gemellocattivo.com

The MAP gizmo was designed to work on any ECU. The primary output is the lowest reading, which should be the cylinder in intake at the moment. This gives a much stronger MAP signal than you'd normally see on an ITB setup....but it's just a simple analog output so it's not specifically cylinder aware.

I do have it setup with 13 outputs, the main plus each individual sensor....but only the main is connected to the ECU at the moment and I don't have 12 open analog pins...although maybe I could rearrange to get them. I'd like to use the CAN expander for this...i need to get that working first though.

There is a very clear and repeating pattern in the main MAP signal vs the crank position, see log below, I have the MAP range set at 30-50 on the graph, the signal it going about 43 to 47. So, yes I can work with this....but it's not simple. This kind of thing needs to be done fast to be meaningful do it eats a lot of processor time.....but it's no different than what I was trying to do with the NBO2 sensors and have now removed do I could put this in.....that might be the easiest option but I'm not exactly sure at what rpm the processor will run out of time and this would start to delay more important stuff.....probably worth trying though.

I would expect that once you get your idle sorted, you could turn this feature back off for tuning higher RPM.

Or go to a dual ECU, with each one controlling one bank.

There really isn't a way to turn off a feature easily....i can ignore the result but the code will still be running. Load g a new model is the only way to truly shut it down.

2 ecus wouldn't help much either.....90+% of the work the processor is doing is generic and would need to be duplicated in both ECUs so no real time savings.

I had just rewritten th NB code to be a but cleaner and faster...ill turn that into MAP code and thinking about it i can probably slow the loop down ...it shouldn't really matter if it only updates every few cycles.

I think I now have cylinder MAP readings and a fancy new tab in the tuner layed out like the cylinders that reads out each cylinder's %error. The ECU does the work to calculate the MAP valuse, but the math to covert to erro% is in the tuner since the ECU doesn't care abotu that.

I set the angle window width to 30 degrees...hopefylly wide enough to let me readings most every cycle but narrow enough to not mix up the cylinders. I also use a latency angle which is where the cylinder is at peak MAP relaive to TDC....I picked 480 to start but I may need to play with that and the window size to get it working right. Cyl 12 is really just a reference....this is the cylinder the actuator is attached to so I plan set set everythign else to match 12 and the error% shown is relative to 12.

Next time I fire it up I'll see if it works.

I decided to change the way MAP is calculated to make it an average of the cylinder values. The way it was the mixture could and probably did vary quite a bit cylinder to cylinder depending what the map reading was from an what cylinder it was used on.

Best would be use the correct cylinder butthe processor can't really do that.....I don't think.... but I can apply a correction to each cylinder based on its most recent MAP reading.....let me see if this stuff works right first them I'll figure out the correction part.

and seemed to have fixed it so it won't start. I'll change that back and do some logging before making anymore changes like this.

Fixed it really good, then!

Being able to change anything you please in the ECU is definitely a double edge sword.

I pulled the plugs yesterday and 7 and only 7 was clean white....like its not running...maybe wet with fuel. I loaded the test program from way back and it has fuel, no question. I hooked up the timing light but didn't get the engine running to test that...could have checked during cranking but I've checked that many times and really wanted to see it on the running engine.

I did a little cleanup that included changing the MAP back to how it was and adding a new term connected to nothing but the data log for the average I was trying to use so I can compare the 2 and adjust things if need be. Its the weekend so hopefully I'll have 4 shots at testing with a fully charged battery.

oh, and the virtual channels for the MAP error weren't doing anything...they must not be what I thought they were so I added that stuff into the ECU model so they should be working today.

Bad coil? Do you have a test program that lets you spark the coil at will?
Manufacturing error in the plug wire?

Even with a hole in the piston it would still burn *something*.

I have a test program written to spark them, but actual access to the coil drivers is not in the FW i have ...its a newish feature in ELC that hasn't been added to a released FW....I get beta ELC versions so I see it and can put it in the program but no longer get beta FW since they went to a release over the web system I could connect it to a different driver and add a 5V pullup I guess.

I know there is no hole in the piston because I just did a compression test.

It makes sense 7 is dead looking at the lambda data.....but I'm really confused by the crank speed data as 7 is not bad. last run it showed everything looking pretty good but 3&4 dead, like 15% slow at 2000rpm dead and the plugs in those 2 looked awful, all glazed with fuel so I believe they were dead. so WTF or 7?

I have more coils, I have more plug wires....maybe I should just swap them out. When I was looking for something the other day I discovered I have an inline spark indicator, I'll install that this morning and make the R&R call based on what I see after that. But the crank speed data looks fine and 7 isn't next 3 or 4 in the firing order so that data still confuses me.