Removing cats...ECU pin 20 or 21

[bootie3367]

As for the pin business, I didn't touch mine when I de-catted and never had a problem.

I accept that you have never noticed a problem, but doesn't that imply that the generally accepted premise that the lack of back pressure produced by the presence of a cat requires a different engine map !

So why did they produce an ECU with the ability to switch between cat and non cat

Or perhaps the difference is only noticeable if you 'give it some 'ammer' 

Aye Mel, I was told many moons ago that without the cats there would be a good change that the car would run lean at WOT. As such I asked for my spare ECU to be chipped, this was done eventually, some three years later and in the intervening time I never noticed any high temps when running at high speed for long. I still haven't fitted the 'chipped' ECU yet as I keep having to fit the cats once a year.
We'll see how it goes next May when I take her down for the annual test.

[melinx]

I would imagine that it would have to run VERY lean at WOT to exceed the ECU's ability to compensate by reference to the Lambda sensor output

[turbojo]

The ECU runs open loop at WOT, it doesn't run off the lambda at all nor take any notice of it. However when the car is running in closed loop it will slowly adjust the LTFT due the the extra air flow at cruising and then carry that % on at WOT so in essance it will 'chip' its self.

[melinx]

The ECU runs open loop at WOT, it doesn't run off the lambda at all nor take any notice of it.

Please quote the source of this information I'm sure that we would all be interested and it may contain further information of interest.

[turbojo]

I don't have a source of this information though a quick google will tell all.
I was the R&D dept for a LPG supplier as well as fitting the occasional lpg system so I have quite a indeph knowledge of engine management systems.
All cars that run the old style lambda sensor run open-loop at WOT, they have to as a engine wants a mixture of around 12/10:1 (jap stuff and turbo/supercharged are more the 10:1) at WOT and a normal lambda will only state rich or lean. But not by how much hence why the car will go 'open-loop' and run off the ecu's prestored map.
When crusing the engine management system tries to keep the mixture stoichiometric (air–fuel mixture is approximately 14.7:1) using the lambda and it flicks from fractionally rich to lean many times a second (hence the lambda 'switching') and the car is running in closed-loop. The ecu compares how much fuel the engine's needing at that point in time to switch the lambda and compares that to what it should be against the fuel map and the difference is the STFT (short term fuel trim - usually a %). This is a constantly changing figure, as fast as the lambda switches.
If the engine always needs for instance a x% increase in fuel against it's map for a long duration of time and for a variety of loads and rpm's then it will pull the LTFT (long term fuel trim) out by the average x%.
The LTFT carry on into the open-loop mode so if you've removed the cats and fitted a filter so the car always needs +5% average STFT the LTFT will go +5%, and the car will also get +5% fuel at WOT.
My 24v carlton stays closed-loop until the dual-ram kicks in (4k) and up to 3/4 throttle and my old v8 range rover is closed up to 3k and 3/4 throttle for instance.

Plug a analog multi-meter up to the lambda and take the car for a run and you'll see what I mean.

The only exception to this statement is cars that have a wideband lambda. These work much faster than the standard type and actually give a mixture reading so very modern cars run a constant closed-loop with the lambda stating the actual mixture so it can be kept at 14.7:1 or 12:1 on the nose. Hence why fuel economy has improved in recent years, even on sporty stuff.
People running Megasquirt with a wideband can (and do) run the car with no map stored in the ecu (google it!).
A/F units that you can buy for cars to see what the mixture is doing in open-loop all need a wideband lambda fitted.
Joel.

[PeterC]

We got 3 cats, JD and Brandy both black and sisters and Nibbler a black and white male rescued cat. Their emissions are clean as they go outside.

Thats irresponsible just dumping the problem in the environment. You will soon be expected to fit them with a cat-litter-converter. Incidentally I like cats they are warm however a problen is they dont like getting wet.

[kevinfourlegs]

If its raining they go out to use the loo and then come straight back in.

[turbojo]

My dog doesn't even do that, she'll hold it in until it stops!

[kevinfourlegs]

Your dog will have a long wait if you live in Wales. It can rain for days down here. That's why Wales is such a green and lush land, its the precipitation.

[melinx]

Plug a analog multi-meter up to the lambda and take the car for a run and you'll see what I mean.

The only exception to this statement is cars that have a wideband lambda. These work much faster than the standard type and actually give a mixture reading so very modern cars run a constant closed-loop with the lambda stating the actual mixture so it can be kept at 14.7:1 or 12:1 on the nose. Hence why fuel economy has improved in recent years, even on sporty stuff.
People running Megasquirt with a wideband can (and do) run the car with no map stored in the ecu (google it!).
A/F units that you can buy for cars to see what the mixture is doing in open-loop all need a wideband lambda fitted.
Joel.

If you connect an analog meter to the Carlton Lambda sensor you will probably totally flatten the output voltage
and run the risk of damaging the Lambda sensor: The sensor should not have a current of more than 1 microamp drawn from it, analog meters draw considerably more than that; only a decent quality digital meter should be used.

I fitted a narrow band a/f ratio meter to my Carlton a couple of years ago : -

http://totalcarlton.com/forums/index.php/topic,10017.msg80116.html#msg80116

They are still available : -

http://www.jaycarelectronics.co.uk/productView.asp?ID=KC5300&keywords=kc5300&form=KEYWORD

Under heavy load, for instance, climbing at high speed up the long climb on the motorway up to Saddleworth Moor
the meter still shows that the ECU is responding to the Lambda sensor output and maintaining a 14.7 to 1 stoichiometric ratio.

It regularly shows 'L' (lean) on moderately high speed overrun when the ECU shuts the fuel off and has only once shown 'R' (rich) when asking for high acceleration with a cold engine (that's a silly way to mistreat an engine anyway) normal acceleration figures for a warm engine show that mixture drops to 12 to 1

[turbojo]

Hi, I had never seen a narrowband A/R meter, thats a new one on me though it can't be very accurate as a narowband has a very non-linear response, though it would give a good guide I would have thought.
with analog meters it depends if it's of the high resistance variety, most new one's are.
My 24v stays in closed loop for the majority of the time, almost to the floor if under 4k as I said earlier.
12:1 is a rich condition, and is in keeping to what I said earlier as well, it's mostly turbo'd jap stuff that can get down to 10:1 which is why FCD's can be used on them as there is room for a little extra boost.

[melinx]

with analog meters it depends if it's of the high resistance variety, most new one's are.

The meter needs to have around 10 megohms input impedance: ANALOG meters with that high an input impedance cost around £300 !

[turbojo]

er, 20k/v is fine for checking them and one can be had for £15. As such narrowbands don't mind being shorted for long periods of time anyway, wideband, now thats a different story. Either way they can be checked cheaply and all carltons run openloop at wot which is what this was all about. I can give you a 2hour readout of my carlton with rpm, load, injector pulse duration and lambda all on the same graph. Actually i can post dozens of them from various cars all of which show forced changes to the mixture and how the petrol ecu responds under different conditions. Also if i change the lpg pressure at wot it makes naff all difference to the petrol injector duration where as at POT the duration changes as the STFT make adjustments. Anyway nuf said as i don't want an argument!

[melinx]

er, 20k/v is fine for checking them and one can be had for £15.

I know you have said "nuf said" but I have to dispute that statement because it's misleading

The specification sheet for the Bosch LSM 11 "Lambda" Oxygen sensor says "Load current max. +/- 1 uA"

A 20k/volt meter would draw 45 uA at the maximum 0.9 Volt output from the sensor.

http://www.bosch.se/content/language1/downloads/Sensorkatalog_sidor_58_till_60.pdf

[melinx]

See Lambda Sensor Testing in the 'How To' section

" Cheap voltmeters will not give accurate results because they load down the circuit and absorb the voltage that they are attempting to measure. A acceptable value is 1,000,000 ohms/volt or more on the DC voltage. Most (if not all) digital voltmeters meet this need. Few (if any) non-powered analog (needle style) voltmeters do. Check the specs for your meter to find out. "