Repair-It

So it's pulling to the right a little bit. Sometimes more than a little bit. The strut job does not impact alignment. Loosening and tightening the upper ball joint also does not impact alignment as the joint sits in a machined taper - it can't seat multiple different ways. And again the strut is not a structural member as far as the location of the wheel (it does bear the weight of course but only vertically).

I decided to look into how level the car actually is or isn't. What I recall was when driving around with the leaky front right air strut, I had raised the whole car but only the left side really raised properly. The right couldn't make it above ~normal ride height. So the car had a slant. When driving like this I noted that there was a very strong pull to the right. Remembering this, I took a look at all the level info in DAS as a starting point. All 4 air struts are independently controlled for air pressure but height info works like a tripod: there are two front level sensors but only 1 rear level sensor. I don't know how it decides to apportion air between the two rear struts (GM rear load leveling was actually just one feed plumbed to two air shocks so the pressure was the same in both - not so on the S550).

I was definitely able to manually create a ride level that got the car to steer straight so I figured maybe I was onto something. That said, even with the car self leveling to an indicated "+9mm" ride height on all three sensors, it still had the gentle pull to the right. I pulled it into the garage and started measuring wheel well height. This is where things got interesting.

What I noticed was, with the car indicating level, the front right was lower than the front left by about 0.25". However, the right REAR was 0.5" lower than the left rear! Remember that the car doesn't know rear height left to right. So this tells me that if the rear is out more, then it's probably causing the issue and sort of pulling down the right front (or pushing up the left front if you think about it).

Despite going through all of the diagnostics I didn't see any leaks or anything amiss. That said, when the car settles in the driveway, it also appears to me that the right rear is lower than the rest. My theory is the right rear is starting to go. If I preemptively replace the right rear, this problem might go away (and avoid a future, surprise failure like the front right).

I should see what WIS has to say about doing the right rear air strut. Hopefully it's as easy as the front. It's not a great time of year to lay the car up for another 2 weeks though, since the temps are plunging below freezing and salt will appear on the roads any day. That means no E63 and truck only. I guess I could drive the truck for 2 weeks.

Just wanted to continue my discussion about the rear air struts to another thread so I don't confuse myself in the future.

So the good news is, WIS makes it seem more involved than it is when comparing to this Suncore video:
https://www.youtube.com/watch?v=ZPRFvsGxup8

Side note, I left a comment here 2 years ago for some reason (and it's the only comment).

The bad news is, this is still way more involved than the front air struts.

WIS attached for reference.

And here's the Arnott video: https://www.youtube.com/watch?v=dI0Ecjy8-3c

They have a slightly different method. I don't know why they remove the brake caliper though. But they drop the lower control arm from the knuckle rather than at the subframe.

Also, the video claims that Arnott offers a brand new unit rather than a rebuilt air spring on a used shock.

I don't think the air struts are the cause of the pull. I think the level calibration got out of adjustment and I guess it happened when I installed the rebuilt front right air strut. I have no explanation for how that occurred though.

So the symptom was a right pull and the ride height measurements showed that the front right was about 0.25" lower than the front left. But the rear right was pushing 0.5" or even 0.75" lower than the left rear. Given how the car behaved when I was diagnosing the failed front right strut, I believe this car just pulls to the side with the lower ride height. And it's pretty sensitive, it turns out. Further compounding the situation, my Michelin Pilot AS4s are pretty worn, around 3/32" remaining, and have gotten more trammy which may well exaggerate a pull condition like this.

I still maintain that replacement of the strut cannot impact ride height calibration. The struts do NOT contain any ride height sensing internally. That is all done by the two front and one rear level sensor that is connected to control arms. The rear struts DO appear to somehow track actual ride height (particularly evident when running diagnostic tests) but I have concluded that this is a calculated value as there is no possible way to directly measure rear left/right difference.

All that said I tried to run the DAS procedure to calibrate ride level, see if I could level the car out, and see if THAT fixed (or helped) with the right pull.

I don't fully understand the AIRmatic logic but I do deem it fancy. Essentially DAS lets you individually set ride height at all 4 corners. I went for 28" wheel well to ground (through wheel center cap) at all 4 corners, which seemed about right.

That screen looks like this and they want you to do it with the engine off which requires a charger. I had a 20A and that was still barely enough as I got low voltage alarms from DAS while the compressor ran. This took a long time to dial in and eventually the left rear stopped responding to lowering commands. I ran the diagnostic test on strut level control and the left rear AND right front failed the test. I guess I ended up going on a test ride to see if I really broke something or the car was otherwise now unable to level itself (potentially worse than before). However, it seemed to do about the same. I pulled back into the garage and re-ran the strut level diagnostic test and guess what, now it was fine. At first I thought maybe there was a contaminant in the system but then I thought maybe the level control valve body thing that has the air valves could be starting to fail?? I have no clue. What I can say is it is good I didn't just try to reman a rear air strut because it would have done the exact same thing.

Anyway now that the left rear was responding I tried to complete the level calibration. Once you set the levels, the next screen asks for inclinometer readouts. On my car the front is taken at the spring control arms and the rear is taken at the rear axle shafts. Of course I don't have this "Romess" fancy electronic tool but see the attached PDF for what they are asking the tech to do. I think they use the angles to determine absolute ride height and I guess the car wants to know what absolute wheel position is, for proper system control and response. In theory it may also stop a tech from (as easily) raising or lowering a car beyond what MB wanted (to ensure ride, safe handling and not damage anything).

Since I didn't have this device I just gave it angles that were in the middle of the range as that is what some videos suggest. I did it and it took the calibration. Went on a road test and.....it was about the same as before. Super.

So I pulled it back in the garage, went back for 4 corner 28" wheel well height, and slightly tweaked the angle input. Essentially I gave the left side 0.1 degree less inclination. It also was successful (I tried a larger deviation and it wouldn't accept it, FWIW). This was the second successful calibration attempt. Did a road test and finally the pulling is significantly better. And wheel well height measurements are much closer now. It may still be a touch low on the pass side but it's now within like 0.25" at all 4 corners. Again it's more sensitive than I'd expect but I do think the worn tires are exaggerating the effect.

I am excited to try a commute so I can evaluate on a wider range of roads but this is very promising. And again I am glad I didn't just throw parts at it. I'll have to keep my eye on the level control valve issue and maybe price out a valve block just for fun, assuming that would be the correct fix.

kevm14 wrote: ↑Sun Dec 08, 2024 2:44 pm I'll have to keep my eye on the level control valve issue and maybe price out a valve block just for fun, assuming that would be the correct fix.

Oof, pricey. https://www.mercedesbenzpartsshop.com/o ... 2123200358

If it means anything this is NOT on the shelf at MB Warwick. I think these are more robust than the previous gen. Actually I think it really does mean something because the application list for this is insane. I do think the P/N changed at least once though so it was probably revised (form/fit/function), which is fairly common.

There are probably some alternative options but not sure how I feel about aftermarket (or used). I wouldn't go this way until it was more definitively acting up (and not just during level calibration as mine was). I have seen zero AIRmatic events or codes in the system or warnings on the dash so nowhere near ready to throw this at it, but saving it for later nonetheless.

EDIT: this is a very common part. Rockauto shows 3 listings. Two Arnott and one VNE.

Arnott 1: https://www.rockauto.com/en/moreinfo.ph ... 51&jsn=323
Arnott 2: https://www.rockauto.com/en/moreinfo.ph ... 51&jsn=324
VNE: https://www.rockauto.com/en/moreinfo.ph ... 51&jsn=325

Seems that both Arnotts are compatible (by P/N at least). At half price I guess I couldn't totally write these off...

Or you could go full Chinesium for a hilariously cheap $55: https://www.ebay.com/itm/145817713228?_ ... BMyLjA__Rk

I'll pass on that but it's amusing.

Looks like I could get a used genuine one for around that same price which I might choose between the two, especially since the donor vehicle can be as new as 2019 (X166 GLE/GLS, yup).

Yeah this is fixed, which is good, because the right pull was really starting to annoy me. I also tweaked the headlight aim per WIS. Just had to lower the right one a tad.

However, aesthetically, the front may be a bit too high now. Maybe 28" front and rear was not the best plan. Maybe I wanted like 27.5" front and 28" rear. Well I'll live with it for a while and see if I still get itchy and want to try again.

Front seems too high. Probably need to try again dropping 0.5"? I wish I could do the inclinometer thing so it really knows what the absolute ride height is. I think it's important for suspension control.

EDIT: Maybe I can, on the cheap.
https://www.harborfreight.com/digital-a ... 63615.html

What I'd still like to know is exactly what the AIRmatic module does with this information. I assume that it uses the actual inclination of the measuring points to tell the computer what absolute ride height you are at. But the first step seems to be to set the ride height so if they just wanted to calibrate a new level sensor voltage to absolute ride height, you could just take an inclination measurement and the computer would know enough. Since the first screen lets you dial in all 4 corners, this tells me that part of the plan is to set the actual ride height and THEN tell it where in the total wheel travel your chosen ride height actually is.

Bought this instead: Klein Tools 935DAG Digital Electronic Level and Angle Gauge, Measures 0 - 90 and 0 - 180 Degree Ranges, Measures and Sets Angles https://a.co/d/7g8hEqk

A real brand, better reviews and cheaper than HF. Imagine that.

I was visiting Old Lyme so I decided to drive on the lift and measure angles.

This really needs to be done on an alignment rack that's very level side to side but I just was experimenting and didn't run the leveling calibration. I probably have enough info so if I do the actual calibration again, I'll feel confident my angles are in the right zone. They probably were already based on my results.

I also think I may need to compensate for the lift as the angles on the left side of the car were lower than the right at the front though the rear seemed tilted the opposite way. I did also do wheel well height measurements to correlate to the angles (with the engine running) so I may be able to do some kind of intricate math/estimating to show that the lift was a bit off left to right.

These just show what the measurements look like. I will post the hard numbers after. Left front: 1.9° 28-1/8"
Right front: 3.1° 28-1/16"

Left rear: -2.1° 27-7/8"
Right rear: -2.5° 28-1/8"

For the front, higher angles mean higher ride height.

For the rear, higher angles also mean higher ride height but it's negative so it will get closer to zero numerically before it crosses over and goes positive.

So remember that nominal values for the front are around that 3.0 degree range and for the rear around -2.0 degrees.

The left rear was maybe 0.25" lower in ride height than the right, so the angle should be less, but it's not. The left front was slightly higher in ride height than the right front but the angle is less. So this suggests the lift was actually tilted down on the right side if I did my visualization correctly.

The fronts were within 1/16" so maybe for sake of the argument pretend that's the same. It would take subtracting 0.6 degrees from the right and adding 0.6 degrees from the left to get them to match. Of course the left was a tad higher so we can consider that at the end.

For the rear, if we add 0.6 degrees to the left, that would put it at -1.5 and the right if we subtract 0.6 that would put it at -3.1. But the ride height is different. It is lower on the left which would result in a numerically lower angle. This actually suggests that at the back of the lift, it is low on the left side. Sigh. If they were the same, it would take 0.2 degrees added to the right for -2.3 and 0.2 degrees subtracted from the left for -2.3. However, the right measured a bit taller so that should be greater than -2.3. Perhaps -2.1? And the left measured lower so that should be -2.5? I don't know if DAS would accept such variation.

If we just try to average stuff, that puts the rear around -2.3 and the front around 2.5. I used around 2.9-3.0 for the front and -1.9 to -2.0 for the rear so a bit different. If we believe DAS, the acceptable range for the front would be 2.5 to 3.6 and the rear -1.6 to -2.3. This would put me on the low limit for the front and low limit for the rear. Does that mean my ride height is on the low end of spec? I thought the front was a bit high if anything. The rear seems similar to what it was before I touched any of this.

What I should have done is raised the suspension and re-measured everything. Then I could have a conversion between angle and ride height.

I don't know exactly what I learned here. Perhaps that my initial entry of using angles in the middle of the range was probably fine. I don't know what a few tenths of a degree really do to the suspension calibration (maybe nothing). And using a sloppy old lift to do this is probably not better than just guessing the angles.

Checked the garage level. It is sloped to the right by about 0.6 degrees at the rear and about 1 degree in the front, using the angle finder on my 4 foot level. Pulled the car back in and checked angles. Turns out I was easily able to attach the angle finder with the car on the ground to the front lower control arms and the rear axles. I tried to compensate for the garage floor slope and the angles mostly made sense except for the front left. This is what I found on the lift so maybe the geometry of my front left is off somehow.

I reset the height to the same 28" rear and 27-5/8" front. For angles it seemed like that was slightly below the lowest limit so does this suggest that what I think is a proper ride height is actually a bit low? Either way, I left the heights and entered the lowest possible angles (at least according to DAS). I guess I could have entered the exact angles and let it tell me it couldn't accept that.

This shows that I dropped it about 15 mm in the front which makes sense.
Everything seemed fine but it did what it did the first time I tried this process: it just reverted to what it was before. So sneaky.

So I did the whole thing again but used the same angles that I used when I had luck last time, and this worked. The front ended up around 27.5" and the rear stayed around 28" and it seems more even front to rear. It still drives nice so I'll continue to drive it and observe to ensure I am still happy with this. I will have to tweak my headlight aim again.