It's a bad practice; really bad.

The sensor is supposed to be in "free-air" for the calibration. We are talking about an import device here that requires accuracy. The entire purpose of the free-air calibration is so the sensor knows what 20.2% oxygen content "feels" like. Remember, you're calibrating the sensor, it has no idea what-is-what unless you give it an absolute. Do you not zero your bathroom scale before you weigh your fat ass? Makes you feel better when you weigh 10 lbs "less." If you're not metering the atmosphere, then you're calibration will be incorrect.

I, unlike AFRman here, have no idea when my AFRs are correct unless my WB tells me. If I tell my WB that 14.7:1 AFR is actually 14.0:1 because there's residual gas floating around the pipes, then it'll always be incorrect. But how will I be the wiser?

Also, the proper procedure for the calibration is holding the button down for three seconds, then it starts to calibrate. IIRC, it calibrates for something like 5 seconds. That's kind of a long time to gamble that the ECU is cutting enough fuel long enough depending on the conditions. (Also have you ever turned off your car and continued to meter the exhaust? I betcha it doesn't read lean right away, nor after ten minutes)

Then after all that you're even supposed to power off the WB and turn it back on after 30 seconds to store the calibration into memory.

Why not get off your ass, unscrew the sensor and let the read the atmosphere, do the free-air calibration and screw it the fuck back in. I tend not to try to gamble with things that need to be "accurate."

Sensor calibration will do precisely jack shit with regards to a discrepancy between the display gauge and the MS reading. Those two should always indicate the same regardless of how far out of calibration the sensor might be.

What kind of display gauge are you using?

The problem you're having is either wiring (ground offset for example) or software settings in either the MS or the LC1. If you have the ability to generate a precise reference voltage, you can test the MS's calibration by applying exactly 2.5v to its input, which should cause a reading of appx 14.9:1.

Fortunately, the LC1 allows you to custom-tailor the calibration of the outputs, so you can correct for offsets. Just make sure that whatever you're using as a reference (ie: the display gauge) is actually correct before you go tweaking the outputs. This is the thing I like best about the XD-16 gauge- since it communicates digitally with the LC1, it's guaranteed to show you precisely what the LC1 is actually calculating, and cannot be affected by offsets or output miscalibration.

It doesnt take 5 seconds to calibrate, and you dont have to power cycle it for it to remember anything. How is it going to store data when you cut the power?

Next, I wouldnt be recommending the shit unless I was certain it worked. If you dont believe me, go cold start your car and get a recording of your AFR gauge for the first 10 min. Then go for a drive and recalibrate exactly how I describe. Wait till tomorrow and restart the car again. The readings will be the same aside from whatever minor descrepency due to temperature changes.

Youre right on one thing, if you just shut off the engine there is exhaust gas lingering. But now go floor your engine and rev to 7k rpm stationary, and turn the key off. I bet you it will be clean.

I had mine running in my exhaust without power for 2 weeks, which you apparently are not supposed to do, guess what the shit still works. Explain what difference its supposed to make if its powered or unpowered, the damn thing is already floating around in something like 500+ F temps.

Bottom line, this thing is not that fragile.

The one that came with the LC-1, "DB" model, blue if that makes a difference :giggle: .

DB Gauges: Air Fuel Ratio Guage

All of my grounds for LC-1 controller and gauge are on the same lug, at back of intake mani with the ECU grounds.

If it were <0.5 discrepancy I wouldn't care, at least not until I got into hustler territory or start WI.

>0.5 causes a little concern, although at least the MS is getting the leaner reading, and this is what goes on the datalogs. Although being nearly a full point too rich can be bad news as well.

I'm about to just recalibrate, and reset the linear settings on wideband controller. I've done it twice already...but still consistently see leaner values on MS dash gauge versus actual gauge. The gauge may just be that far off, or it may be the amount of time it takes the eyes to travel from one to the other, but it's pretty consistently off, and worth figuring out.

^ Check to make sure you have the configurator set up correctly, and slow down the update speed in the LC1 for both ports to see if there really is a discrepancy.

This might explain it. Ideally, the wideband (and gauge) should be grounded as near as possible to the ECU. This means splicing into one of the MS's ground wires, not merely running a separate wire to the head. It seems counterintuitive, however the goal is not to give the LC-1 the "best" possible ground, but to make sure it sees the same ground as the LC-1.

Even if they're grounded on the same bolt?

EDIT: MS and LC-1 grounds have seperate wires, but not grounded to seperate parts of the head. Those two lugs are on top of each other, with the same bolt attaching them to the head. I should really know how important grounds are from car stereo days, but for some reason they always get so little attention, and not just from me I've noticed.

I'm about to just cut every single ground wire, weld(er, solder) them ALL onto on big lug, and torque it down to 22ftlbs.

Yeah, even if they're grounded to the same bolt. It's the voltage drop across the MS's ground lines (due to sinking high-current stuff like injectors) that's the concern. IOW- the potential at the MS side of the ground wire may be slightly higher than at the head side of that same wire.

Ok, thanks for that explanation...that's enough to explain the small difference, despite what seems to be a commonly used setup.

So if I splice into the MS ground wire -at the head- or solder them all together I should get the same potential for all the ground wires?

I barely made it through my electrical engr. classes, but I don't see any reason to splice further upstream.

Youre getting wrong readings not because of bad grounds, you have your software set up incorrectly.

I have a shared ground wire running from the throttle body to behind the exhaust manifold. From those points they split off to the MS/LC1, to my DB Gauge and other gauges, and IAT sensor. Everything works.

I took a look at things, and everything is spot on...I've done it three times before, and even helped talk other people through the setup.

I'm willing to try anything again though, as I'm tired of being that last little bit off.

I agree that this is most likely. Just figure we should make the ground system as ideal as possible before attacking the software.

I just did a free air cal and the LC-1 display swings like crazy.

On output 2 I should be ok with the default settings - right ?

As long as you wired the brown wire to the gauge.

Yes, it is wired right.

Funny thing is that all this was working correctly until I tried to configure the LC- 1 no 1 output. :bang:

Those tabs can be really strange....I've set output 2 only to reopen and see it's back to default.

Once I got it right I decided to leave it alone...which is why I'm hesitant to reset it.

Just keep trying until it sticks.

Well, I never did solve the discrepancy between the Glowshit gauge and Megatune.
I just unplugged the gauge and it lives on top of my toll box.

On a brighter note, my friend helped by coming up with a good AFR table. We drove around the twisties while logging and came up with a good tune. The car runs really good now.

Now I need money for a turbo ;)