say if it just settles on nonzero correction for each cylinder or if the ego correction walks around randomly?

You'd have to go to pure alpha-n to not make the MAP reading matter at all, which I think does not work very well. The MAP measurement is really sensitive, if you idle at 60kPa then a MAP error of 0.6kPa would result in a lambda error of ~0.01 (although if it was random it would be different every ignition and I wouldn't expect the O2 sensor to pick that up).