NA is best - 3.1L M20 w/ ITB's

[Northern]

My innovate came set up to make the aux output and laptop cable read something weird. First thing I'd do is get LM programmer running and make sure your WB is actually spitting out the right readings to the gauge and MS

[digger]

yeah the PW is not even right for a fully sequential setup with requires approx double the PW over batch

[nando]

Are you sure? When i went from batch to sequential, my base PW did not change - the injectors were still firing once per rev. And if you are firing more than once per rev (not usual with sequential), the pulse gets cut in half, not double.

The scaling could definitely be wrong for the injectors but its probably not from changing to sequential fire.

[digger]

yep sorry its probably not called batch strictly speaking? when they all fire at once every rev, the name escapes me...compared to firing every 2nd rev.

when i did that the PW in the primary fuel map basically doubled (not exact because dead time wont change) , the duty was the same because you have 720* in which to inject vs 360*

[nando]

2 / alt is the same as 1 / simultaneous - its only the timing that changes, they still usually only fire once per rev.

You can of course configure them to fire twice per rev, but its not often done as the pulse at idle gets too short to control properly (esp. with large injectors).

[Raxe]

Yeah that first picture was set up very very wrong, I've spent the last two afternoons fixing it little by little as I figured out what needed correcting.

- Switched the pins on the TPS connector (whoops) and I'm actually getting a decent ADC range now.
- Changed to alpha-n and disabled all of the MAP input.
- Changed the ignition timing to a proper table and redid all of the table scaling for alpha-n.
- Actually wired my wideband to the MS and set it to read from there vs the narrowband O2.
- Mechanically synchronized the throttles a bit better with a feeler gauge and set a reasonable idle but I still need to do a proper vacuum sync now that it runs.

I've been playing around with the VE and AFR tables and managed to get a decent idle mixture of 13.5-14.5 at 1100rpm, I'm still having a hell of a time with anything under load but it's baby steps... I'm pretty much starting from scratch with no experience and learning along the way.

Is this looking better?



I don't think the scaling is off? Otherwise I'm really not sure what to do about the pulsewidth.

[nando]

Probably too much advance for idle - it will make the car drive jerkily at low r3vs because the advance makes it accelerate.

[Raxe]

Probably too much advance for idle - it will make the car drive jerkily at low r3vs because the advance makes it accelerate.

Sweet thanks, 19 was a guess and didn't kill it so I left it alone. I'll try lowering it a bit.

I'm also thinking now might be a good time to give VEAL a try and see what it can do to help ease things along.

[digger]

2 / alt is the same as 1 / simultaneous - its only the timing that changes, they still usually only fire once per rev.

You can of course configure them to fire twice per rev, but its not often done as the pulse at idle gets too short to control properly (esp. with large injectors).

proper full sequential is fire every cycle (2 crank revs /every cam rev), the main purpose is so you can fire with respect to inlet valve motion which happens every 2nd crank rev. the PW potentially changes when you do this depending what you start with of course. 1 big squirt every 720* vs 2 littler squirts every 360*. mine used to idle at 1.8ms now its 2.8ms 850rpm 30lb injectors which would probably be a factor of 2 when you subtract dead time but it wasnt really what would normally be termed "batch" to start with. i dont know what you get out of the box with MS pNP

anyway on with the show...

[nando]

right, what I said was if you fire twice in a cycle (one crank revolution), your pulse width gets cut in half, it doesn't double. I'm not really sure that you want to fire twice in a cycle anyway, sometimes that's used as a form of AE, but even factory sequential tunes only fire each injector once per cycle. Instead you use an injection timing map to trigger them based on load an RPM, to like you say, hit the back of a valve or target an open valve when appropriate.

If he was using 1 / simultaneous (all 6 fire at once) or 2 / alt (basically just 2 banks, but each one still only fires once per cycle) and then changed to sequential without changing anything else (like the # of pulses per cycle), then his base PW would remain the same, because in all 3 modes the injectors still only fire once.

[nando]

Sweet thanks, 19 was a guess and didn't kill it so I left it alone. I'll try lowering it a bit.

I'm also thinking now might be a good time to give VEAL a try and see what it can do to help ease things along.

if you look at a factory ignition map, the idle area is more like 8-10 degrees. A modern BMW engine is like 6.. lol. :)

with the cam you might need a little more, but not 19. Also, with the cam, you will probably find that it doesn't like to idle at 14.7:1. Instead, focus on trying to get a smooth idle with a high vacuum, rather than trying to hold a specific AFR (will probably end up being high 13's low 14's). Should be able to idle at ~800-900 just fine which is a little more comfortable than 1100.

[digger]

mine idles fine anywhere between 15-20, gives a bit better vacuum as you need the throttles to be closed further to pull the idle speed down, in general mine likes a heap of timing at light load no doubt due to the cam and combustion chamber characteristics

i'm sure some other member could share their fuel and ignition MAP and you could scale from there? it would streamline the process

[ForcedFirebird]

Look at others' timing maps in this thread. They are in the ballpark.

http://www.r3vlimited.com/board/showthread.php?t=390283

[Raxe]

Look at others' timing maps in this thread. They are in the ballpark.

http://www.r3vlimited.com/board/showthread.php?t=390283

This is great, thank you. Hadn't seen that thread before.

mine idles fine anywhere between 15-20, gives a bit better vacuum as you need the throttles to be closed further to pull the idle speed down, in general mine likes a heap of timing at light load no doubt due to the cam and combustion chamber characteristics

i'm sure some other member could share their fuel and ignition MAP and you could scale from there? it would streamline the process

Good to know, I'm finding it's running too retarded at light load 2500-3000rpm right now (evident by the super hot exhaust and afterfiring), yet the numbers apparently work well on less modified engines. More cam = more advance?

Any tips on how to scale for ITB's? Most of these boosted maps start at 10-30% load and go past 200%, obviously I need to start at 0% and end at 100%. I've read that the first ~30% is more important than the last 70% because of effective WOT and lack of IAC, and that they need relatively little fuel at idle vs much more as soon as they open.

[digger]

In general big cam with overlap has more contamination of exhaust into intake due to reversion at low load and rpm so it burns alot slower therefore you need to light the fire a little earlier. of course this assumes the setup is correct and if you ask for X* you are actually getting X*. If the thing is offset then..........

The other thing is if the throttle isn’t sync'd then a couple cylinders could be running too lean and this slows the burn down so it happens very late and can even continue into the exhaust. if you get lean enough it will still be burning by the time the inlet valve opens and you can get backfires and popping in the inlet.

You should invest in a sync tool they are so easy to use rather than screwing around with pressure tapping’s etc that require tiny restrictors to dampen all the pulses and it tells you actual airflow instead of vacuum



yeah the problem with ITB is the fuel curve is very steep at low throttle position so you need more load point down there.

also this thread is back alive, Fporro has a map that might interest you there

http://www.e30tech.com/forum/showthread.php?t=33028&page=12