Freelander 1 Jatco auto gearbox problem

Hippo said:

Removing fuse 4 would remove power from the auto's computer but I don't see how this would make a difference?

The Main relay RL4 in the engine bay feeds power to the auto's computer via fuse 4 when RL4 is on. RL4 is only on when the engine computer requests, when the ignition is on. When the ignition is switched oft RL4 stays on for something like 3 minutes (I was playing with it recently on mine but can't remember exact time) then switches oft, and the engine computer remains powered. Rave confirms this and says it allows time for error codes to be stored in the eeprom. but that's sort of thing is don't within a blink.

As the engine computer is permanently powered by a different fuse, I don't see what removing fuse 4 would do that's different, other than the engine being in "ignition on" mode as opposed to "ignition oft" or "sleep" or "alarm on" mode, if these conditions exist. Removing fuse 4 and therefore killing power to the auto's computer is what the engine computer does anyway, when shutting things down after the ignition is oft.

Click to expand...

Interesting. When I replaced the fuse I heard a definite "click" for want of a better term so something was energised, ignition was off. Still changing perfectly. It's a 51 plate - does this make a difference?

TF

Maybe the auto's ECU doesn't store adaptive information after power disconnection?
I know the engine ECU doesn't loose it's adaptive information, even after extended periods of disconnection.

Rave refers to the auto's computer storing fault codes etc in an eeprom. It also has some other programmed info which the hawkeye can see, which is probably in the same eeprom. Adaptive info is one thing but where the control/response of the auto is feked up to the extent somehow "clearing" it with a power down would put it back to default set-up is something the programmer should protect against. Needing a reset would indicate it's somehow trying to work off parameters which don't work, or derived from "calculated" data which points it to do silly things. I would hope any working "magic" it has to look at adaptive stuff would not allow the auto to do daft things beyond the limits of good operation.

Removing the adaptive information would only restore the ECU to its factory settings. It would work just fine, maybe better if some corrupted adaptive information is stored in the ECU.
If a new engine or transmission is fitted, the adaptive information should be reset anyway. This insures the ECU only adapts to current values not old or irrelevant ones.

I would hope the adaptive info wouldn't put the auto into a state where harsh or incorrect changes were possible.

Hippo said:

I would hope the adaptive info wouldn't put the auto into a state where harsh or incorrect changes were possible.

Click to expand...

Sometimes an erroneous adaptive value can be stored and read as a correct value. This happens with engine ECU's, so I can't see why it couldn't happen with the gearbox ECU.
This is why the adaptive values can be reset.

GrumpyGel said:

You say the problem is when the box is gong down from 3 to 2 - is this because the car is slowing or is it kick-down to increase speed?

If its when slowing down, do the problems occur when you still have some throttle on or when your foot's off the accelerator completely? If it its when your foot is off the accelerator could the problem be related to engine idle speed? This is affected by having the air con on, so if your idle isn't too flash when the aircon is on, maybe its confusing the box's ecu? Might be something completely unrelated to the box - eg air flow sensor.

I should qualify these thoughts by adding that I haven't got a clue what I'm talking about!

Click to expand...

No not kick down. When it's slowing down with foot oft the accelerator and foot on or oft the brake as it slows. When it changes from 3 to 2 there's a jerk. Not always, but more likely to happen with the air con on than oft. When the air con is oft it still happens but a lot less than when it's on. Air con fans are powered via the same fuse as the auto's computer/sensors and engine sensors etc. I have compared sensor readings with me hawkeye taken previously to current live data readings and can't see enough of a difference in values to suggest a fault or direction to work in. The accelerator is progressive in it's reading on the hawkeye but I appreciate this isn't fool proof and no datum to work to. I may play with this some more or take a guess and replace the pedal switch as they're cheap. Fault happens less often after taking the engine fuse box apart and squeezing everything in there. Having to wait longer between playing to see if it's cleared completely or not now, before tinkering again. If I change too many things at the same time I'll lose the logical approach and won't know what solved it, if I resolve the problem.

Nodge68 said:

I suspect it's something to do with light throttle and varied loads on the engine. Mine has a slightly thumpy change when cold while going down hill.
The UK spec V6 doesn't have an air flow sensor, it uses a map sensor instead

Click to expand...

I have been looking at accelerator throttle response on me hawkeye. It's progressive and smooth. I have similar thoughts about loading and powering the engine to suit the load. I may take the acoustic cover oft and look at all the sensors again to see if they'e lose/wet/something. It's a good job I recorded all the menu's with the engine oft cold, oft warm, running cold and warm, and at 60mph. Certainly handy to have for comparison.

Epicuser said:

I have a jerk on my auto going from 4th to 3rd. It only happens when I use engine braking to slow down, so have foot slightly on the gas. It gets down to about 1100rpm then jerk changes down to 3rd. Happens every time under these circumstances, and seeing as I drive in 30mph limits too much of the time, it happens lots.

It doesn't do it when I take foot off gas and brake. Certainly nowhere near as noticeable.

Click to expand...

I've changed pedal sensors before due to dash light coming on. I found they didn't fail as expected. The switch seemed to have a transient period when moving. The contact slides normally but occasionally would open as the switch bent. I may take the switches oft and mess with then and refit. My fault is similar to what you describe. Mines on down change from 3 to 2 I think. I'm running LRN402 in the auto so it's the proper stuff. I was changing it more than once a year at one point. Reason being it seemed to help resolve the fault for a while then it slowly came back, getting worse over time until I changed the fluid again. I haven't changed it for some time now. So I think the fluid change was making the fault less noticeable as opposed to fixing the problem proper.

This auto fixing things has been going on for some time now. I'm busy with other stuff at the same time. I also have to wait to see if my tinkering resolves the problem as I've managed to get the fault to happen a lot less often now. That's good but you have to wait longer to see if the fault is still there.

One of the things I was concerned about was the occasional bump when putting it into drive or reverse, or taking it out of drive or reverse. I wondered if the rear diff was jumping about but it wasn't. I've seen the pinion/ird rock up or down when pulling away (filmed it) so that got me thinking about the engine mounts. In particular the lower tie bar at the bottom rear which connects to the subframe. It seems to be a common failure part.

I also had a vibration on the steering wheel at about 50+ mph. On certain roads this was 40+ mph but I assumed this was the road itself. It may well have been a contributing factor or a cause of the vibration coming out as it would happen on certain roads at certain places.

I changed from grabber AT2's to Pirelli scorpions long before the current auto fault popped up. I hated the rumble from the AT2's from day 1. Especially when venturing to bala 4x4 in Wales for pay and play. 60mph roads were orrible. Comparing the ride at 70mph in ma hippo to a new Freelander 2 proved ma hippo had problems. I know the FL2 has betterer suspension but cruising at speed should be comfortable in both vehicles. The tyre change reduced the road noise and vibrations but there was still something I felt was wrong, but not enough to pinpoint as it didn't happen all the time. In the end I filmed the engine while starting/stopping and pulling away. There wasn't much movement other that what you would expect. I started waiting 5 seconds between gear lever movements from R to N, then to D as opposed to R straight to D. I was thinking this may give the auto and it's pump more time to feed fluid to where it's needed. This sort of made the "change into gear" bump less or unnoticeable. But it was still there. This problem was cured for a while with an auto fluid change.

In the end the fault was tracked down to the lower tie bar which supports the bottom rear of the engine. I tried rocking the engine and pushing ma hippo with it in gear and couldn't see much movement on the engine or bar. It's a heavy car so not easy. I bought a new lower tie bar (I know you can polly bush them but I wanted a new propper one and don't like the idea of using poly bush) and played with it. The comparison showed my old bar was faulty but only when removed from ma hippo. This surprised me but after fitting the new one ma hippo feels smoother and more luxurious to drive. I didn't realise there was a constant background vibration as it was always there and the ride got betterer after the AT2 tyres were removed so I assumed that was how it should be. Video below of the old and new mounts for comparison.

Freelander 1 - Lower tie bar failed - YouTube

I will continue with my search for the auto problem and keep this fred alive with updates.

Been thinking about me fault some more and looking in rave and found the answer to the question of saving data as below.

...The EAT ECU uses a 'flash' Electronic Erasable Programmable Read Only Memory (EEPROM). This enables a new or replacement EAT ECU to be externally configured. EEPROM also allows the EAT ECU to be updated with new information and market specific data.

To input new information and market specific data the EAT ECU must be configured using TestBook. The EEPROM allows the ECU to be reconfigured as many times as necessary to meet changing specifications and legislation.

The EAT ECU memorises the signal values of the gearbox sensors and actuators. These stored values ensure optimum gearbox performance is achieved at all times.

This information is lost if battery voltage is too low, for example if the battery becomes discharged. The EAT ECU reverts to default readings on first engine start after a battery discharge or disconnection. The EEPROM facility in the ECU allows the stored values to be re-learnt, ensuring optimum gearbox performance.

If these signals are not within the EAT ECU stored parameters, the ECU will make adjustments to the operation of the gearbox through the actuators to provide optimum driveability and performance.

The inputs from the sensors constantly updates the EAT ECU with the current operating condition of both the gearbox and the engine. The ECU compares this current information with mapped information stored within its memory.

The ECU will make any required adjustment to the operation of the gearbox through the following actuators:

Gear control solenoid valves
Lock-up solenoid valve
Line pressure solenoid valve.

The EAT ECU also interfaces with the following:

Engine Control Module (ECM) via the CAN
Instrument pack via the CAN
Diagnostic socket via the ISO 9141 K line.

So the TCU does learn the gearbox parameters. These parameters are lost if the battery voltage is removed so implementing it's default values.
If the TCU detects an error in its adaptive values, it corrects them.
So disconnecting the battery might be worth a try.

I've disconnected the battery many times whilst playing with it and no change.

I can't see how the auto's computer stays powered when fuse F4 loses power when the main relay switches oft. There must be another power supply to keep the auto's computer alive if the above rave description is correct. Unless I'm missing something in the circuit.

So, how do we reset the gearbox back to default them?

Good question.

The computer uses an eeprom. This is a memory device which the computer can store information in. When power is removed the eeprom remembers the data. So disconnecting the battery won't make a difference to its content.

The eeprom seems to store "market specific data" which can be changed using LR's T4 test book. I don't know what info this is but I can guess it's the serial number of the computer, date of manufacture, hardware revision, software revision etc. I eggspect there will be additional info which could include emission code if they need to get the auto/engine to work differently for different countries. My Hawkeye can't see this data.

Further to the above the computer can store data on "sensors and actuators". This achieves optimal performance of the auto which would suggest fine tuning of factory default data for the auto. I assume this is to create perfect gear change timing, gearbox braking, altering shift patterns etc but this is just a guess. The auto will have its own pre-set values stored in the eeprom and use these to start with when first powered. It will also have an "adjusted" set of data which is specific to that particular auto which it creates when first powered and then starts to use the auto. It's this data which is said to be lost if the auto loses power like disconnecting the battery or if battery voltage drops too low. This is the most important bit. 1 of a number of possibilities I can think of:

1. The auto stores the adjusted data in the eeprom and its therefore still available when the system looses power, but when re-connecting the battery the engine computer says to the auto's computer "I lost power so you have to reset yer data to factory settings anorl" when the auto's computer is next powered up.

2. There's a delay in storing updated adjusted data when the engine or ignition switches oft. If you disconnect the battery before the delay times out you lose the updated data or this action initiates a reset flag for when the auto's computer is next powered. If you don't disconnect the battery then data is stored in the eeprom after a certain amount of time and therefore protected for later use.

3. The auto's computer has another power feed which I haven't spotted.

4. Something else.

As said this is all guess work but based on the derived logic.

Time factor?

Ignition off, fuse out, ignition on. 10 mins later and ignition off fuse replaced.

So; if there is a back up it lasts a few minutes to hold whatever data there is, then it's lost. You probably wouldn't have noticed having the battery disconnected previously as there wasn't a problem with the changes?

TF.

The market dependent stuff will be things like. ..

Cruise control
Abs module type
3 door or 4 door
Sunroof
Horn sound
Auto or manual box
Auto door lock
Alarm beep or flash of lights
Etc

These can be programed into the ecu.....coding the ecu

Adaptive values are learnt over time by the engine ecu maybe the auto box too?...

On my vw just clearing the engine ecu of a fault code will reset all the adaptive values back to default...the car must then be driven for a while so that it re learns the adaptations.

There is also a procedure that resets them in the workshop involving the vw diags and reving of the engine among other things....

Its mostly to reset the lambda sensors to their readiness states..there are 4 inorl Also the egr and a few other things...I cant remember what they all are now.....

Here you go as an eggsample.....The Vw diags are brilliant. ..If only landys had similar

http://www.ross-tech.com/vag-com/tour/readiness.html

jedi said:

The market dependent stuff will be things like. ..

Cruise control
Abs module type
3 door or 4 door
Sunroof
Horn sound
Auto or manual box
Auto door lock
Alarm beep or flash of lights
Etc

These can be programed into the ecu.....coding the ecu

Adaptive values are learnt over time by the engine ecu maybe the auto box too?...

On my vw just clearing the engine ecu of a fault code will reset all the adaptive values back to default...the car must then be driven for a while so that it re learns the adaptations.

There is also a procedure that resets them in the workshop involving the vw diags and reving of the engine among other things....

Its mostly to reset the lambda sensors to their readiness states..there are 4 inorl Also the egr and a few other things...I cant remember what they all are now.....

Here you go as an eggsample.....The Vw diags are brilliant. ..If only landys had similar

Ross-Tech: VAG-COM Tour: Readiness

Click to expand...

Ah yes, the wonderful VCDS. I had a few people work on my Octavia mk 2 and my Yeti, performing magic that the dealers could not do

Hi Hippo,

just read yesterday you are having trouble with your auto. I called the LR guy who fixed mine today to refresh.
Did you check that the cables are ok? You know when I had trouble last year with my Jatco I changed all kind of solenoids, then a cable into the Jatco and still had same problem shifting up and down the gears. It was better but not really good. Then a LR dealer tried his luck, changed all kinds of solenoids again, check the cable I changed which was ok. He opened the gearbox and checked that as well. Always got sometimes good sometimes poor values while measuring which they couldn't explain. Then they took off the battery compartment and underneath was a cable damaged which damaged also one to the Jatco as far as I understood. It scratched open on some metal there. Both fixed and it worked again. Wouldn't know how to take it off though.

Kind Regards
Matthias

mbrokof said:

Hi Hippo,

just read yesterday you are having trouble with your auto. I called the LR guy who fixed mine today to refresh.
Did you check that the cables are ok? You know when I had trouble last year with my Jatco I changed all kind of solenoids, then a cable into the Jatco and still had same problem shifting up and down the gears. It was better but not really good. Then a LR dealer tried his luck, changed all kinds of solenoids again, check the cable I changed which was ok. He opened the gearbox and checked that as well. Always got sometimes good sometimes poor values while measuring which they couldn't explain. Then they took off the battery compartment and underneath was a cable damaged which damaged also one to the Jatco as far as I understood. Both fixed and it worked again. Wouldn't know how to take it off though.

Kind Regards
Matthias

Click to expand...

Hi. I've looked into this and I think my auto's wires are ok. My barrel connectors are tie wrapped together just in case they come apart again. I too have had the battery etc out so I could see the wiring loom and all looks ok. My resistance values are solid and don't change. Problem is the fault is intermittent, but I've been working my way through stuff and it's become less intermittent. I think it's down to loose connections somewhere. More likely several and the more I find and cure them, the less often the average loose or bad connection causes the fault to occur. Some of which I have probably corrected but you don't realise until you notice the average number of "fault occasions" drop over time.

Tempus Fugit said:

Time factor?

Ignition off, fuse out, ignition on. 10 mins later and ignition off fuse replaced.

So; if there is a back up it lasts a few minutes to hold whatever data there is, then it's lost. You probably wouldn't have noticed having the battery disconnected previously as there wasn't a problem with the changes?

TF.

Click to expand...

I don't think the auto's computer has a battery backup inside. It stores data to eeprom which doesn't lose data when power is removed. This is a betterer option than battery backup and it's essentially permanent unless you get a hardware failure or over write it. So the extra cost of an internal battery wouldn't be required.

The method you describe is similar to method 1 but the battery is connected after the fuse removed so the Freelander is powered, but with no auto computer power while the ignition is switched on to position 2. I haven't tried that on mine that I can remember. This fault has been going on a for over a year and steadily getting fixed.

A lot of people advise to allow electronic stuff to power down for several minutes as they think it's still powered inside due to resedule voltages but this more often isn't the case, after the initial power off + a few seconds. There may be some filtering including capacitors etc inside the computers case which would still have some power after the fuse is removed but this decays within seconds as opposed to minutes. If anything the battery being connected without the auto computer when the ignition is on position 2 should be enough to trigger a "reset" command when next powered as below.

The only time a required delay in power loss is required is if the engine computer is programmed to see loss of the auto computer for a required N seconds/minutes, and then it initiates a "cold reset" of the auto's "adaptive" info in order to restore default details when it's next powered. I would be surprised if this were the case as car computers are normally quite rugged in their operation, meaning they can cope with power being removed or part of the system loosing power suddenly, and cope ok when it's all powered again. Sometimes that may mean powering the whole lot off then on to be ok. If this power off mechanism is used then it would be initiated within seconds as opposed to minutes.

I think the adaptive info thing is a red herring in relation to my fault. All the adaptive stuff does it fine tune wear or production tolerances. It can also look at things like average % acceleration and hold gears longer/shorter depending on pedal demand, and apply guess work to make the car more responsive depending on average historic use. Things like shift patterns can be modified to do this. Holding gears on acceleration etc. Rave doesn't go into detail on this so we don't really know but it's an interesting topic. If adaptive info is cleared I don't think an owner/driver would notice the reset of any adaptive config.

jedi said:

The market dependent stuff will be things like. ..

Cruise control
Abs module type
3 door or 4 door
Sunroof
Horn sound
Auto or manual box
Auto door lock
Alarm beep or flash of lights
Etc

These can be programed into the ecu.....coding the ecu

Adaptive values are learnt over time by the engine ecu maybe the auto box too?...

On my vw just clearing the engine ecu of a fault code will reset all the adaptive values back to default...the car must then be driven for a while so that it re learns the adaptations.

There is also a procedure that resets them in the workshop involving the vw diags and reving of the engine among other things....

Its mostly to reset the lambda sensors to their readiness states..there are 4 inorl Also the egr and a few other things...I cant remember what they all are now.....

Here you go as an eggsample.....The Vw diags are brilliant. ..If only landys had similar

Ross-Tech: VAG-COM Tour: Readiness

Click to expand...

Ma hippo was repatriated from Japan. When it went for its first service at a main dealer (I still remember the service girl after 10 years) I asked them to look all over it and do a few additional things I'd spotted like enable the volumetric (I fitted the alarm senor in the roof meself) and alarm activation when key/fob locked. I assume they don't nick things in Japan. They don't fit the internal CDL lock all doors switch anorl but the wiring harness is behind the blank. The dealer spotted the "territory" settings were wrong and set them to the correct version. I assume this is UK spec as opposed to Japan spec. I guess it includes emissions related stuff anorl as that's links to wear.

My Hawkeye shows some settings but probably not all. Things like 3 or 5 door are detected by the vin number which the computer is programmed with. If it's a 5 door it ignores the "roof fitted/not fitted" signal for example.

We know the engine has adaptive stuff as it's required due to tolerances in build and wear and tear. All engine computers have this I assume. We think the auto has adaptive stuff too and this may be similar to the engine, but probably still not noticeable in operation to a Freelander owner/driver. Rave refers to it but doesn't go into detail.

The vag com stuff looks interesting. I will look up what they have when I have more time.