NGK AFX analog values far off from display???

[ccrp]

Guys,

yesterday on the dyno I got puzzled beyond believe. I have two brand new NGK AFX which are inputted to HPtuners in Lambda (main) and AFR (Secondary) and suddenly on a Kenne Bell supercharged C5 the discrepancy between display and analog output has been far off. Like 12.1 display and 13.7 analog output. On two cars before all was OK.

If I used my own configured PID or the HPtuners premade PID did not make a difference.

We connected a calibrated 5 gas tailpipe sniffer and it conicided with the display, put another Innovate to other exhaust pipe and that has in between and the pre-installed Autometer wide band was way off.

after a new calibration in fresh air on all instruments another try with similar results.

what the heck? Anybody got an idea here?

calibrated NGK today again and when showing CAL- output voltage is 4.94V like always..

I simply have no explanation for yesterdays behaviour!

[DSteck]

Where area you grounding these?

FYI, seeing what it reads when it says CAL will not tell you if it is calibrated right. You have to check it during the startup sequence.

[ccrp]

+ and neg are connected straight to battery with clamps which I have on both (actually I have 4 AFX in total).

can you explain about the startup sequence?

[DSteck]

While the unit powers up, there's a 20 second count down. During that time, it will spit out three different preset voltages.

[ccrp]

yes they do:

at the beginning 0.987 / 0.982 volts, then 3.97 / 3.96 volts, then 0.05 / 0.04 volts and finally again 4.93 / 4.92 volts

that correct?

[DSteck]

Yes.

[ccrp]

OK, thanks. and what mean those voltages in detail?

I can assume that 3.96 would be Lambda 1, 0.05 Lambda 0.62 8AFR 9.0) and in the end the upper end AFR 16. But what is the first value?

[DSteck]

I don't know exactly. I just use it when it spits out 1 lambda.

[LSxpwrdZ]

I have had a few cars with some sort of electrical issue that I literally have to use an external standalone battery to supply the wideband. I've only had it happen twice in the last little bit but it was def the vehicles grounding or charging system issues.

Restarting the wideband and logging the startup on the 20sec countdown is how I normally detect this type of issue.

[10_SS]

OK, thanks. and what mean those voltages in detail?

I can assume that 3.96 would be Lambda 1, 0.05 Lambda 0.62 8AFR 9.0) and in the end the upper end AFR 16. But what is the first value?

I forget the order exactly, but it spits out all three readings upon power up:
1) It's lowest possible reading
2) It's highest possible reading
3) Stoich (1 lambda).

I simply calibrate/verify with HP tuners using these three values by cycling power to the WB.

[directnosfogger]

Hey guys - sorry to revive this from the dead but I'm trying to make sure my numbers are calibrated appropriately.

I see a readout of 1.001, 3.98, .020, 4.85. My low and high voltages are not coming across as I would have hoped, but can I adjust for this within my custom input PID?