Widebands That You Would or Wouldn't Buy

[Hentai]

+1 for ECM AFM1000 units



likes

1)wide range of lambda readings
I have the A unit and it reads from .55 to 1.24 lambda, AFR stoich point is 14.56
They also have other units that read down to .3 lambda and units to read leaner
2)Long harness with barrel connector for the o2 sensor
3)Set output 0-5 volts
4)Outputs 3 AFR\lambda locations at startup\warm up so you can correctly configure in the datalogger\ecu for any offsets on the output of the wideband controller
5)Has a known pressure compenstation curve from ECM
6)Can work in 24v system and has a good range for voltage input, 11-28v

Dislike

1)no screen, this is more of a unit for data collection\tuning, less of a unit for your average car person.
You can purchase a optional display from ECM for the unit.

General Info

-Unit has wide range of lambda output options purchasable from ECM
-free air cal
-uses a NTK Lab grade sensor with barrel connector, each have a specific cal number that can be used to look up at ECM's website
-Blue box houses the controller
-Multiple harness length options

http://www.ecm-co.com/product.asp?a1000

[DelSolid]

+1 for ECM AFM1000.

I used the NTK branded version of this box and I always considered it bulletproof. The only con I had on it was the price and the lack of a waterproof housing or connector.

Does anybody know the price of the unit now? That link didn't show a price. Is it cheaper now or is it still ~$1,500?

[HelmutVonAutobahn]

5)Has a known pressure compenstation curve from ECM

Yeah. But it's about 4% per PSI of exhaust back-pressure. At just 4psi, that's +/- 16% !!! So, without an exhaust pressure sensor and compensation mechanism, it is not very accurate at all Unless you have next to ZERO back-pressure.

That's the big problem with NTK sensors. They CAN be very accurate, with the proper pressure sensor and related circuitry/code. But, without it, They are only accurate at one specific pressure.

Ironically, AEM has a back-pressure sensor compensation kit for their Bosch sensor devices. While the Bosch sensors ( which also have a specified pressure curve ) have around 1/4 of the error, due to pressure, as compared to the NTK units, need it less...

[Hentai]

5)Has a known pressure compenstation curve from ECM

Yeah. But it's about 4% per PSI of exhaust back-pressure. At just 4psi, that's +/- 16% !!! So, without an exhaust pressure sensor and compensation mechanism, it is not very accurate at all Unless you have next to ZERO back-pressure.

That's the big problem with NTK sensors. They CAN be very accurate, with the proper pressure sensor and related circuitry/code. But, without it, They are only accurate at one specific pressure.

Ironically, AEM has a back-pressure sensor compensation kit for their Bosch sensor devices. While the Bosch sensors ( which also have a specified pressure curve ) have around 1/4 of the error, due to pressure, as compared to the NTK units, need it less...

Based on the compenstation curve given with the AFM1000
.8 at 4psi would be around .8229
or
11.648 would be 11.98

Not nearly the 16%, but of course this is linearized curve.

Their formula for it.

"Pressure Compensation Equation for Lambda
Lambda(corrected) = (Lambda(measured) + B x P) / (1 + B x P) [Equation 3b]
where: Lambda(corrected) = the Lambda corrected for exhaust pressure.
Lambda(measured) = the Lambda output by the AFM.
B = 0.000627 for Lambda < 1.0 (rich).
B = 0.000830 for Lambda ≥ 1.0 (lean).
P = the exhaust pressure in mmHg above the pressure at which the sensor was calibrated (using the CAL POT on the AFM while the sensor is held in air). Equation 3b is valid for -152 mmHg < P < 532 mmHg."

I have a turbine inlet pressure sensor to the datalogger with the o2 sensors.

I don't see anything ironic about aem having a sensor for the 4 channel units, it is really needed on turbo setups, which will see above atmo pressure in the turbine\manifold

I haven't seen a sheet for the LSU 4.2\4.9\ADV sensors for pressure compenstation

[DelSolid]

I haven't seen a sheet for the LSU 4.2\4.9\ADV sensors for pressure compensation

Here you go.

This is all the Bosch 4.2 spec says about error due to backcpressure

They expanded their information in the 4.9 spec.

and more in the ADV

[DelSolid]

It's shocking how much more error the ADV has over the 4.9 and 4.2

[DelSolid]

Ironically, AEM has a back-pressure sensor compensation kit for their Bosch sensor devices. While the Bosch sensors ( which also have a specified pressure curve ) have around 1/4 of the error, due to pressure, as compared to the NTK units, need it less...

I don't see anything ironic about aem having a sensor for the 4 channel units, it is really needed on turbo setups, which will see above atmo pressure in the turbine\manifold

Exactly. We put the pressure compensation into the 4-channel device because it is specifically designed for individual cylinder measurements and a huge percentage of our customers are turbocharged. My personal car (2JZ 6Cyl with a GT47) has 7 AFR sensors on it, 6 in the individual cylinder pipes before the turbo and one in the down pipe. The back-pressure measurements before the turbo are typically 30-35 PSI above ambient so compensation is very necessary, even with the LSU-4.2. We have measured the back-pressure in the down pipe and it was essentially zero, but that was to be expected because it is a 5" diameter pipe that is only 30" long and a relatively straight shot to the outside world.

[Hentai]

It's shocking how much more error the ADV has over the 4.9 and 4.2

I would "suspect" it has to with the point of the ADV, being for pre turbo applications. I've seen one in a oem car. Also I believe the design tolerates higher temps than the 4.9 and 4.2 does. It would surprise me in order to meet the goals it was setup for that was one of the penalties it had to take.

[Tre-Cool]

1st post, so i'll make it a strong one.

I've been tuning mostly factory GM based ecu's for over 8 years or so, so have used a few different sensor kits, so these are my impressions.

1st wb was an LC1, it died after 4 months. (this was back in 2006 or so)

Moved onto using some of the PLX SM-afr kits, was primarily using the serial output back then before modding them to use serial comms into my efilive flashscan cable. It's only at this point do you find out how bad the difference is between the analog to serial data. even a large difference between the display guage and analog signal was enough for me to give up on them.

Started using the Techedge 2j2 (never ended up upgrading to the 2j9) again used these primarily for the direct serial output into efilive.

Started using the MTX-L as permanent gauges & for tuning in a few cars due to it's small form factor. Generally had 2 permanently mounted in the car on v8's, 1 for each side of the motor. Generally they have been pretty good but over time they started getting lots of E8 errors, probably due to the heat as they are mostly running on supercharged v8's. Still using LSU 4.2 sensors though.

Started buying the ecotrons ALM kits, 1 x alm-led and 3 other guage kits. the response and data I was getting was much better then the previous stuff, but I put this down to the 4.9 sensor if anything. Retuned a few cars from other shops and fuel economy has always improved on the cars & closed loop fuel trims get's within 3%. Probably not much to do with the sensors/kits but more to do with how the previous tuners were reading from tailpipe or the sensors they were using.

I even removed the old LSM-11 "wideband" from my dyno and put in the ALM-LED in it's place, I don't get much difference between wideband readings from what I fit in the car header to reading out of the tailpipe on the dyno. the LSM seemed to read leaner and leaner after each run, but it's a pretty old sensor.

Now looking at the new AEM x-series line & the 14.7 kits to test out.

Cheers
Dave

[jalai]

It's shocking how much more error the ADV has over the 4.9 and 4.2

Interesting, I have LSU4.9 document where is this pressure comp. curve.




Also, Bosch MS6.x ECU software has correction like this.