The Lucas ECU - adaptive or learning abilities?
The Lucas ECU - adaptive or learning abilities?
Author
Discussion

pupp

Original Poster:

12,972 posts

302 months

Saturday 24th June 2006
quotequote all
Anyone able to give me a quick rundown on what parameters the ECU 'adapts' or 'learns' after a reset? And over how long? Within what range? Or point me in the right direction if this has been covered elsewhere.

I disconnected mine for a while last night and the car has completely changed character now after reconnecting... no fault codes displaying and nothing obviously 'wrong' with its running behaviour, just very different and I'm wondering if it'll drift back to how it was?

I seem to remember reading something that said a particular temperature had to be reached to enable fully effective adaption?

Boosted LS1

21,200 posts

290 months

Saturday 24th June 2006
quotequote all
If it's anything like a saab you'll need to load it up and do some hardish driving

Boosted.

shpub

8,507 posts

302 months

Sunday 25th June 2006
quotequote all
Pages 108-110, 140-143 in the bible. Adaption starts when the water temp hist 90. below that it runs on the reset defaults.

brummiewedge

5,284 posts

251 months

Sunday 25th June 2006
quotequote all
I see, i tried a white tuning resistor in mine and as soon as it warmed up it died the first time i dug it into the carpet, put the green back in and it ran. Do you also know what tune resistors do what? Do they alter the same map, or do they put it on a completely differant one??

paulathome

686 posts

248 months

Sunday 25th June 2006
quotequote all
One tune resister is for pre cat engines.
The other for catalytic converter fitted cars.
Hope that answers your question.
Paul.

brummiewedge

5,284 posts

251 months

Sunday 25th June 2006
quotequote all
was looking for a bit more detail as to WHAT it does to the map. I know pre cat is green (UK non cat), and the white ( UK with cat ) was the original one with cats, does it just move the fuel up or down on the same curve etc?

trackcar

6,453 posts

256 months

Sunday 25th June 2006
quotequote all
changing the tune resistor changes the actual map, there's two on the chip (and a fair bit of free space lol).
The learning abilities of the lucas are very very poor, it's an ancient ecu so was probably state of the art when introduced though. John ravenscroft talked me through the ecu capabilities when i was with him once at the factory, I seem to remember he didnt rate it even back then, over a decade ago now..

essentially the ecu learns ONE adaptive value, at tickover. The engine speed must be stable for several minutes (best thing is probably to leave the car idling and go and make a cup of coffee), the temp must be high (90degrees sounds about right, if the cooling fan is cycling in and ot it's hot enough). the throttle pot signal must be stable ie no blipping of throttle. If you have the ecu software you can then watch the adaptive value changing as the ecu trims the idle mixute off the lambda sensors.

the ecu then applies this value across the whole RPM range .. so if you have any issues at idle you screw up the whole fuelling scale. Any lambda trim in real time is then applied over and above this alredy-trimmed fuel map number.

it's such a poor system it's amazing it works as well (or badly depending on your viewpoint) as it does ..

GreenV8S

31,021 posts

314 months

Sunday 25th June 2006
quotequote all
trackcar said:
there's two on the chip

I think there are more than that, five rings a bell?

trackcar

6,453 posts

256 months

Sunday 25th June 2006
quotequote all
GreenV8S said:
trackcar said:
there's two on the chip

I think there are more than that, five rings a bell?


hmm .. I only found two .. I have a neat little emulator prog that highlights ascending numbers so it's easy to find map addresses. This was on a std TVR chip. If there are other maps in there they're very fragmented. I will look at my Tornado chip later and see how many i can find in there, i seem to recall MA saying he only puts 2 into his chips though aswell?

chassis 33

6,194 posts

312 months

Sunday 25th June 2006
quotequote all
From a thread about 6-9months ago I seem to remember Mark saying theres space for 6 ECU maps if required, you can get 5tune resistors for the CUX, red, yellow, green, white and blue, so I guess no tune resistor will give the 6th map, however this may only apply to Landrover ECU's where the same ECU was sent all over the world.

Regards
Iain

Griffter

4,143 posts

285 months

Sunday 25th June 2006
quotequote all
Can anyone confirm that the ECU is the same for any 14CUX unit, regardless of application - any vehicle-specific changes (eg to the map) are effected ONLY in the EPROM chip (and courtesy of the tune resistor)?

Edited by Griffter on Sunday 25th June 17:04

GreenV8S

31,021 posts

314 months

Sunday 25th June 2006
quotequote all
Griffter said:
Can anyone confirm that the ECU is the same for any 14CUX unit, regardless of application - any vehicle-specific changes (eg to the map) are effected ONLY in the EPROM chip (and courtesy of the tune resistor)?


That's my understanding.

Cider Andy

1,889 posts

255 months

Sunday 25th June 2006
quotequote all
Griffter said:
Can anyone confirm that the ECU is the same for any 14CUX unit, regardless of application - any vehicle-specific changes (eg to the map) are effected ONLY in the EPROM chip (and courtesy of the tune resistor)?


Sounds right to me, too. As Lucas made the same parts for different car manufacturers, their products usually contain that manufacturer's part number, as well as the Lucas part number. And it may be that the Lucas part number is different between two identical items supplied as OEM kit for different car makers (it certainly looks that way for the 5AM airflow mass meter). But it may also apply to different revision numbers among similar Lucas hardware so beware! Certainly I've seen second-hand 14CUX ECUs for sale on eBay from the same supplier at different prices for units with different part numbers.

Most likely, I would think the same ECU with different firmware (or chips) would carry a different part number, so the ECU hardware itself would be identical.

shpub

8,507 posts

302 months

Monday 26th June 2006
quotequote all
'Fraid the reality is slightly different. There have been several revisions to the units over the years including the provision of new/additional signals. Look at the various variations of 14CUX wiring diagrams. The Eprom software also varies and its revision is typically compatible with the revision of hardware that it is using.

Been playing with this to get the Griff through its emissions. Take an Eprom and plug it into different ECUs and low and behold different emisssions and running characteristics. OK I ve a modded engine which was causing the issues but I was surprised at this. Looking at the ECU itself and there were some differences. It took two days of testing with a Gas analyser to get it sorted.

Eventually found a combination that worked and got through. Generally the Eproms will work in any ECU - engine will start etc but it may not be running in a the way you think it is.

trackcar

6,453 posts

256 months

Monday 26th June 2006
quotequote all
hmm that's interesting .. you'd think a fuel number is a fuel number is a fuel number for any 14cux.

shpub

8,507 posts

302 months

Monday 26th June 2006
quotequote all
Yes but a fuel number is only part of it... The lickle computer still has to convert it to an electrical pulse or work out where in the map it should be taking those values. Those algorithms were often tweaked.