Fury internal lambda inaccurate reading

[warmup]

I found that Fury's internal lambda controller reading is always 0.5 AFR leaner than LM-2 LSU4.9 and AEM UEGO LSU4.9.

The test used two LSU4.9 sensors for 3 devices, sensors were cross swapped during test, same result, updated Fury fw. to 5.4.3, same result.

LM-2 and AEM read consistently 0.1 AFR diff.

I also compared LM-2 to several other AEMs(no pictures), very same result. I think LM-2 is something can be trusted.

Any suggestion?

 

 

Edited August 29, 2015 by warmup

[mapper]

First, check all ground points and get sure, that all devices have the same ground. 

Second, i found the AEM widebands are very inaccurate. The read on low AFR values much too rich and on high afr's much too lean. And there is not seldom more than 0.5 AFR difference. 

I always calibrate them against my Dynapack wideband (MOTEC PLM), which I think is very trustworth. 

 

[Dave Kriedeman]

Hi Warmup,

when the wideband is up to running temp and reading, if you hit the R key for Runtime Values and go to the ECU STATUS TAB.

In the bottom right corner is there any errors etc from the Internal Lambda 1 data.

Is all the data coming through OK.

What are your settings under Internal Lambda 1 SETUP in ECU SETTINGS.

Do you have your settings setup as,

LAMBDA SENSOR CONTROL = ON

RUN WHEN STALLED = NO.

Are you running 5.4.3.2630 firmware and have updated your ecu with this FIRMWARE.

Regards

Dave.

Edited August 29, 2015 by Dave Kriedeman

[warmup]

Hi guys,

Thank you for you replies.

Mapper: for the three devices in this test, they are all working in "digital" way, so I guess analog signal does not matter that much, I grounded them well though.

Dave: Yes, I'm running 5.4.3 2630 firmware. and yes, the two settings in "Internal Lambda 1" were set exactly like what you posted.

LAMBDA SENSOR CONTROL = ON

RUN WHEN STALLED = NO.

Runtime value:

Int. Lambda 1 Status = Operating

Int. Lambda 1 Error = OK

Int. Lambda 1 Temp = swinging around 780.

 

To Mapper:

AEM is much better than before since LSU 4.9 upgrade, what I have seen is that AEM(LSU 4.9 Version) reads leaner only when actual AFR is higher than 14.2.

I think that range is less important.

 

[Dave Kriedeman]

Hi warmup,

I forgot to ask, where abouts are the comparison sensors mounted, in relation to the onboard LSU 4.9.

Regards

Dave.

[warmup]

Hi warmup,

I forgot to ask, where abouts are the comparison sensors mounted, in relation to the onboard LSU 4.9.

Regards

Dave.

 

Something like the attachment.

10 o'clock and 2 o'clock at the same distance to the turbine housing.

Edited August 31, 2015 by warmup

[Simon]

When we tested here with ours, LC-1, LM-2, Motec LTC, New NTK and our Allen four gas, they were all basically identical to each other.  I would be suspicious of the wiring or install.  Especially if there is a constant offset over the entire range.  All controllers should read lambda 1 at the same point, there is no excuse for not getting that right.

 

[Adamw]

I maybe stating the obvious, but have you checked all devices are set to convert to AFR using the same stoichiometric ratio (14.7)? I think the AEM might not be adjustable but G4 and LM2 certainly are.  It might be a good idea to set all to display lambda units to eliminate that possibility.  If you have nitrogen or argon (welding gas) handy you can also use that to test all devices read same at lambda = 1.  Most devices use 14.7 by default but there are some exceptions that I've come across such as the NGK afx for instance uses 14.56.

I have heard reports that the AEM reads a little rich on the rich side of stoich but have never had a known reference to test against.

Edited August 31, 2015 by Adamw

[warmup]

Adamw: the stoiches were all set to 14.7, when I set both LM-2 and Fury to show lamda, Fury read about 0.03 lambda leaner, I just used to use AFR as the unit. btw AEM is adjustable for multiple fuel types, but can not show lambda.

Simon: where should I get start to check? Since they are all "digital" devices.....  

An extended answer of Adamw's question is that if I set stoich to 14.2 in Fury, the AFR reading offset will be in a acceptable range like 0.1ish, lambda is still off though.

but that doesn't make sense because I think you should not fix an "offset" by changing a "multiplier".

Edited August 31, 2015 by warmup

[Adamw]

I agree with you then warmup, this can't be blamed on "wiring" when all three are standalone devices with "integrated" displays - there is no way wiring can influence any of these displayed readings.  Since it does appear to be more of an offset type error rather than a multiplier error then I still think it would be worthwhile doing the inert gas test I mentioned earlier as that at least confirms the "zero" point. 

So you now have an interesting situation - which one do you trust??  After my personal experiences with innovate over the years I would tend to trust the fury more, but that's just my gut feeling...

[Dave Kriedeman]

Hi,

Ideal positions for the wideband sensors are 10 o clock, 12 o clock and 2 o clock.

Anything lower than this can be influenced buy H2O, a by product of the combustion process.

How old are your other sensors and are you using a genuine Bosch LSU 4.9 sensor, not an EBAY copy.

Unfortunately we are comparing apples to bananas here not apples to apples.

Regards

Dave.

[Simon]

With the wiring side the areas to look at would be any joins.

Crimps between the loom and sensor plug and the pin connections at the ECU connector.

It is possible that there is a bad connection.