Lock & Load

I will be installing my CZ unit today with the latest map downloaded and installed by Jeff( mazdamaniac ) so it will be interesting to see the results .

B...free
michael

r0tor

iTrader: (1)

I thought he said that the REVi significantly leaned out his car?

Nemesis8

iTrader: (3)

No, the CZ leaned out my car. That chart was with the CZ pulled out. I wanted a base log for the REVi.

r0tor

iTrader: (1)

i thought when you went from a cz to cz+revi you had to readjust things alot because it went lean?

Nemesis8

iTrader: (3)

I can't remember - I'm starting to lose my memory.

Anyway, I'll have to re-tune the map because of the new flash I will get this week. When SCO gets the knock-retard sensor in the sCANalyser software update, I will be one happy camper when it comes to tuning the upper end.

Rasputin

The 0.75 Lambda value seems totally possible. I've measured that kind of value too in WOT third gear acceleration, 20°C ambient. You will get catalyst protection overfuelling as soon as the catalyst temperature approach 900 °C. It looks like the maximum acceptable catalyst temperature is 960 °C on my European RX8. In line with what I was told by catalyst specialists last March.

I've sen Ford engines with even worse (richer) lambda for catalyst protection back whern I worked there in UK (due to closed-coupled cat). I remember a 9.5 AFR!!! on a South American CE14 Escort engine. Totally unjustified though (a calibration error I think).

And I totally agree with Hymee, LAMBDA is the parameter that should be used in lieu of AFR as the stoechiometry AFR changes with different blends of fuel and with octane rating (different Hydrogen and Carbon content). The PCM actually uses Lambda as a main input, not AFR.
If you are more used to AFR, no problem but make sure you know the exact stoichiometric AFR of the fuel you're using when you're programming your PCM or Piggyback. And keep an eye on the catalyst bed temperature as above a certain temperature, the cat dies instantly.

And the figure you posted is definitely a switching HEGO type of sensor signal, not a wide-band UEGO as used as main sensor in the RX8 (the second sensor, after the cat, is a switching HEGO sensor). Most cars use standard HEGO sensors but UEGOs seem to become more popular now with new engine management techniques like DI and Lean-Burn... It was first available in production on Honda V-TEC engines in the mid-90s.

Fabrice (back from holiday)

r0tor

iTrader: (1)

look at what real AFR's look like from dyno's...

https://www.rx8club.com/showthread.p...ght=dyno+sheet

https://www.rx8club.com/attachment.p...chmentid=54799

ect ect...

they are not in the 11's....

zoom44

the second one is saying 216hp where is the thread that is in?

the first thread the last dyno- 163 on a dyno jet means there is something wrong with the run. plus the a/f on that run never gets below 13 so that cant be right. its also very smoothed

also you will get different readings from a sensor in the back and a sensor up at the fron of the cat. im not taking sides mind you i just have issues with those to dyno charts

here's my recent chart from scanalyser

dannobre

iTrader: (3)

I think you will notice the flat spots at the bottom of the valleys.....and at the 85K+ area...the sensor is maxed out in those areas...or the PCM doesn't read higher than that.....

Not relavant I know...we don't want it that rich anyway.....:D

Nemesis8

iTrader: (3)

Dan, bring your new hardware down this weekend and lets do some testing and tuning

zoom44

yes we've talked about that before. maybe i am not understanding what rOtor wrote when he posted

dannobre

iTrader: (3)

Just arrived 10 min ago :D Will see how it works !

r0tor

iTrader: (1)

my point is show me any NA renesis that runs at or below .75 lambda read by something else other then a scanalyzer (like a dyno or another WBO2 sensor)...
because I have yet to find a single one, yet everyone using the scanalyzer is posting extremely low lambdas/afr.

Only FI cars are in the 11's and thats because they had to use a piggyback to get that low.

r0tor

iTrader: (1)

<- beats head against the wall

for the last time, the only afr ratio you are ever going to see displayed from an instrument is a constant x lambda. I can show you guages from the same manufacturer that read in lambda or in afr - they are the same damn guage except for the scale written on the guage face. So in all reality it doesn't make a single difference if someone gives you an afr from a guage or a lambda from a guage because the afr was derrived from a constant that everyone in their brother uses for gasoline and therefore makes no frikkin difference for tuning.

I don't care about slight chemical variations in the gas formulation effecting the tru afr because the damn guage isn't smart enough to correct for that - your just getting lambda given to you in a different scale!!!

dannobre

iTrader: (3)

For me the ONLY thing important is that the engine doesn't knock with the A/F, or Lamba that I tune too.

I could care squat what the # is...A/F...or Lamba...as long as I can measure a repeatable value to tune too.

The problem I see is....does the PCM dump raw sensor data from the sensor...or use a corrected value based on the lack of linearity of the stock WBO2 sensor....

So far I can only see that Canzoomer found the Lamba and A/F values out of the scan tools WAS NOT linear in the areas I want to tune too.....and he applied adjustment tables based on comparrisons with a corrected Lamba meter.

Until I see anything to the contrary.......the output is NOT LINEAR...and both the CanScan and Scanylyser are uncorrected in the areas I want to tune too

So be careful out there guys...

This weekend Nemesis and I will post runs of lamba values from both sensors...and see how they compare to the corrected values from the CZ software. I just wish my WBO2 meter was here :D

sco

See my graph earlier in the thread. I posted current vs lambda as reported by the PCM.

Why does the conversion need to be linear? I've never seen an O2 sensor that returns linear results. The graph I posted matches the one in the Mazda workshop manual and also matches the graph for other WB O2 sensors I've seen.

dannobre

iTrader: (3)

Sensors are adjusted to be linear so the results of the guages/ outputs are RELEVENT and REPEATABLE...so they can be calibrated between brands and units

This goes for any guage/sensor in any industry.........

The difference between .75 and .76 Lamba should be the same as the difference between 1.15 and 1.16 Lamba.....


I think you will find that commercially available WBO2 meters use software calibration to achieve this linearity :D.

otherwise why bother ?

sco

The results can be relevant and repeatable without a linear conversion. What needs to happen is a certain current reading from the sensor always returns the correct lambda reading. What's important is that the lambda reading is correct. If the lambda reading is correct then sampling in the same way with a different (correct) sensor should produce the same results. A linear conversion is not required to achieve this.

The graph I posted is of a capture I did over a number of minutes.... it was a drive through city streets. Lots of acceleration and deceleration, some WOT some not, and some closed loop. Notice how clean the graph was despite readings in different scenarios. This shows that the conversion was consistent.

dannobre

iTrader: (3)

So how does a non-corrected sensor output the correct Lamba if the output is not correlated to the current output ?? I must be stupid or something cause I don't get it...............:D

I can see how you would always recieve the SAME Lamba from the SAME current without a correction...just not the CORRECT lamba???????

rAte

Sorry to jack the thread, but this discussion reminds me a little of the V!shnu VS Dyn0flash threads on the EVO boards.

It is really funny...

Great info from both sides BTW!

sco

Using a lookup table... in the case of the RX-8 it is using a lookup table that matches the graph I posted earlier. So what it does is get a current reading. It then looks up a table that says for current reading X, it means lambda reading Y. It doesn't have to apply a linear formula. It could be applying a formula, but that formula is not linear.

dannobre

iTrader: (3)

This is too funny :D symantics it is :D I wasn't implying that there would be a formula....just a data conversion :D

So...does the " lookup table" correct the values in the PCM..and supply corrected values to the output of the scan tools???
Or are the outputs uncorrected current values from the sensor as they appear to be...as you said yourself that they mimic the sensor output expected from the factory service manual??

sco

I don't understand what you mean by "correction"? What gets "corrected" and why does it need to be "corrected"?

The data reported by the PCM and in accordance with the OBD spec is sensor current reading as well as the corresponding lambda value.

dannobre

iTrader: (3)

OK.....

If Current is not linearly related to Lamba...

Then......... raw current is not able to be related to Lamba without " conversion"???/

That's all I'm getting at..........

sco

Correct... hence a non-linear function or lookup table is used. That conversion function looks like the graph I posted.