My AEM AFR Gauge display bounces up and down and doesn’t give legible readings. Tuning is impossible.
New NGK Oxy Sensor produces correct parameters when tested on its own (as per Gregory).
When oxy disconnected, both OEM and MT ECUs work well and display keeps within 0.1 units.
When MT ECU is connected, tuning and AFR values function perfectly.
It seems like connecting the oxy causes havoc – I can actually hear it in the engine sound, when doing 1500 RPM (on Park), it fluctuates audibly! In other words, the mixture really is going up and down constantly.
What the heck is happening?

Also, what Calibration position do you use in the AEM Gauge? When I turned it from P0 to P4, the jumping was less intense, but still useless readouts.

Hi Killer, the following may be of help.
I am not familiar with the AEM gauge but I did run a construction kit electronic one for a couple of years on my TJ.

In ECU closed loop mode it is quite normal for the mixture to cycle between rich and lean – about once per second at idle and faster when at moderate load and rpm,
but the average of all the cycles will be at stoic (14.7:1). This cycling is a way of giving tight control over mixture in a system where there is a delay between a change in
mixture and it being registered by the oxy sensor. This operating area cannot be tuned as the ECU will re-adjust itself to bring the mixture back to stoic.
Of course disconnecting the oxy sensor will force the ECU to revert to its current mapping but will be fixed rather than cycling.
Your new oxy sensor may have different output parameters than the old (standard) one – is it a wide range unit?

At almost full throttle, or above about 3000 rpm, the ECU reverts to open loop mode and runs factory mixture mapping, and will show as rich on the gauge. This is
the only area that can be tuned by re-flashing the factory ECU (if possible), or by adding a piggy back unit such as Greddy emanage. Also I don't know how your
MT ECU operates.

Of course you may notice with the gauge a few other interesting areas of the ECU/ mixture control such as fuel cut-off on a trailing throttle and lean cruise mode.

Cheers
By the way, your not as old and grumpy as me!!

Nah, I'm way grumpier... :woot:
I'm familiar with the oscillating function of a std oxy sensor as you mentioned. In my case the gauge is displaying readings almost min to max on the scale in very wild manner, I can even hear the varying mixtures at that stage. It only seems to settle at idle or large loads/RPM.
When I tune it (on-road, live), I can take it so lean that the engine almost starves and the gauge still shows nothing intelligent, just up and down madly.
Got new NGK std oxy yesterday, no diff. And it's same as original, 4 wire. The sniffing AEM is a WB, yes.
Once std oxy disconnected, it runs ok and gauge display settles and any changes made in fuel tune are immediately visible ok at any load and RPM.
And, everything worked fine until about a week ago....
Double tested all sensors in the same circuit, TP, Coolant, AFM - all ok and to parameters.
Absoloutely baffling.
I'm guessing it is a conflict or earthing (all ok tho!) issue somewhere, but.
Plugged it with a M18 bolt now, so driving with "fixed" tune by the MT ECU until I have more time to waste on this. Works fine, unless weather changes drastically.

You mean AFRs cannot be tuned if std oxy sensor is connected?


Hi Killer, the following may be of help.
I am not familiar with the AEM gauge but I did run a construction kit electronic one for a couple of years on my TJ.

In ECU closed loop mode it is quite normal for the mixture to cycle between rich and lean – about once per second at idle and faster when at moderate load and rpm,
but the average of all the cycles will be at stoic (14.7:1). This cycling is a way of giving tight control over mixture in a system where there is a delay between a change in
mixture and it being registered by the oxy sensor. This operating area cannot be tuned as the ECU will re-adjust itself to bring the mixture back to stoic.
Of course disconnecting the oxy sensor will force the ECU to revert to its current mapping but will be fixed rather than cycling.
Your new oxy sensor may have different output parameters than the old (standard) one – is it a wide range unit?

At almost full throttle, or above about 3000 rpm, the ECU reverts to open loop mode and runs factory mixture mapping, and will show as rich on the gauge. This is
the only area that can be tuned by re-flashing the factory ECU (if possible), or by adding a piggy back unit such as Greddy emanage. Also I don't know how your
MT ECU operates.

Of course you may notice with the gauge a few other interesting areas of the ECU/ mixture control such as fuel cut-off on a trailing throttle and lean cruise mode.

Cheers
By the way, your not as old and grumpy as me!!

Been chatting with couple of car maniacs and engineers etc and this is not getting any better - by chatting. Seems just very much of a noo-damn-idea kinda case, cos nobody seems to have any more thoughts on this.
Hand grenade perhaps?