De-catting / Lambda Sensors

[mel]

I'm quiet happy to do the actual mechanical alterations of cutting out the cats and swapping them over for staright through pipes on clamps that can then be swapped back come MOT time. However it's these slippery little bugger on wires that are attached to the cats that I don't get. Can I just tape them up out the way? can I unplug them? or do they feed back some cunning information to the Engine management system that will send it all Bill Gates if they don't get any input?? Now were's my disc cutter

[shpub]

Basically you need to rechip the ECU so that the ECU doesn't report you to the Cat Protection League for cat abuse.

Steve
www.tvrbooks.co.uk

[simpo one]

I think - but disagree if you like - that as well as the differerent chip, the lambda sensors are normally disconnected for the 'ex felis' scenario. However, despite new chip, a friend's decatted TVR idled like a pig until the sensors were RE-connected. On quizzing the original garage who did the work they said 'it seemed better like that'.

Bizarre. Can anyone explain this? I can't!

[joospeed]

yep. if you disconnect the sensors the ecu gets a 0volt signal, meaning lean running, the ecu will apply an adaptive trimming value that it learns at idle and applies this across the rev range, hence if the car is decatted and the lambdas disconnected it will run richer and richer until the rich trimming limit is reached, you need a chip whcih disregards the lambda sensors completely to work. or leave the lambdas in of course which is cheaper.

[budd]

In addition to the decat chip you will also have to change the tune resistor,both I believe are available through your TVR main dealer at around £100.

[tvradict]

would it not be possible to trick the ecu into thinking all is well by shorting the Lambda Sensor?

For there sensor to return a 0 volts signal there has to be a 1 volt signal sent which is returned if it's ok! Run a wire shorting it out and effectively removing the sensor from the circuit, then there should be no problems! Correct?

[david beer]

Is it not true that the ecu does not use the lamdas after 3500 rpm anyway.

[GreenV8S]

quote:
would it not be possible to trick the ecu into thinking all is well by shorting the Lambda Sensor?

For there sensor to return a 0 volts signal there has to be a 1 volt signal sent which is returned if it's ok! Run a wire shorting it out and effectively removing the sensor from the circuit, then there should be no problems! Correct?

It's never that easy is it? In practice, lambda sensors are extraordinarily complicated things to use. The sort we use on our cars only send back two values, think of them as 'rich' and 'lean'. The ECU tunes to the optimum value by cycling the fuelling richer and leaner, it knows it's on the threshold when it gets equal proprtions of rich and lean signals from the sensor. So any steady signal is going to really upset it.

In any case, it's only a single resister to change for goodness' sake, cost you about 5p and two minutes to fit. And it gets you to a standard configuration that any mechanic looking at the car would recognise. This seems utterly trivial compared to the work required to remove the cats.

[mel]

quote:
In any case, it's only a single resister to change for goodness' sake, cost you about 5p and two minutes to fit. And it gets you to a standard configuration that any mechanic looking at the car would recognise. This seems utterly trivial compared to the work required to remove the cats.

All the rest of it I get, being from a "metal bashing" type engineering background I view physically removing the cats as a walk in the park (well in comparison to making bike systems from scratch) but its the bloody electrics bit that baffles me, single resistor?? where?? or am I correct that it's easier to make the linkage pipes that will replace the cats capable of taking the lambda sensors?? so that they lambda or sensor or whatever they do the gases that are in the replacement pipe instead of whats in the cats??