LC1 Gone crazy

[Bluehaze]

Sorry, but that's the best description can give.
Apologies to all crazy people, no offence intended

Anyhow, it starts like this; I've been using the LC1 unit with the MSII along with the gauge (G5 analogue). Working well for a few weeks. Driving the car occasionally and running the VEAL to get the fuelling nice. All this has worked well, and I probably have a total of an hours' driving with VEAL and the LC1 forming a great team. The car drives well, but I'd like to continue the refinement process.
Then I started noticing the AFR gauge would occasionally "Stick" or drop down to 10 or go up to 18, even though I know the engine isn't running that way.
This fault would disappear as randomly as it appears, but has been getting progressively worse-ie more often and for longer periods.
Its not a meter fault, (tap-tap) as I've had the laptop connected, and the MS is showing similar signal discrepancies, but more along the lines of swinging wildly between dead lean and ultra rich.
Sometimes the LED from the LC1 indicates warm-up mode, but not often. Most of the time the LED is on solid.
I have set the EGO correction off for the time being, but I wish to continue setting my fuel map using the LC1, as it seemed like a good tool until it went on the fritz.
I have yet to re-calibrate the WB unit, I hope this will fix things, but I'm pessimistic, as this looks a bit weird to me.
Anyone else had this?

PS Inline 3.3l NA engine with port injectors
MSII running V2905 code
WB sensor mounted approx 4 feet from head. (at the far end of the extractors, just before the join with the exhaust pipe)

[trakkies]

If both the LC1 gauge and the readout in MS agree, I'd guess at a faulty LC1? Or a connection problem?

[Bluehaze]

Hi Trakkies
I agree, but I'm fairly confident with my wiring. All connections are soldered, and all the earth connections for the ECU and the LC1 are made by way of all ground wires soldered to a crimp lug and bolted to the engine block. All the other signals to the ECU have a signal to noise ratio of around one percent. It's only the O2 feedback which is all over the shop.
If indeed the LC1 has become faulty, this would be most dis-appointing. I hope this is not the case. If the controller has failed, I'd be a fool to buy another one and risk losing another one after a few hours use.
The only cause for failure I can imagine is the sensor unit itself. This is however mounted correctly, at the one-o-clock position at a distance downstream such as not to cook it.

[Joe Clark]

I used to run an LC1 and had similar issues, apparently the older ones were quite bad for throwing a wobbly if they were subjected to noise, I think what they did was perform a calibration while running, or they would need the setings for the outputs reflashed.

I ended up moving to another type of wideband and never looked back!

[trakkies]

Hi Trakkies
I agree, but I'm fairly confident with my wiring. All connections are soldered, and all the earth connections for the ECU and the LC1 are made by way of all ground wires soldered to a crimp lug and bolted to the engine block. All the other signals to the ECU have a signal to noise ratio of around one percent. It's only the O2 feedback which is all over the shop.
If indeed the LC1 has become faulty, this would be most dis-appointing. I hope this is not the case. If the controller has failed, I'd be a fool to buy another one and risk losing another one after a few hours use.
The only cause for failure I can imagine is the sensor unit itself. This is however mounted correctly, at the one-o-clock position at a distance downstream such as not to cook it.

I'm not sure about the LC1 as I have a Tech Edge. That has independent signal and power grounds, so the signal ground is at the MS DB37 as with other signal grounds, not the engine block. Just to prove it's not a noisy ground, can you disconnect the LCI from the MS and see if that makes any difference to its readings? But it does sound like a faulty LC1.

[Matt Cramer]

Where is the LC-1 getting power and ground? We've found they work best if they get power from a 12 volt source that is hot in Run but not Start, and if they are grounded straight to the engine block.

[Bluehaze]

Funny you should ask Matt
I went away and had a good think about it.
When I installed the unit, I read that
1-don't wire so as to get the unit heated up before starting, as condensation may be blown onto the tip, causing damage.
1-don't run the motor with an un- powered unit, as exhaust gases may cause harm.

With this in mind, I wired from the fuel pump circuit. As it turns out, I also wired this feed to the ignition coil supply.
There was method in my madness here, but I considered that perhaps the load from the pumps caused a low voltage condition, and of course the coil will make heaps of noise on the circuit, which used to run the original NB heater. (naturally).
The grounds are all run seperately to the block, soldered to a lug along with all the other grounds, I couldnd have a more solid ground.

Now, I wired the supplies to the unit from a switched ignition feed, and connected the computer to the unit.
The software showed the unit had reverted to default, ie channel 1 NB output, ch2 default LC1.
I've got a gauge, so set CH1 to same as CH2, and did a heater and free air calibration and hey presto!
She's working again.
I'm very happy I managed to fix the unit.
So Matt, does that sound like a reasonable hypothesis re the unit going nuts and defaulting itself?

[trakkies]

The feed to the ignition will have lots of 'dirt' on it. If you wish to use this because of how it is configured, add a relay with the coil fed from that circuit, but the actual supply clean. Same as you'd do (should do) with MS.

[Bluehaze]

Just an update.
Looks like the noisy circuit was on the money. Since connecting to a cleaner supply feed, the AFR gauge has worked perfectly without flipping out.
Even went for a trip to Philip island on the week-end, a 300 km round trip and it was a delight to watch the gauge working as expected. I've also given veal a good workout, and the car runs better and better.
Thanks for your help.