[MikkoR]

Ok, finally had time to continue this project. Last August I got it running, very shortly, but running anyway. Currently there are cooling system and exhaust installed, so I could try to get it running better and, after all, to idle well.

After one burnt fuse and some trial and effort the engine starts, but idles about 2700. I had to give quite/too much fuel before it starts. I'm pretty sure that ICV setup is not correct and it is not closed when cranking and "idling". I have a 3-wire Bosch pwm valve, typical for BMW. I think when idling the active VE table cell should be around the bottom left corner. Mine stays about 50 kPa, and 2700 rpm, and the cell is in the middle of the table.

Does anyone have a working setup for BMW M20 ICV? Years ago I had a MS installed to M20, but I don't remember any more how it was set up...

Log attached if someone could give a look. At the end of the file is the section when the engine runs, marked as "Run". The ICV curve in the log is quite reasonable, but I think there's something wrong with the setup.

Edit: Is there a way to test the valve? + Added msq

[MikkoR]

Forget the latest post. I had some mistakes in idle setup, it was doing just the opposite to which I wanted . Now I understand how the ICV reacts. Anyway, I don't know if the frequency is correct, or which it should be? I have still open loop as suggested in the first phase.

It starts quite nice. Idles better, but still around 1200-1300 and cannot get it lower. Exhaust smells (and looks) quite rich, advance shoud be somehow in the ballpark. Also decel fuel cut off enables sometimes, I have to switch that off.

I have wideband installed but not yet connected to th Speeduino, maybe it will give some more info.

Which should be about the correct vacuum at idle?
Correct ICV duty cycle at idle, approx.?
How to get the idle to the level of 750?

[MikkoR]

Any good choises for an external MAP sensor? Two reasons:

- The integrated sensor quite small inlet, hard to get vacuum hose tight enough. All outlets in the manifold are bigger.
- I'm planning to locate the speduino in the trunk. May be too long distance for proper response, although by the laws of physics the vacuum should be immediate.

I'd prefer a shorter hose and do the needed wiring.

[PSIG]

- The integrated sensor quite small inlet, hard to get vacuum hose tight enough. All outlets in the manifold are bigger.

Just use a hose union adapter to the smaller hose.

- I'm planning to locate the speduino in the trunk. May be too long distance for proper response, although by the laws of physics the vacuum should be immediate.

I use semi-rigid nylon MAP tubing (about 8AWG/ 3mm diameter) and put it into the wiring harness like a wire. No "hose" look, and it couples directly to the MAP sensor with a piece of silicone hose. There are MAP line kits like this from vendors with case bulkhead fitting, couplers, clamps, etc.

Semi-rigid_MAP_line_kit.jpg (73.08 KiB) Viewed 176 times

Wired MAP or on-board is up to you. I like on-board as it's cheaper, much easier, less prone to electrical noise, and testing shows it works even at longer distances (30M roll of hose). Do your thing.

[MikkoR]

Ok... Problem is that I've not found any hose which will fit tight enough to on-board MAP sensor, neither any small enough clamp to tighten the hose.