DIYEFI.org Forum

So Innovate as you know announced LSU 4.9 with a firmware upgrade and a sensor change (with connector change) here: http://www.innovatemotorsports.com/support/49.php

I've ordered an LSU upgrade kit for my MTX-L (since my LSU 4.2 sensor died after a few years of use about 4-5" away from the turbo in the downpipe) (I am in Australia and the package is coming from the USA, so it'll take a while).

I emailed Innovate asking if the LSU 4.9 still requires calibration as the LSU 4.2 and they responded that the LSU 4.9 just like the 4.2 requires calibration according to their calibration schedule.

My question is why does the LSU 4.9 + innovate still require calibration, and how does it calibrate without having a reference air system like the LSU 4.2

Don't get me wrong I'm glad Innovate allows their MTX-L to be upgraded via firmware, but I was expecting more reliability out of the LSU 4.9 without the need of calibration.

To the controller, there is little difference between the open-air reference, used by the 4.2 and the sealed reference, used by the 4.9. Electronically, the 4.9 requires a "bias current" to replenish the sealed reference. Which is no big deal.

Innovate controllers require "free-air" calibration because they to not use the factory calibration resistor that is included in the sensor's connector. That is, they need to figure out what the ratio of pump-current to Oxygen is. The controller knows exactly how much pump current it is applying. It just needs to know how much Oxygen that represents ( to calculate AFR/Lambda ). The factory calibration resistor represents ( indirectly ) this ratio.

To be fair, unless there is some sensor damage. One free-air calibration procedure, when the sensor is new, will provide AT LEAST the same accuracy as using the factory calibration resistor, over the life of the sensor. "Constant" calibration is only needed if the sensor is aging rapidly, due to some sort of abuse; such as overheating, lead poisoning, silicone poisoning, thermal shock, etc.

If you find that you need to re-calibrate often, something is wrong.

Which begs the question : "Why don't they just read the calibration resistor, once, when the sensor is new; and, let you do a free-air calibration if/when you need it ?" That would probably make everybody happy

Factory laser trimmed resistor, too, right? :-) They wouldn't go to that extreme for no reason. The reason, though, is likely production line use where a free-air calibration is unacceptably manual and slow.

Actually, free-air calibration is how they trim the calibration resistor on the production line.

They attach the new sensor to a VERY accurate controller. Then, with the sensor in free-air, they LASER trim the calibration resistor until the controller reads the correct Oxygen value ( actually, pump current value ). They use some formula to calculate the exact Oxygen percentage for the current conditions.

And, poof! every sensor, as a complete unit, is identical.

That said, there is no real reason why Innovate can't just read the value of the calibration resistor and divine what the actual pump current was at the the time of the LASER trimming. That's pretty easy. And would eliminate one of the big complaints that the market has had about their products for years. But, they seem to be stuck in an extended cycle of cut-n-paste engineering; with declining performance.

The ability to free-air calibrate a sensor is a very good thing. But, it does not need to be mandatory. Especially, not for a new sensor.

Sure, but not on the vehicle at assembly time, which is what I meant. This would not be feasible.

That would be a complete mess.

Exactly. Manually controlled free air calibration, as done in aftermarket widebands, would be infeasible on a production line (of cars). Glad we're on the same page, now. :-)

Yes. I think that is one of the issues that kept the Innovate design out of the OEM market, along with sensor longevity. It's a shame, because both issues are easily solved.

I don't recall where I read this or heard it but I have heard of certain widebands being free air callibrated under certain decel throttle conditions where there is no longer any fuel burning, but tons of air moving though the exhaust system, I believe its also could be cal aganisted the known narrowband sensor

I don't think "free-air cal on decel" thing has ever been successfully tested. There is always some contamination of the air stream. Injector leakage, fuel pooling, EGR, PVC, blow-by, etc. And the airflow is minimal, as the throttle plates are ( mostly ) closed. Also, the heater control is in transition.

The narrow-band gives no useful information for calibration. The wideband already knows where lambda 1.0 is. That is "zero" for the wideband. And, also why one should NEVER judge a wideband as "accurate" because it reads "14.7" at idle on a "known good" vehicle. Every wideband is accurate at 14.7 AFR (Lambda 1.0 ).

That is analogous to a volt-meter reading 0v when the probes are shorted together. It's better than NOT reading 0v. But, gives no indication that the scaling is even close.