Blown06

iTrader: (32)

I've been running the AEM UEGO Wideband in my truck since the beginning and have always trusted it. I have had no reason not to. The truck runs great and I haven't blown it up due to false A/F readings.

Well, I bought an Innovate LM-1 Wideband controller to be able to tune some other cars and decided to hook it up to the truck and see if I could get it to work. All seemed good so I decided to have the AEM hooked up at the same time to compare outputs. To my suprise the AEM seemed to read about a half point leaner and also was not nearly as steady as the LM-1. I didn't have a chance to actually drive the truck. All logging was done in the driveway.

Questions/Concerns:
1. My bung is in the 6'oclock position (upside down). I realize this is bad, but thats where it ended up. Could this be a cause for concern that the sensor for the AEM could be going bad.

2. I'm not sure which wideband I actually trust. What are ya'lls thougts?

3. When setting up the custom pid for the LM-1, I have volts set at 10 and the offset is .63 instead of the recommended .625. It idle and free rev, HPT and the handheld controller seem pretty close. Could the variance change under load and WOT?

4. Any other recommendations?

Thanks in advance for the feedback.

Frost

iTrader: (45)

1. 6-oclock is the worst possible place in the pipe to put it.. 8 to 4 is ok, 5-7=no.
2. The offset seems to be static in most cases. You can force the output of the LM1 (with the log works software) to equal 5V and log raw voltage in HPT to observe % error.

Here's another issue that you didn't mention but is worth bringing up. As the sensors begin to age, they typically report leaner conditions. The LM1 uses free-air calibration for the sensor to correct to "fresh air" whereas the AEM unit is reliant on the cal resistor in the connector. The cal resistor is matched up to the sensor's characteristics at time of production. Over time though, through the course of many conditions, the sensor's output will change but guess what; the calibration resistor will not (can not). As a result, your reading will be skewed. Your LM1 handles aging sensors by calibrating them against free air for better accuracy over time.

I do wish there were a way to slow down the updating of the AEM unit...

Blown06

iTrader: (32)

Thanks for the reply.

In your #2 answer. Is this the same as setting the output voltage to 2.95 at 10 afr and 2.95 at 18 afr in the LM Programer? Forcing the output to reflect 14.7, and noting the difference in HPT? Do I have to setup a custom histogram for this? If so, can I get a screen shot from someone?

Frost

iTrader: (45)

Yes, same thing

I don't have any special histo, I just log raw voltage to the EIO input of choice and look at it with a DMM at the same time. The % error is what I am looking for. You can check it setting the voltage high and low to verify the %error is nearly consistent.

Doc

iTrader: (9)

The Free air calibration with the Innovate products are only half of the picture. You can command 1vdc and 4vdc accross the AFR spectrum and log the voltage offsets for them respectively and see what your datalogger's analog inputs are really seeing and adjust their inputs for even better accuracy.

Thus, setup LM programmer to command 1vdc at 7afr and 20afr (a flat line) and see what you get... I bet it is not exactly 1vdc.

Then setup LM programmer to command 4vdc at 7afr and 20afr. Take the average offset of the two and adjust your calculated pid accordingly.