john banks

I've got really good connections now and get up to 0.86V (only got this once) which I suppose is satisfactory.

But I don't get this repeatedly, and there is no real difference changing back to the original ECU. The car has only done 12000 miles so I doubt the lambda is tired - certainly fuel economy is reasonable and the car runs very smoothly.

Surprising that I really had to be caning it at high revs to get 0.86V, and that is tricky on the road - even though I have a UK gearbox it just flies round the rev counter and you are hanging on for dear life....

Not really a very good way of estimating fueling as the sensor is so far out of its band, but some of you had already said that!

Maybe I'll just get a Knocklink.

[Edited by john banks - 12/1/2001 2:21:51 PM]

john banks

....target zones 26 to 31 - ie lambda targets. Car is MY00 UK spec.

Maybe someone knows or could check for me next time they have their tuning module plugged in?

This will give me voltages to look for from my lambda sensor when I fiddle with the boost given all the worry about full boost at part throttle with boost controllers the I-club guys are having. I am armed with an accurate DVM.

Thanks very much.

GavinP

John,

MRT have some of the default settings on their site (see towards the bottom of this page):

http://www.mrtrally.com.au/performance/brochures.htm

Thanks

Gavin

R19KET

They are lambda voltage targets used by the ecu, if you have the closed loop lambda function turned on.

My advice is to turn this function OFF, and map the fueling manually, zone by zone.

Personally, I wouldn't turn it back on, and I know that Bob Rawle maps the same way (unless he's changed recently ??)

The main problem with closed loop lambda control, using a standard narrow band lambda sensor, is that when the sensor gets hot, it reads lean, and the ecu will add more fuel to compensate. Then you will run VERY rich.

With the Motec, you can program the zones where you do want closed loop, for example, all the cruise zones, but switch off the boost zones.

With the Pectel, you can have the wideband option, and the ecu will compensate for egt', but just the sensor costs about £400.

Unfortunately, this is not an option on the Link.

Mark

john banks

Thanks very much Gavin and Mark. Just the info I need.

To make it clear I am not actually using a Link ECU or even a Lambda link. I just found a particularly useful table in the Link manual for the six zones and thought if I knew the default voltages they were suggesting it would be handy when I fit a boost controller. I know it is only rough with a narrow band sensor and will go off given time and heat, but at least it will read towards lean and give me a margin for safety when I check it - certainly a little better than guessing. I will also record some values before I put the controller on. I just want to know with the boost controller on that I am not running stupidly lean on part throttle as there is a lot more boost than the original (PPP) ECU thought of. I already run 17PSI with the PPP ECU at WOT. My concern is running the same at 1/4 throttle with a boost controller.

Bob Rawle

I still map the same way.

Link, however, have just introduced partial lambda control, now its possible to run a version 5 (MY99/00) Link under closed loop fuel adjustment below 0.1 bar yet still run from the fixed map above that, and also disable lambda control on idle as well if required which makes for a stable idle when larger injectors are used, its also coming on the 44s and MY97/98 ecu's.

Its still necessary to get the whole map right at the outset, oh the Link also work's very well with the wide band sensor that Pat H has mentioned in the past, the Bosch LSM11. This is also used by the Motec M48/M4/M8 series ecu. The M800 uses the wide band LSU 4 sensor which is a 5 wire item, both these Bosch sensors are less than £150 each.

Its easy to control the sensor temperature, all you need to do is pwm the heater voltage against an X and Y axis, an aftermarket extra injector drive would do it for example, if the ecu in use did not support it. So revs and EGT would be a good pairing. Or revs and boost if EGT could not be sensed. M800 ecu is ideally placed for this. (as is the Pectel of course) The standard oe sensor can be controled like this, when doe it works as well as the wide bands with the right afr meter.

john banks

Thanks Bob.

So for reference:
0.2 to 0.6bar boost 0.83 volts = just over 13:1 around 4.5% CO
0.6 to 1.0bar boost 0.85 volts = about 12.5:1 around 5.5% CO
over 1.0 bar boost 0.87 volts = about 11.5:1 around 7% CO

If anyone tries this you need to use a high impedance say >1Mohm digital volt meter to avoid ruining the lambda sensor (ouch) with pretty good accuracy for it to be even rough (mine is +-0.5% reading + 1 digit so could be out by just over 0.01V). This along with all the other errors mentioned above will make it very rough and ready, but will hopefully reassure me.

When I get the boost controller I will post results with PPP and factory boost controller and then different configurations.

R19KET

Bob,

Thought you might be

It's about time Link made some improvements, shame it's taken so long. It should make life much easier.

AFAIK, the LSM11 isn't true "wide band", but is much better. I was under the impression that it was the ability of the ecu to compensate for temp', that was important.

Interestingly, when we were dyno'ing the engine, Dave W. decided to ditch the Motec/LSM11 readout, in favour of the Pectel/NTK Uego (spelling) readings. He felt they were far more consistant.

Mark.

Trout...

I have been experimenting with running lambda ON when cruising around town/motorway etc. As in any given zone, say 325 the fuel required whether in 2nd in town, versus cruising on the motorway.

The problem when in town is that if you are plodding along on the boundary between two zones, you can get an anomolous effect where one zone gets leaner and leaner, and the other gets richer and richer. I have seen zones go from say 96 to 116, which with 740s is a hell of a lot of extra fuel!

Does the new Link chip deal with this problem - or there some flaw in the way the car is being mapped (very possibly - Mr Bodgit).

Trout

Bob Rawle

Map needs to be very much spot on to start with David, then the ecu does not have to adjust much, if that is done then the problem you describe does not become a major issue, still some work to be done on it but, as Mark said, steps in the right direction but no substitute for a correct map.

john banks

Given things again I might have got a Link ECU rather than the PPP. Mods tend to snowball. However, you guys are suggesting the mapping requires a lot more experience if you are setting all the zones manually. What attracted me was the way it was supposed to be able to automap a row at a time using lambda. Oh well!

Seems these narrow band lambda sensors are not very useful then.

Adam M

Bob, you mention he M800 uses a bosch lsu 4 wide band lambda sensor.

Is this a true wide band lambda sensor unlike the lsm 11 and if so, where can you buy them at this price? I was under the impression that true wbl sensors were £350 + vat, and on top of this the wbl option on the M800 was in the region of £500 on top of the ecu. It was this that put me off the thing.

Trout...

If only I could learn to get my map spot on accounting for variable EGT, atmospheric pressure, heatsoak, gear by zone, ambient temperature, speed, etc!

But there again, if I could get it so spot on accounting for all of these compensations then I wouldn't need Lambda ON I guess

Trout

R19KET

Trout,

I wouldn't worry about it too much. It doesn't seem to matter HOW GOOD the map is, the results are the same. That's why Bob leaves it open loop.

Mark.

john banks

Very odd - the highest output I get from the lambda is 0.84V on WOT 1.2bar. Seems a little low to me. This is after running gently to let the temp settle for a while and it falls if you hold it on boost for a long time as you all said. Any comments? Probably I should get it checked on a RR. I am taking the earth from the ciggie lighter - naughty perhaps and giving abnormal results?

Trout...

The value is low - but may be reasonable if you have a very hot lambda probe, if it has been cooled for several minutes then it is too low.

Although, by repute, and I ain't checked this myself - there may be an offset, which you could well be worsening by using a poor quality earth. Realisitically check with connections as close as possible to the sensor for best results - these are very small voltages.

Trout

john banks

I'll wire it better, and also turn down the boost to standard and see the separate effects. Thanks.

john banks

Wired the earth properly and now get 0.82 Volts. If I turn the boost down to 0.9 bar I get 0.81 Volts. Very weird. Maybe my multimeter is not as accurate as it claims. I'm not desparately worried. I find it reassuring anyway that I get the same or higher lambda voltage when I turn the boost up to 1.2 bar.

R19KET

John,

I think you need to cross check the voltage, because if you are getting .82volts, that's VERY lean, and I'd be very concerned.

Either there's still an earthing issue, and you're getting an offset, the lambda is a bit tired, or your fueling needs addressing.

Mark.

Trout...

John,

if you were caning it you would get lower values as the EGT rises the lambda output falls. So you may get say 0.87/0.86 for the first minute or even less and then it would fall away quite quickly to 0.8/0.7 even.

Trout

john banks

With more runs, I am consistently getting 0.84-0.85V at WOT and 17-17.5PSI suggesting ballpark 12.5:1 and 5.5%CO. Reading on I-club with discussions about best power being about here and best safety being even 11:1, with others touting 13:1 and the sensors being very inaccurate, I reckon my figures look OK, and the 0.87V target as standard on the Link ECU is conservative for safety - around 11.5:1. With some cars running 2% CO and not blowing up instantly on the rollers on dyno.scoobynet then I suspect I am OK. Any other comments welcome.

All I can imagine is that if you did extensive tuning with the link in closed loop mode you would be overfuelling like, well, er some standard UK Scoobies! Amazed to see several MY99 on the dyno site with >10%CO. Can't imagine that being fun either.