Base idle adjustment

[phill01]

I have a 4.0L 1997 it always starts from cold ok but as you drive off the revs rise to near to 2,000rpm which is too high. When warm it normally ticks over just under 1,000rpm I say normally because it sometimes speeds up to close to 2,000 again? Even when it does tick over at 1,000 it is still slow in responding, ie. when you pull up,dipping the clutch (of course) the revs fall to 1,500 then after a couple of seconds being stationary they finaly drop to just below 1,000. I have another car with the same rover V8 with SU carbs and standard points ignition and it ticks over at a steady 600rpm, if I want it can be slowed to less than 500rpm....when I built the car I wanted everything back to basics so I am in control not the black box electronics ECU...Any ideas.

[zzr1200]

Sounds like a possible air leak, check all the hoses from the dizzy and manifold but try checking the idle air bypass control screw which should be about 1.5 turns out on a CAT car, mine was out 3 turns when I got it and it ran a lot better once I reset it. The other thing it may be is the coolant temp sensor(the one on the inlet manifold, mine (also a 1997) idles at about 1800rpm as it warms up and gradually drops to a normal idle within about 1 minute of start up.

Hope this helps.

Paul

[phill01]

Thanks for that but which is the idle bypass control screw? is it the screw near to the throtle linkage on the oposite side to the throtle position sensor? I have looked for any air leaks etc and everything apears to be in very good condition. The car has been serviced at Fernhurst at 24,000m last Feb. and has now covered 27,000m (I use the car every day) they cleaned the stepper motor and could not find any faults.

[greenv8s]

phill01 said:

Thanks for that but which is the idle bypass control screw? is it the screw near to the throtle linkage on the oposite side to the throtle position sensor? I have looked for any air leaks etc and everything apears to be in very good condition. The car has been serviced at Fernhurst at 24,000m last Feb. and has now covered 27,000m (I use the car every day) they cleaned the stepper motor and could not find any faults.

Yes that is the base idle screw. It should be set so that when the air bypass hose is squashed flat, the engine just barely lumps along at about 500 rpm.

Your idle control problems may also be caused by a faulty road speed sensor, or failure of just about any sensor in the fuel injection system. If the problem 'goes away' when you disconnect the stepper motor, at least you know it is a fault in the control system and not an air leak.

[HarryW]

Another minor point, but well worth doing, when you've reset the base idle setting is to clear the ECU adaptive memory of anything it learnt about tickover previously, i.e. disconnect it for 30secs.
It does make a diference as your adjustments might not be apparent straight away unitl its re-learnt again.

Harry

[K4TRV]

HarryW said:

Another minor point, but well worth doing, when you've reset the base idle setting is to clear the ECU adaptive memory of anything it learnt about tickover previously, i.e. disconnect it for 30secs. It does make a diference as your adjustments might not be apparent straight away unitl its re-learnt again. Harry

The Rover/TVR ECU DOES NOT HAVE ADAPTIVE LEARNING (or an adaptive memory !!!) - It has limp-home modes which are triggered by parameters going outside the envelope and these generate Fault Codes.

Re-setting the ECU clears these fault codes, but if the parameter/sensor is still faulty, the code will be generated again.

I certaily DOES NOT re-learn anything, except when Mr Adams re-programmes it !!!!

Trev McM
Pedant !!

[HarryW]

Fair point there K4, its probably more of a habit of mine, that when anything is changed then faults are cleared just incase they mask what you have just done.
I know the Lucas/ ECU is not the brightest of compts but it does an OK job under close loop all things considered.

Harry

[greenv8s]

I believe the 14 CUX DOES have adaptive fuelling, although it is relatively crude and nothing like the self-mapping capability that you get from GEMs and the like. As I understand it, it has a simple adaptive setting that is updated when certain specific conditions have been met, I don't remember what the exact conditions are but as I recall they include coolant temperature and lambda readings.

[K4TRV]

HarryW said:

Fair point there K4, its probably more of a habit of mine, that when anything is changed then faults are cleared just incase they mask what you have just done. I know the Lucas/ ECU is not the brightest of compts but it does an OK job under close loop all things considered. Harry

Harry, Totally agree that it does an OK job. I just get touchy when people start using 20th Centuary IT terms on the ECU.

Trev

Green, It is true that the ECU uses the logic that being able to revert to a "limp-home mode" because something goes outside a fixed range can be called "adaptive". If you specify a number of alternative modes following the excursion outside of the set parameters, it can "look-like" adaptive performance. Crude, but probably what the 14 CUX does - just don't believe it is "really" adaptive in the 3-d mapping/fuzzi-logic sense.

I think we agree, just not adaptive


Trev McM

[shpub]

greenv8s said:

I believe the 14 CUX DOES have adaptive fuelling, although it is relatively crude and nothing like the self-mapping capability that you get from GEMs and the like. As I understand it, it has a simple adaptive setting that is updated when certain specific conditions have been met, I don't remember what the exact conditions are but as I recall they include coolant temperature and lambda readings.

Yes around 90 degrees coolent temp. The adaption or (to be more PC) changes it makes adapts or (to be more PC) changes the mapping based on the information it receives from its sensors. It is crude but it does change the maps by adapting or should I say changing the way the look ups are done. By the way many of the adaptive multi-D systems do the same thing but only a little bit smarter. Very few ECUs have the horsepower to do the calculations on the fly and they result to using look up tables to get the various paramters. The adaption is usually done by changing the way the tables are looked up or by processing the results afterwards. What it can't do is measure the engine's health and perform its own mapping. Many of the ECUs can now but they still rely on look up tables for the main fuelling and ignition data.

BTW When I managed the applications team that supported Lucas during the ECU development... this feature was referred to with the a word
PS that was some 20 years ago...

>> Edited by shpub on Wednesday 7th April 13:47

[K4TRV]

shpub said:

greenv8s said: I believe the 14 CUX DOES have adaptive fuelling, although it is relatively crude and nothing like the self-mapping capability that you get from GEMs and the like. As I understand it, it has a simple adaptive setting that is updated when certain specific conditions have been met, I don't remember what the exact conditions are but as I recall they include coolant temperature and lambda readings. Yes around 90 degrees coolent temp. The adaption or (to be more PC) changes it makes adapts or (to be more PC) changes the mapping based on the information it receives from its sensors. It is crude but it does change the maps by adapting or should I say changing the way the look ups are done. By the way many of the adaptive multi-D systems do the same thing but only a little bit smarter. Very few ECUs have the horsepower to do the calculations on the fly and they result to using look up tables to get the various paramters. The adaption is usually done by changing the way the tables are looked up or by processing the results afterwards. What it can't do is measure the engine's health and perform its own mapping. Many of the ECUs can now but they still rely on look up tables for the main fuelling and ignition data. BTW When I managed the applications team that supported Lucas during the ECU development... this feature was referred to with the a word PS that was some 20 years ago... >> Edited by shpub on Wednesday 7th April 13:47

I think we said it does a good job - doesn't time fly and when you look back, it actually did an even better job that we thought!!

Tks for that clear explanation. I still twitch at the "a" word tho'....;-))

Trev McM

[greenv8s]

Since mine has been mapped it should be spot on using the default adaptive setting and any other setting is likely to be worse, so I tend to think of the adaptive 'feature' as an extra way for it to go wrong! In my case resetting the ECU to clear everything back to the defaults is generally the right answer. For most people, this is something you should only do if the ECU has registered a fault and reverted to limp home mode, after you have corrected the fault.

PS in discussions with Mark Adams I have got the impression that the 14CUX actually compares very favorably with many much more recent ECUs and you need to spend a lot of money on an aftermarket ECU to equal the performance of the trusty old 14 CUX.

>> Edited by greenv8s on Wednesday 7th April 15:52

[joospeed]

K4TRV said:

The Rover/TVR ECU DOES NOT HAVE ADAPTIVE LEARNING (or an adaptive memory !!!) Trev McM Pedant !!

[panto on] oh yes it does [/panto off]

from john ravenscorft at the factory ..
*the lucas ecu will learn ONE trim value for fuelling and applies this across the whole rev range. The car must be at a standstill, it must be up to temperature (can't remember exact but over 80 deg i would think) ie cutting in and out on the fan switch. The throttle must be at idle setting and no faults must be present. The ecu then trims the idle fuelling back to lambda 1, and this percentage trim value is then applied to the rest of the rev range.

So yup, it does adapt. Yup it is very crude (JR was quite dismissive of the system compared to modern alternatives).

[HarryW]

joospeed said:

K4TRV said: The Rover/TVR ECU DOES NOT HAVE ADAPTIVE LEARNING (or an adaptive memory !!!) Trev McM Pedant !! [panto on] oh yes it does [/panto off] from john ravenscorft at the factory .. *the lucas ecu will learn ONE trim value for fuelling and applies this across the whole rev range. The car must be at a standstill, it must be up to temperature (can't remember exact but over 80 deg i would think) ie cutting in and out on the fan switch. The throttle must be at idle setting and no faults must be present. The ecu then trims the idle fuelling back to lambda 1, and this percentage trim value is then applied to the rest of the rev range. So yup, it does adapt. Yup it is very crude (JR was quite dismissive of the system compared to modern alternatives).

Glad you chipped in with that Joolz, as I wasn't too sure of my ground. However I was advised by The Man (Mark Adams) to always clear the ECU memory after any base idle adjustment and let it run until the fans cut in. Makes sense now .

Harry

[k4trv]

HarryW said:

joospeed said: K4TRV said: The Rover/TVR ECU DOES NOT HAVE ADAPTIVE LEARNING (or an adaptive memory !!!) Trev McM Pedant !! [panto on] oh yes it does [/panto off] from john ravenscorft at the factory .. *the lucas ecu will learn ONE trim value for fuelling and applies this across the whole rev range. The car must be at a standstill, it must be up to temperature (can't remember exact but over 80 deg i would think) ie cutting in and out on the fan switch. The throttle must be at idle setting and no faults must be present. The ecu then trims the idle fuelling back to lambda 1, and this percentage trim value is then applied to the rest of the rev range. So yup, it does adapt. Yup it is very crude (JR was quite dismissive of the system compared to modern alternatives). Glad you chipped in with that Joolz, as I wasn't too sure of my ground. However I was advised by The Man (Mark Adams) to always clear the ECU memory after any base idle adjustment and let it run until the fans cut in. Makes sense now . Harry

The word is "very crude" - you've all said it.

I think it's quite inappropriate to think the TVR ECU is "Adaptive".

As I said earlier, if you specify an alternative mode following the excursion outside of the set parameters, or to match a criteria (As Jooz said)it can "look-like" adaptive performance. Crude, but probably what the 14 CUX does - just don't believe it is "really" adaptive in the 3-d mapping/fuzzi-logic sense.

Just stop thinking that the ECU is adaptive which is synonamous with Fuzzi logic and 3-d mapping and your there!!

The ECU really does work, but it is not adaptive in today's meaning of current ECUs.

Trev, still a pedant ;-)) and similing!!

[HarryW]

Oi pedant NO

H

>> Edited by HarryW on Wednesday 7th April 22:39

[JonRB]

I've got the same problem on my 1997 Chimaera 500.

It was idling too slow and stalling, so Dave Batty adjusted it to idle a little faster. Ever since I have had this problem occasionally - not sure of the circumstances but generally after "pressing on" so it may be something to do with temperature.

I have no idea if the ECU was reset, so may do this.

Good job the car isn't due an MOT until August because according to the printout I had on one of my other cars that has just had an MOT, the emission test says that the acceptable paramters for normal idle is 450 - 1500 rpm.

[GreenV8S]

I suspect we've just about done this one to bits now, but I'll just follow up one possible misunderstanding from your last post.

The ECU has two maps, the 'normal' map is normally set up for closed loop operation and is used when the ECU thinks everything is working, the alternate map is set up open loop and is used when the ECU recognises a fault which prevents it working in closed loop mode.

Quite separate from this, the ECU has the trim setting which Joolz described. The trim setting is the bit I'd call adaptive, and yes I think we can agree it is crude and barely deserving that word but strictly speaking it IS adaptive.

I just mentin this because you have dismissed the 'adaptive'-ness as just being a switch between two maps, and that isn't what we're talking about.

[K4TRV]

GreenV8S said:

I suspect we've just about done this one to bits now, but I'll just follow up one possible misunderstanding from your last post. The ECU has two maps, the 'normal' map is normally set up for closed loop operation and is used when the ECU thinks everything is working, the alternate map is set up open loop and is used when the ECU recognises a fault which prevents it working in closed loop mode. Quite separate from this, the ECU has the trim setting which Joolz described. The trim setting is the bit I'd call adaptive, and yes I think we can agree it is crude and barely deserving that word but strictly speaking it IS adaptive. I just mentin this because you have dismissed the 'adaptive'-ness as just being a switch between two maps, and that isn't what we're talking about.

Green V8S, Done to death I fear. Its a good ECU, that I have no qualms about, and good to know how the closed/open loop elements fit together.

Sorry HarryW, once a pedant......;-p

Trev McM

[phill01]

Thanks to you all for providing all this information, I pleased to say with your help the fault was simply a poorly adjusted bypass control. Adjusting the screw to 1.5 turns out from fully screwed in has solved the problem.

Thanks again to you all, from phill01.