Elanscan log following emission test fail

[tomatcbl]

I own a federal elan. In June, the car failed its Maryland emissions with a high hydrocarbon (438 ppm vs state limit of 200) and high CO% (2.34 vs state limit of 1). I replaced K&N high flow air filter with a new one, replaced plugs, changed 02 sensor, PCV valve, oil and adjusted timing from 12 btdc to 16 btdc. Retested today and failed again with HC at 745 and CO% at 1.86.

Posted on problems forum and there's only on f in Geof Smith suggested posting an Elanscan to assist diagnostics. Test involves logging from start-up - two periods of constant 2500 rpm following return to idle (the emission test protocol) and then a short drive on highway.

Any suggestions appreciated

[GeoffSmith]

Tom, as I suspected, your O2 sensor output is marginal and is not reaching 4V consistently. If you look at the CO potentiometer/Oxygen sensor output, it warms up and starts oscillating at about 150s, but the output only ever approaches 4V at 550-610s, 640-690 and 1000-1210s. Flag 42 bit 0 (Closed loop) shows that the ECU is going closed loop during these times, but it should be consistent at all times you are not under boost conditions. What you need to achieve is the two outputs resembling those between 1065 and 1100s in your log. You are obviously running open loop for most of the time which explains your high HC and CO figures.

I think you said you'd replaced your O2 sensor so perhaps it is the wiring: I believe the ECU input is a 10kohm resistor to 5V so if you can find a 10kohm resistor and connect it from pin A of the O2 sensor to the O2 sensor body (see EMJ.4 Code 13 for connections) and see if the ECU is reporting 2.5V.

A last minute thought: I wonder if the sensor has been contaminated by e.g. thread lubrication compound. If this is the case, I think you may be able to clean it by heating it as hot as you can get it with an air gun (don't use a flame as it will contaminate it with carbon).

[saska]

Tom, as I suspected, your O2 sensor output is marginal and is not reaching 4V consistently. If you look at the CO potentiometer/Oxygen sensor output, it warms up and starts oscillating at about 150s, but the output only ever approaches 4V at 550-610s, 640-690 and 1000-1210s. Flag 42 bit 0 (Closed loop) shows that the ECU is going closed loop during these times, but it should be consistent at all times you are not under boost conditions. What you need to achieve is the two outputs resembling those between 1065 and 1100s in your log. You are obviously running open loop for most of the time which explains your high HC and CO figures.

Whilst I agree with most of your comment, my experience is that on closed loop 1499's the O2 goes slightly over 4.5V (not 4.0V) and oscillates all the time (the crossover). Tom's scan shows the O2 staying relatively high at ~820-910secs and also ~1000-1030 secs (no oscillation/crossover at all).
It could be duff O2 sensor, could be wiring but could also be earth. Exhaust manifolds/downpipes get rusty as do the earth braids going to the engine.
Since if the earth braids to the engine were faulty a load of other sensors would be kicking faults (which they aren't) then my money would be on the O2 sensor, the wiring to the O2 or the earthing of the downpipe. Most of which you've already suggested.

[tomatcbl]

Thanks for advice. Given that this is the second 02 sensor, would seem to suggest a wiring or earth issue rather than sensor or thread lubricant.

[HJ2]

Beware!

My wirig went over the fixture for the headlight pod. Over the years the isolation on the wires and the paint on the bracket were gone and caused all sorts of O2 problems, so check this as well

viewtopic.php?f=79&t=16507&hilit=time+to+shorten&start=40

[tomatcbl]

Thanks for link - wouldn't have looked for break in insulation initially - but would make sense for symptoms.