[minim]

I have this issue with EAS bumping up and down all the time not findings it's position while driving. It is doing this every 10 seconds when it first start to oscilliate. Even at idle (look in the log from 10min out) it constantly adjust a little but but then it's 4-5mm change. Still enough to keep the compressor going almost constantly. I've tried everything I can think of and read through all the guides and forums I could find without finding a solution so I'm trying here. Every suggestion will be appreciated

Front is going up/down between 124 and 105 in standard height and 104 to 88 in highway - 19bit and 16 bit / 29mm and 24mm
Rear is going up/down between 123 and 109 in standard height and 104 to 88 in highway - 14 bit and 16 / 21mm and 24mm

I have tried following:
-4x Brand new sensors from island 4x4 (2 x Rear Height Sensor 97-02 (OEM) ANR4687G £119.98 2 x Front Height Sensor 97-02 (OEM) ANR4686G £119.98 )
-Measuring constant 5V reference voltage at the sensor. No interference/bad supply showing when using oscilliscope to debug either. Anyways, sensors is just confirming the error so they do work from what I can figure out.
-Brand complete new valve block with solenoids, pressure sensor and so on..
-Overhauled compressor. Fills the tank from dead empty in 7-8 minutes with the door open (so it blocks the constant adjustment)
-Brand new EAS driver pack
-Used EAS ECU from ebay (could ofcourse have the same error as my old unit..)
-Calibrated new sensors with height blocks according to rave
-4x new air springs
-Cannot spot any air leaks with soap water. Also if I read off the sensor heights, open a door then they stay in place without going up/down. Compressor then also fills the tank and it stays full without the compressor starting again.
-If I let it stay on a level concrete floor it's constant adjusting so the compressor runs all the time. Not major adjustments but small and just enough to keep compressor going.
-Soldered the EAS connector down on the left/front, inside of the car in case of bad contact in the connector that is prone to water ingress there. two of the leads was black copper but I cut it untill it was nice and clean copper and solder+shrinkwrapped it.
-Visual inspection of the ECU connector looks good. Voltage supply at the ECU also looks good without disturbance.



Dump of the .CSV file. Just save the content and you can open it in Nanocom CSV Viewer, excel or similar
https://pastebin.com/eeLvs6Z4

 

[pwood999]

If it's constantly adjusting while idling & stationary on level floor, then my suspicion would be leak(s) that you haven't found yet ?

How long was the door open test ? Maybe a Nanocom recording while stationary might show something ? Do traces with door closed & open.

Was the valve block & solenoids really "brand new" or refurbished ? I have seen internal leaks (between NRV's) and around solenoid bases during bench tests, that don't show up when the valve block is in the car. Besides you cannot see soapy water bubbles underneath the block when it's in the car !!

 

[minim]

Yes but only by a little. Then it's just going up/down with 4-5mm. Enough to keep the kompressor starting/running much. I really can't see that it can be a leak. I can leave the car for 2 weeks without it dropping at all (I swapped the relay for a regular one so it doesn't adjust when ignition is off).

It takes 6-7 minutes from empty to the compressor stops. I refurbished the compressor and it's really effective after the fix. From 10minutes and out in the log it's statoinary. I can't spot anything then either. I would think that if there was bad sensor values it should show as irregular values in the log but who knows.

The block and solenoids is brand new from dunlop bought from island4x4. I tried to replace all the o-rings in my old unit and noticed bad NRV valves so I just bought a brand new unit to not have to think about that. Error is exactly the same as before. Basically the only old components in the system now is wires, ecu and air lines.

I had a friend log his vehicle and he actually had the same up/down problem on his but he didn't notice it before logging with nanocom. Maybe someone could do a log of 20 minute ride on a nice paved road and see if this is normal? It could be that the ECU just is programmed this crappy so it's constantly adjusting.

 

[pwood999]

What are the height values set to in the ECU ? Target values for L-R in each height setting should be within 6 bits of each other, or the ECU will continually hunt to level the car.

Read the "EAS System Info Document". I cant upload it here because the forum blocks large PDF files.

FYI, The ECU will always try to adjust the car to the target values . . . . but it also opens both the Front L-R valves together at intervals to balance the car by equalising the air-spring pressures. If the ECU settings are more then 6 bits different, the combination of L-R balance will always be followed by more adjustments !!


"After all doors are closed, systemleveling could take place if any height sensors actual value differs from its target value bymore than ± 2 bits. The target value is determined by the lowest sensor input. "

"The leveling operation opensboth front corner valves for only three seconds to allow pressures to equalize between thefront air springs. This will be repeated every time the vehicle speed drops to zero. In stop-and-go traffic a lot of extra valve activity may be heard. "

 

[95classiclwb]

https://www.usermanuals.tech/d/land-rover-range-rover-electronic-arir-suspension-rover-manual/index
this is the factory document explaining in detail how the system works, further info is available on RSW and Paulp38.com.

small changes are basically normal, the system is very advanced for its age but the same intelligence is its achilles heel. attempting to adjust 4 corners on solid axle with in exact heights while on uneven surface is nearly impossible to achieve, thus the constant level fight of the system.
as for the constant running of the compressor, you may have a system leak or the refurbishment of the compressor may have been a full success (it happened to me, I ended installing a new compressor).
seems you have addressed all possibilities of weak links by soldering connectors, renewing bellows, replacing sensors and block.
as you state you have ruled out all other possibilities. your log shows a right rear a bit over active, it is the longest pipe in the system it is possible it may have a pin hole leak, small enough as not to be heard but constant enough to provoke changes.

read the documents, and double check all your work.
best of luck.

 

[minim]

What are the height values set to in the ECU ?

Thanks for this!!! I've never seen this thing about 6bit. Where is this info? I've read through all the manuals I could find on this system without finding this. I have this difference on all from my notes. So I need to manually adjust them to be within 6bits? I'm tempted to just even them out for now and see if that solves the problem. I'm driving 350km tomorrow and would be nice to be able to do so without worrying for overheating the compressor.

I don't understand the logic behind this. When I set a target value they should scale to follow this target value. It doesn't make sense that we are allowed to set anything manually if there is some limit there. I would never program a feature like this for my customer in software we develope. That would be considered a bug.





I think I've read them all :S The constant running is caused by the constant adjusting. The compressor stops after 6-7 minutes from a completely empty system if I let a door stay open. If I keep a door open while driving it also stays at level. I can leave the car for 2 weeks without it dropping down so I can measure it from the wheel arches. I don't use the car a lot so I actually measured the wheel arches and remeasured after it had sat for 2 weeks in the garage and it was at the same millimeter still after the valve block replacement.

I will test what pwood999 suggested. This seems kinda logical but I could not find any info about this in the manuals.

 

[LarryS]

As long as your target values are within the expected range then calibration will do the rest (that's the point of course, imagine if it didn't, kind of leaves the technician and the vehicle stuck). If you look at a range of sample targets heights of real ecu's you will see differences with greater than 6 bits.

Only thing I can see is your 'Left front' consistently seems to be furthest away from the target height and 'dropping', again difficult to see just from data i would suspect its not holding pressure, swapping the springs and sensors - one at a time - left for right would be my suggestion.

 

[minim]

This is how I understood it also but I've tried every other option out there so I figured why not try that also so it's out of the question. I manually calibrated the sensors so they where within 2 bits off eachother and then I did a full calibration with calibration blocks again. Now they where within 1-2 bits on all stages so all looking good. New testrun and the result is EXACTLY the same as before. Constantly working adjusting up/down all the time.

I then figured I had to go back to basics. I logged this from when it was idle on a flat concrete floor. As you can see it is going up/down with roughly 5bit (1bit is 1.5mm approx I've been told). I noticed that most of the times it where going down when the compressor was stopping but that could be me just missing. According to the log the exhaust and inlet valve is opening/closing with 20seconds between and every other time it's exhaust or inlet so it confirms the height measurement. It's not far from target either so it's overshooting alot when it's adjusting. Almost like the opening time of the valves is way to long and the hysteresis for error is to little. Funny thing is that the log is showing the compressor running all the time but I had my hand on it clocking it to approx 10sec stop, 20sec running. So in reality when I do the other logs while driving it could be starting/stopping in the same way without me noticing.

Front left target is 116 and it goes between 114 and 119
Front right target is 116 and it goes between 115 and 120
Rear left target is 115 and it goes between 113 and 117
Rear right target is 115 and it goes between 114 and 118








Am I overthinking this? I work with PLC controls systems and program stuff like this for a living. A system adjusting/regulating this bad I would never let my hands off. To me this looks like bad tuning that will cause eccessive wear on the valves and compressor.

 

[ColoradoCrow]

Any chance your NRV could be sticking or have a hair caught in them/too much lube to cause an non proper air seal?

 

[minim]

I don’t know since I didn’t open the new block that came pre assembled. But I had this error before with the old block also and it didn’t change at all with a new block. It would be unlikely that a new block also had a issue with NRV valves but I guess I don’t know before I pull it open. How likely do you think it is ?

 

[ColoradoCrow]

Unless I have seen the NRV’s myself I wouldn’t trust anything. Especially since you have ruled so many things out.

 

[minim]

Ok. I don’t have time to do it before departure tomorrow but I will do it in the evening. Worth checking as you say but it does seem like the system is outputting wrong control (to long) signal to the valves for me. There is nothing between the valve and the spring so if the valve closed fast enough it should not overshoot like it does.

 

[pwood999]

I don't know if there were any ECU software revisions ? Maybe the hunting up & down is expected ? I just jump in & drive mine !! If it stops raining later, I will do the stationary test, and see if mine doe the same thing.

One other suggestion: Unscrew the solenoids on the top valves, and repeat the soapy water leak test directly on the valve itself. Also check if the valve block has the crappy orange o-rings, or the better black viton type.

 

[pwood999]

Also check this thread on LandyZone https://www.landyzone.co.uk/land-rover/constant-height-adjust-after-bag-change.163069/

Is your ECU version ANR4499 or RQT100040 ?

 

[minim]

Good tip and looked like just my error when I read about it. I checked this now and the old ECU did have a ANR number but the "new" used I got from ebay has RQT100040 Both have the same issue. I'm tempted to get a new unit but it would suck to spend that much without an improvement. I also have 60-smt bit when I leave it on bumpstops if I recall correctly so this should be good.


I replaced all o-rings with viton type in the old valve block and it didn't change the behaviour at all. I then moved around on the NRV to see if I could change anything but it didn't change. I then figured that it could still be worn parts inside so I decided to get a brand new OEM block from dunlop pre assembled and still the same error is present. Is it likely that both my old refurbished block and the brand new one with new internals/valves/block has the same error? It just sounds so little likely but I'm checking everything so I why not


The dryer tip they mention in that thread I've sorted out. I didn't see any desiccant in the block when I refurbished the old one but just to rule it out I replaced the whole dryer unit with a new one since I read that many had problems related to this.


I drove 5 hours today and it was fine. Logged approx 2hours of them but I didn't bring the computer that has nanocom log viewer so I can't check the logs yet.

I don't have control over is the sensor signals when driving. They measure good when stationary and when I look in nanocom both stationary and while driving. Nanocom does have a slow update rate to the log data so it could drop some "frames" I guess and I wouldn't see if it oscillated. This is a voltage feedback so maybe It would be worth connecting to a sensor with an oscilloscope when driving just to see if they all are stable and comparable. I have a 4 channel scope so I can log all at the same time if that can tell something.


Would be interesting to see a 10-20 minute run on a highway if someone get time to do that. The log I got from my friend had the same symptoms as mine so this might be a feature It is just to put in a SD card in the nanocom and click on the floppy disc icon once in the EAS ecu "input section" and type a file name and click save. The file won't work directly but I can give instructions on how to modify it or modify it for you to get the correct format so nanocom software can open it.

 

[pwood999]

Reading through the thread again, your EAS system seems to be in generally good order, i.e.

- Stays up when parked.
- Pump fills tank in 8-10mins.
- Works while driving.

Maybe just ignore the 5mm adjustments when stationary as "normal Land Rover" because they used a flawed ECU algorithm ???

 

[minim]

while driving it's constantly adjusting 20-30mm. I will go out and check the valve block now and see if the factory fitted NRV valve is wrong.

 

[minim]

Good thing I brought some tools on this trip

Pulled the EAS valve block out of the car and inspected the NRV valves today. There was heavy use of loctite on the block but nothing of it inside the valves and block. It basically looked good so I put it together and tested some again.

Did the NRV valve test where I let the compressor go until the tank was full and inspected if it still bumped up and down. It did.. So after not very long the compressor started running again (only stopping 7-10seconds so exhaust valve could open). Target heights was within 3-4 bits. I did notice that my front left was 4-5bit lower than front right even if I calibrated this the other day. Tried setting the target height 5bit lower for this and then it was going around the target just as before so it didn't help.

When I look at the height values while it's doing this jumping it's just like in the logs. It overshoots each time it adjust both up and down. Looks like the valves are open to long both when lowering and going up. This can't be how it's supposed to work.

I didn't bother doing at test drive today as it's prolly behaving just in the same manner when driving since it did this when stationary before also.

Edit; the NRV that's in the little block on the last picture is in wrong orientation in the picture. I didn't fit it together like that.

 

[minim]

Used a spare ECU and soldered onto the 5V output pin to the sensor, 5V feedback pin and gnd of each sensor to log some samples tomorrow. Is there any other input/output I should check? I do notice that I have 555rpm at idle according to EAS ECU but in the Bosch EDC ecu it reads 750rpm like it should be. I measured this rpm signal before but it's not easy to know if it's ok or not as it is a "pwm" (not true pwm) signal from BECM so without knowing the spec it's hard to tell if it's off or not.

 

[minim]

This is a log session of all the four corners. The last part of the video (from 2min out) is logging the 5V output from ECU and it also shows the same noise as I get on the feedback when EAS is adjusting. I don't have any referense but that is a whole lot of noise! When I move from high to standard I measured 500mV differense and the noise is at times way beyond that. I wonder if thoise noise pulses are making the EAS adjust. Could it be a ruined noise filter on the 5V power supply of the ECU? Will look into the circuit tonight and see if I can add a RC filter to try to cancel out some of it. Weird that both ECU I have should have the same issue but could be more common that we know since most don't do log runs and notice small dipping behaviour. As long as it doesn't alarm it's good right



Channel 1 yellow - Pin 5 - Left front height sensor
Channel 2 cyan - Pin 23 - Right front height sensor
Channel 3 Purble - Pin 22 - Right Rear Height sensor
Channel 4 Blue - Pin 4 - Left Rear Height sensor