[LAV1000]

Marvin,
Start with idle stop screws in the the possition you have them now.

Unlock idle advance and see what happens, AFR readings.
You probably need to adjust fueling.

And then slowly work them down in steps like half or quarter turns.
Don't start with bottomed out ones.


Tuning is al about small steps.
Take a tune on which the engine runs and adept it in small steps.

ITB linkage:
BMW also used only cables to link them, nothing wrong with it.

[LPG2CV]

Yes, I forgot to tell you to unlock the advance

However, the advance may well be adding extra torque and pulling the idle rpm high. Turning it off, is going to add extra torque, as when the table comes into play, you will add an extra degree of advance. As a test, can I suggest you leave it locked, but reduce it to 10*, and see if that reduces the idle considerably.

How are the TB's connected to MAP, is it via just one TB? Or, are they linked with tube and a Tee?

In Engine Constants, change the MAP sampling to Minimum, as this may help pull the map at idle a little further down the table.

As you make improvements, move the block of nine cells, so the engine is idling on the centre cell. The reason for this, is the cells interpolate. That is, values for VE are calculated between the cells, dependent on how much rpm/map is traveling from one cell to another. No change in cell values in the block, gives you consistant VE values for getting base levels. You could do the same for Advance, but it is easier to just lock it.

[Marvin2Shoes]

The TBs vac lines are linked via a tee.
There is no idle control at all at present.
As a hypothetical.... I have two NO2C's - if I ran both ecu's I would have two idle stepper drives, two MAP sensors. All I'd need to do is take the output of the crank/cam sensor daughterboard to drive both ECUs. Treating each cylinder as its own single cylinder engine. Sound like fun? It'd fix a heap of issues
I'll try those suggested changes tomorrow (not the hypothetical ones). Many thanks for the help.

[apollard]

As a hypothetical.... I have two NO2C's - if I ran both ecu's I would have two idle stepper drives, two MAP sensors. All I'd need to do is take the output of the crank/cam sensor daughterboard to drive both ECUs. Treating each cylinder as its own single cylinder engine. Sound like fun? It'd fix a heap of issues
I'll try those suggested changes tomorrow. Many thanks for the help.

BMW V12s were run like that with two ECUs. However, as I recall, there was some coordination between them.

[Marvin2Shoes]

BMW V12s were run like that with two ECUs. However, as I recall, there was some coordination between them.

They would be co-ordinated by the common crank/cam sensor. I'm guessing it'd require a double up on air temp sensor. I have no idea how the oil temp would be shared... Maybe feed the values out of one board into the other on an auxiliary pin... It's not like a few milliseconds will affect oil temp/pressure. Both ecu' s would be connected to their respective throttle body so MAP would be simple and precise. As each throttle body has separate tubing, separate pressure gauges would enable syncing of individual idle screw adjustment. Each ECU would control its own idle stepper motor but uncertain how this would be tied together. It could be a complete nightmare to get that to work properly.
I think two cylinder boxers aren't exactly an ideal mix with Speeduino. It seems no matter which way there will be shortcomings.

[apollard]

There are shortcomings with all ITBs. Most use a small balance tube to overcome the idle issues and make balancing easier. My '91 M5 uses this technique, with a ~8mm tube running between TBs. Some have tried blocking this tube (supposedly to eliminate low speed throttle touchiness) and balancing the TBs becomes virtually impossible.

I would suggest a 2-4mm hose run between the ITBs and connect the MAP in the middle.

Also, you are definitely going to need to change the screw setting, you can't change pretty much everything except injectors and sensors and not have to change things. For example, the wife's 2012 X5 uses CAN bus to tell the ECU when the car is decelerating at closed throttle, vs idling while stopped vs idling while creeping forward. I can watch the timing and injector PW change just by lifting my foot off the brake. Engine parameters don't change as fast as the ECU changes the timing and PW. It even changes timing at the same time as the HVAC activates the A/C compressor to make the idle seamless. Very slick, but complicated. So it's definitely communicating. You removed all that, changes to the ITB settings will be required.

[LPG2CV]

There is no idle control at all at present.

So, just to clarify, there are no forms of OEM idle control on the ITB's?

You really don't need to try running 2 NO2C's. There will be absolutlely nothing to gain.

You can balance the TB's by either obtaining vacuum gauges, or you can just run the map sensor from one ITB, and then the other to check balance.

you don't need to close the idle screws completely, just close them both a sixth of a turn at a time each. I say a sixth because screwdrivers often have flats on them to turn with a spanner.

[Marvin2Shoes]

1. Some have tried blocking this tube (supposedly to eliminate low speed throttle touchiness) and balancing the TBs becomes virtually impossible.

2. I would suggest a 2-4mm hose run between the ITBs and connect the MAP in the middle.

3. It's definitely communicating. You removed all that, changes to the ITB settings will be required.

1. All my inline four motorcycles had separate carburettors and I took vac lines from the port side of each carb to balance them (from memory).

2. I currently have tube from each throttle body to a tee and that connects to the MAP on the ECU.

3. I have thought about the underlying issues with two ECUs and that would be a nightmare. Much reprogramming with custom code to get the two systems to run together.

[Marvin2Shoes]

1. So, just to clarify, there are no forms of OEM idle control on the ITB's?

2. You really don't need to try running 2 NO2C's. There will be absolutlely nothing to gain.

3. You can balance the TB's by either obtaining vacuum gauges, or you can just run the map sensor from one ITB, and then the other to check balance.

4. You don't need to close the idle screws completely, just close them both a sixth of a turn at a time each. I say a sixth because screwdrivers often have flats on them to turn with a spanner.

1. The OEM idle controls are stepper motor driven. I can't run two steppers from Speedy.

2. Two NO2Cs would be even more of a nightmare than the current set of issues.

3. If I disconnect the tubing and just run one TB to the map (no tee) while measuring the other one with a gauge?

4. My idle screws are external hex-headed, the same as a bolt.

[PSIG]

Those steppers are descendants of the old GM stepper IACs. They don't draw much power. I'll bet you can drive both of them in-parallel from the one stepper module. Check the ohms of each coil in yours.

They (GM) have coil windings that are about 45 to 50 ohms, so at 12V are only a bit over 1/4 amp each. At 45 ohms with 2 in parallel would be 22.5 ohms resistance, and 0.53 amps total for the module to handle. That's well within the capability of the DRV8825 stepper module. At least the original Pololu version that has 1.5A capacity per-phase without heat sink.

Sounds like some testing is in your future.