Repair-It

Looks like my right front level sensor is flaky. Or it is while the car is in the air. I'm sure I damaged it when the linkage came around backwards when we were doing the struts. Shows level as like -4" when the other side is +3. If I touch the level sensor a bit it snaps back to normal. But it feels loose and probably broken inside. Damn.

A2129050403

https://www.mercedesbenzpartsshop.com/o ... 2129050403

I don't know if this just controls headlight level or if it is also used for damping control. There is no front Airmatic so it wouldn't be for front level control. Pricey mistake. I could also take a look at it with the car on the ground. Maybe it is in a happy zone at that point in its travel. But, I should probably just replace it and be more careful next time...

Ended up centering the steering wheel after 2 tries. The wheel was about 8 degrees off to the right when driving straight and it turns out the tie rod adjusters move the steering wheel about 8 degrees per full turn. So handy.

Then I was able to teach in the steering stops. I had a lot of trouble until I realized it really wants you to manhandle the steering wheel past lock so it can, uh, learn motor torque or perhaps be confident where the stop limit is. During normal operation maybe it just stops boosting at the stop rather than try to work against itself like hydraulic steering. It still clunks the way the other one did with the engine off but perhaps less so? When not running, there is play. When running, there is no play. When driving, there is play. Get all that? I measured about 3 degrees of steering wheel play before anything happens on center when driving. I think it's better than the old rack. Maybe the old rack was 5 degrees? I should have measured. Like with my CTS-V rack replacement for the same reason, the on center play didn't really get fixed but it felt more positive on turn-in. That applies here, as well. Was this worth $1800? I dunno. Maybe not? But the curiosity would have killed me. Would the $2700 reman at the dealer been any better? I don't really know. Would I have rather spent even more to get a better result? Not sure. If I could convince them to order one I could grab the input shaft and rock it and that would give me some indication of play. But I'm probably done with this experiment. Now I have a new rack (and motor and belt and tie-rods).

I did put my old rack in the vice and tried playing with the preload adjustment thingy just to see. It didn't really do anything. I loosened it and tightened it and the apparent play at the input shaft didn't really get worse or better. So I guess I'm actually relieved because if it magically made it better I would have felt some regret in not just trying that, considering how easy the rack was to remove.

Turns out the steering wheel was still off a small amount. Center was slightly to the left. So I decided to check via scan tool since the car constantly tries to find center to provide a compensation offset for the steering angle. Rather than lug out the Xentry laptop, I decided to try the Launch x431. Well, that did not correctly display steering angle. It was just 0 and didn't change. Other values in the steering module did work so that's just a fail I guess. After that I decided to try the Top Don TopScan. I had to download the Mercedes software (to the tune of 700MB) so that took some time. Then I was able to get into the steering module and it did correctly see those values.
Sure enough, -1.4 degrees. Now I tried to be a hotshot and use my math from before. It would have worked, too, but I remembered it wrong. I thought it was 4 degrees per full turn. As a result I calculated about 2 flats on the inner tie rod to dial out the 1.4 degrees. Problem is, it was actually 8 degrees per full turn if I had just re-read this thread. So after doing 2 flats, it was off about the same amount the other direction. I went back and backed my adjustment off to 1 flat and that was about right. I had the right idea. Actually, FWIW, my intuition was 1 flat but I let the false math guide me. Whoops.

Anyway, here's the proof:

Turns out I can't live with the play in the new rack. Contacted Rockauto requesting a refund. They said due to this not being easy to source they will do a warranty refund. Works for me. Now I need to decide what I want to buy instead. Leaning toward online rack (with -64 suffix) from local dealer. There are some cheaper options on eBay (from Poland) but I just don't know.

https://www.ebay.com/sch/i.html?_nkw=A2 ... m570.l1313

If a rack from the dealer doesn't fix it, I think I will just live with it as I've done everything I can at that point. FWIW I also can't seem to source a rack in the US without that -64 suffix. -80 means reman. One of my struts was a -64 and I think the parts guy just said it's sort of an alternate supplier or that kind of thing. Hopefully not from the same line I got my Bosch rack from. It definitely would suck to pay even more and get the same thing. But it WOULD help with my FOMO.

The forum says the clunk with the engine off is normal though I find that odd. Logically that kind of metal on metal sound perfectly coincides with the play while driving that I feel. Here are two videos I just made:

https://www.youtube.com/watch?v=-wBFOWnTEy0
https://www.youtube.com/watch?v=O1kj0piHv-Y

I just want to enjoy the car and I've done many other improvements that have made a difference. Kind of given it a new life at 111k. I just need this on center play to be significantly reduced or eliminated.

Talked to Mike at Mercedes Warwick parts. He's the internet parts specialist I guess. Told him my story and asked if I should do the new part with the -64 suffix or the reman part with -80 suffix. He said do the reman. They have had more issues with new parts vs reman, for whatever that's worth. He seemed to at least agree with my instinct that a reman should probably have better quality control since it is rebuilt by hand vs an assembly line. Mercedes could certainly reject a new part that doesn't meet spec and decide to box it as Bosch but who knows.

So my mind was set on the reman. There was only one issue: the parts website only let me order he -64 suffix new part that I don't want. While on the phone he was able to change the listing to the -80 suffix reman part! Additionally, he put in a request from their NJ distribution hub to get a rack to Warwick, before I even bought it. So that was wonderful. Then I hung up and ordered the rack. Another $3k BUT, when all is said and done I think the delta cost to upgrade to this from the Rockauto Bosch will be $885 (plus another alignment and plus S&H back to Rockauto, I suppose). It will be totally worth it if it solves my issue.

He said I could fiddle with it before I walk off with it, as well, which gives me some hope I guess. I'm just worried the input shaft will feel like the others (which could be normal, despite the play). We shall see. Should arrive Friday.

Based on how horrible the parts counter experience with New London was on the cat, I think I'll be headed to Warwick for anything else I need. I almost feel like I should've just eaten the $150 more that Warwick wanted for it.

It is occurring to me that at least some of my positive experience with Mercedes over the past 5 years is due to Warwick Mercedes being pretty good to work with. And maybe that's not universal.

Mileage at replacement (removal of Bosch/installation of reman) Picking up the reman rack at the dealer. But then I got distracted. Recall from my alignment that my right front camber was actually too positive by around 0.6 degrees (which was a change from too negative before I did the front struts). I hatched this plan to use my inclinometer, loosen the knuckle bolts and adjust the camber accordingly. I had a brain fart initially and put the inclinometer on the strut but I quickly realized I needed to put it on the brake rotor hub. It was a little tricky to tighten the bolts while holding the adjustment but eventually I got it. What's also interesting is that I did check the left front and it measured 89.9 degrees (almost perfectly perpendicular to the ground). And adding 0.6 degrees negative camber to the right front also put it at 89.9. So maybe the car was engineered (or this is a coincidence) to make 90 degrees dial in to the perfect nominal camber when the car is on the ground (something like -1.9 degrees but I would need to double check). 90 degrees would be 0 camber at full droop if that makes sense.

Left side for comparison Right side before adjustment. Right side after initial adjustment. Right side after torque/angle of knuckle bolts Before I removed the Bosch rack I did some experiments. I had Ian wiggle the steering wheel to see if I could further isolate the play. I put vice grips on the steering shaft and while Ian wiggled I tried to see where the play was. I had him decrease the wiggle amplitude to see if his wiggling continued while motion stopped at the input shaft. As far as rotation goes, that never really happened meaning there was NO rotation outside of the steering rack. Furthermore, I was able to notice that any remaining play was happening in the input shaft rocking radially. What I am saying is, all the slop was in the Bosch rack and it was time to continue.
https://youtu.be/icKRJrHo9Q0?si=5uvHaxfqE8fsU2Le

Removed the Bosch rack. With all three on the floor I compared various things and found some interesting stuff.

With a vice grip on the input shaft and twisting the shaft back and forth (like wiggling the steering wheel), the clunk between the Bosch and original rack seemed similar. Another test was just grabbing the input shaft with my hand and rocking it radially to see and hear the differences between the racks. What I found is that the Bosch and original rack were similar and the reman rack I just bought was better. Kind of a relief.
https://youtu.be/i26fsWF9jyI?si=BDJ7kiLKg9E8U-hZ

After further studying I realized the Bosch rack actually had a Mercedes sticker but had the Mercedes-Benz logo itself scratched off. Both the Bosch rack and reman looked pretty much identical, made in Hungary. But the reman showed Mercedes-Benz on the sticker and the Bosch rack had an additional Bosch part number sticker that the reman rack didn't have.

Original rack Reman rack from dealer Bosch rack from Rockauto. You can very clearly see the Mercedes-Benz logo painted over (maybe whiteout, lol). This leads me to assume that the reman rack did come off the same Bosch Hungary manufacturing line at some point. But then what? Remanufactured by Mercedes? Remanufactured by Bosch to better QC standards? And what decision happens to scratch off the Mercedes logo on a particular part? Because it failed QC (at least Mercedes standards)? Was this reman rack really used in a car before being remanufactured? I have questions but this is interesting stuff.
https://youtu.be/epDV6Ji_llc?si=Bc77wGyJK3_zOcbV

Other labels from the reman box. Got the new rack in and based on my measurements with the old one (tape measure to outer tread sipe), I didn't need to change the toe. So I didn't. Not too hard to swap these at least.

Had the BenzNinja flash software into the rack module and do an initial setup. Then I test drove it. The steering wheel was off a solid 10 degrees to the right but I was able to complete the teach in (can only do partial on the lift). After that I decided to center the wheel so I could feel the steering properly. Electric steering racks do funny things if they aren't centered or haven't learned the center offset yet. Ended up doing 1 full turn plus one flat of both tie rods and that got it pretty close.

So what's the verdict? It looks like the center slop is down to 2 degrees of steering input and it does feel more precise on center. Is it perfectly tight? Nope. But maybe that isn't possible and I do feel good that I did what I could. I am happy and I will return my core, and ship my Bosch back. And get an alignment.

Alignment scheduled for tomorrow morning. I just gave my Bosch rack to FedEx and will return my core tomorrow.

kevm14 wrote: Sat Sep 06, 2025 5:37 pm But then I got distracted. Recall from my alignment that my right front camber was actually too positive by around 0.6 degrees (which was a change from too negative before I did the front struts). I hatched this plan to use my inclinometer, loosen the knuckle bolts and adjust the camber accordingly. I had a brain fart initially and put the inclinometer on the strut but I quickly realized I needed to put it on the brake rotor hub. It was a little tricky to tighten the bolts while holding the adjustment but eventually I got it. What's also interesting is that I did check the left front and it measured 89.9 degrees (almost perfectly perpendicular to the ground). And adding 0.6 degrees negative camber to the right front also put it at 89.9. So maybe the car was engineered (or this is a coincidence) to make 90 degrees dial in to the perfect nominal camber when the car is on the ground (something like -1.9 degrees but I would need to double check). 90 degrees would be 0 camber at full droop if that makes sense.

I remembered wrong. I actually found the photo of my alignment and it was BOTH sides that were below the min spec for camber. So the left still needs to be adjusted. That would also explain why I felt a bit of a pull to the left. Now I'm in a pickle because I have an alignment appointment this morning. I think my plan is just explain the situation and see if they'll get a reading from the rack first. Then come get me. And either convince them to adjust the knuckle (I doubt it, or MAYBE if I pay more labor), or I take the car home and do my inclinometer trick on the left. SIGH.