Fuelling control... possible mod

slindborg

this may ramble and may get a bit disjointed, but give it a go lol.

Right, can I just confirm that you understand that a non-wideband (narrow band) lambda sensor is "centred" around 14.7:1 and switches between rich and lean for the ecu to dial in on Lambda 1.

I'm assuming the fuel trims are learnt while in closed loop control and then when you go WOT it goes open loop but uses the trim values to keep the fuelling correct..... or have I got the wrong end of the stick there?

now here is my idea.
I got a Lambda kit that has an output that can be setup like a normal narrow band, but centred on ANY afr. So technically you could centre it on say 13:1 for N/A which is almost an ideal power AFR. Would this make the ecu learn offset trim values to make WOT runs the same "richness". I would also guess the IGN can advance enough to burn the extra fuel off properly etc.

discuss....

vocky

sounds a good theory, the only way to find out is try it dunno..

slindborg

indeedy.

Spoke to the gas fuel specialists at work and they say that our system does learn the fuel trims based on how the A/F behaves in closed loop, then applies them to when not in closed loop. So could potentially increase the fuelling to cope with a specified A/F ratio, but not 100% sure on the spark side as thats handeled in the US.

Another idea would be to modify the code in the Lambda controller and slew the centre point over revs so it is 14.7:1 at idle then kinda richens up over the rev range (depending on the MAF readings and so on)

Its almost like a remap/ebay resistor mod, but engineerg properly lol

vocky

maybe easier to just alter the map sensor output dunno..

that was how they got around the bigger injectors

slindborg

dosent the apexi thing mod the MAP output and iirc the speedster chap on .org had issues on a long drive that the map reading would slew too far and fire off a MIL fault.

Ill give the single A/F centre point thing a go early next year and see how it goes (bang probably but thats a good excuse for a proper 2.2 build from the US lol)

rabidh

Sounds like a great idea... Is the kit for doing this available? What kind of price is it? If it handled faking of the 2nd post-cat sensor it would be a major help too!

I'd always assumed that the Lambda sensors were a little more complex than just rich/lean. Are the Z22SE ones just Zirconia sensors as described in the wikipedia link below then?
http://en.wikipedia.org/wiki/Lambda_sensor

I've always wondered about the Z22SE ECU fuel trim... I've seen the short/long term values in OBDII - but does the ECU just use a single fuel trim value that modulates the entire map, or does it maintain an entire map of fuel trim values by rpm + MAP input?

slindborg

normal automotive use Lambda sensors essentially switch between rich and lean crentred on 14.7:1
http://www.lambdasensor.co.uk/main/pic/graph.gif

The bit of kit I got is a JAW lambda kit form canada (google 14point7 JAW etc and get the guys site...) and its got a controlable output that can replicate the narrow band output but centre it somwehere else...(I think the 2nd output "could" be made to track what lambda 2 should be outputting too)

Not sure if the fuel trims are applied to a rpm vs throttle table but it would be sensible if they did as it would allow faster transient fuelling corrections rather than relying on the lambda sensor (its a bit sluggish)

can the OBD gear show spark advance? as if the spark advance dosent move then its a largely futile task 🙁

rabidh

Ahh. After writing that I found some kits on ebay for £130ish - although the JAW one is much cheaper even after you factor in the sensor.

I'm pretty sure I've seen spark advance on the simple OBD I have. If the spark advance auto-calibrates with the knock sensor then everything might just work out - although looking at how basic the lambda sensor is I don't hold out much hope!

these are interesting things about the wideband sensor if anyone is interested:
http://wbo2.com/lsu/lsu4.htm
http://wbo2.com/lsu/lsuworks.htm

slindborg

are the ebay ones the NTK/NGK ones?
generally for £30 or so more you get a prettier device, but I'm a function over form kinda guy (hence the VX lol)

just need to make the JAW up and get a lambda sensor. then try the theory

rabidh

The ebay ones were Bosch - looked ok really:
http://cgi.ebay.co.uk/Wide-Band-Lambda- ... dZViewItem
There were others too. I'm function over form too - but to be honest a whole kit sounds more appealing than trying to wrestle two things through customs 🙂

slindborg

the JAW kit with no sensor was $90... just over £44 at the time..... came through customs with no issues at all. no charge either as the value was so low.
the sensor would be even less value so again no problems.

markv

slindborg said:

can the OBD gear show spark advance? as if the spark advance dosent move then its a largely futile task 🙁

Yes, timing advance is supported on EOBD by the ECU, PID 0x0e in mode 1 (see http://www.verboom.net/projects/speedst ... index.html for the list I got from the ECU).

If it is actually changed, I don't know. Can imagine the ECU has a strategy to back off the advance when it senses knock through the knock sensor. However, running on anything higher than 95 octane fuel doesn't seem to change the power output, so probably it does not advance over a certain treshold?

Mark

zafman

my scanguage reads the timing advance and it goes from -8 ish to about 48 shrug.. depending
on what rpm/load is applied

markv

Yup, the ECU obviously adjusts timing. The question is whether it will advance it more when you mod the enigine and/or put better fuel in it. I think it will not..

Mark

zafman

well i've been running the car on 95ron for a couple of weeks now.when i had 99 in it
i recorded what the timing advance was at certain rpm.I did the same test with the
95 in and the timing advance was exactly the same,so you are probably right Mark.
Didn't notice any loss of performance either.

Yello

Im going for the Apexi SAFC NEO and a Zeitronix wideband kit to monitor and "map" my fueling. But do you guys think I will get problems with the ignition if I dont get a decent ReMap?
dunno..

slindborg

tbh the only real way to find out is to suck it and see.

we have knock control irrc, so should be protected enough. Its only a pain if you need MORE ign advance than its willing to give you.

markv

Yup, the ECU has a knock sensor. I think the most it will do is about 40 degrees advance. However, it might be possible to change this without doing a remap. I'm going to try this somewhere in future, after I have my wideband on the car..

Mark