heinzboehmer's 2002 Topaz 6MT Coupe

[heinzboehmer]

Datalogger is picking up readings from both sensors just like before, so I'm calling this a success.

Success indeed! Test drive went great. New MAP location + non-shitty wiring seems to have cleaned up any remaining low RPM weirdness, awesome.

Extremely happy with how the car is running now, especially with all the recent cold start updates to the mullet tune!

[Obioban]

Hmm, not sure I follow.

The screws go through the adapter and engage the threads on the air rail. Don't think heatserts would have helped here.

Oh right, herpity derpity. Ignore me.

[heinzboehmer]

You can use heatserts with PET-CF

Hmm, not sure I follow.

The screws go through the adapter and engage the threads on the air rail. Don't think heatserts would have helped here.

[Obioban]

You can use heatserts with PET-CF

[heinzboehmer]

Also, quick side project to give myself a break from all the CAD work.

I've never been fully satisfied with the Kassel MAP sensor solution. It being so close to the throttle bodies means that the sensor sees huge fluctuations in readings with each revolution of the engine. This doesn't seem to affect drivability much, but it's never sat well with me. Additionally, the harness going to the sensor has to make some very sharp bends to make sure it's out of the way of the throttle linkage. Again, not something that affects how the car drives, but I didn't like it.

A week or so back, I noticed the CSL air rail was ~$400 after tax on ECS (and in stock in the US!), but closer to $800 through european sellers. NLA fear creeped in sooo I jumped on it:




My original plan was to just kinda re-zip tie the harness I had made for the Kassel sensor and call it good. However, I found sketchy patchwork in the middle of the IAT harness that I had forgotten about and this ridiculous choice by Kassel for the MAP sensor connector (on the left):



For some reason, they used the wrong terminals, put no waterproofing grommets in there and picked a questionable wire gauge for one of the three wires. The connector housing is also clearly a knockoff with pretty bad tolerances. I would not be surprised at all if this is the cause for my intermittent MAP sensor code (which, admittedly, never seemed to affect how the car ran).

The stuff I laid out to the right are the closest replacements I could find in the stash. I rewired that part of the harness with those, but I've ordered the correct terminals and grommets and will replace soon.

I was already halfway into building a completely new harness, so I decided to just redo the entire thing. This is how I routed everything:








As you can see, I ran my sub-harness directly into the bottom of the stock engine harness box thing by the brake booster. Not having that extra conduit snaking down the front of the harness box cleans up the engine bay nicely:



I did briefly consider running the wires through the stock conduit that runs under the airbox and then making a slot (with a matching 3D printed grommet) in the hard plastic distribution block bolted to the engine for the IAT + MAP sub-harnesses, but ultimately decided that was way too far down the rabbit hole. I'll revisit and do it that way if I ever run out of projects

I, unfortunately, did not have the appropriate terminals for the IAT, so I left some slack in the harness and ordered some of those as well. Slack is tucked into the conduit for now, but I'll be swapping things when I get the parts. Getting the boots over the conduit is gonna be a fun adventure:



(yes, I also ordered a new boot for the IAT)

Anyway, next up I printed karter16's Bosch MAP adapter out of PET-CF, for it's temp resistance. This filament has a heat deflection temperature of 205 C, so should work well for the application. Also grabbed the complementary hardware:



Just like karter16 noted, the metal sleeve on the Bosch MAP just barely doesn't line up with the air rail. Dremel + a carbide bit made quick work of this though:



Gave the sensor and adapter a quick scuff, then bonded them together with some CA glue. This part isn't really necessary, but I did it anyway cause why not.

Here's the final install:



(yes, I know that's a 90 deg boot, straight one is in the mail)

The screws did start digging into the plastic on install, so they got torqued to "as much as I can turn just the bit with my fingers". I don't think this is gonna be an issue.

Datalogger is picking up readings from both sensors just like before, so I'm calling this a success. Huge thanks to karter16 for sharing the adapter and all the details surrounding it! Made my life so much easier.

[heinzboehmer]

This is frickin cool

Thanks! More to come soon.

Small update for now. I think I've found a suitable replacement for the tubing that BMW uses on the brake booster shut off valve: Festo 570360 - PAN-MF-10X1,25-SW Plastic Tubing

Conforms to DIN 73378 just like the OE stuff, measures the same and looks the same. Only issue I'm running into is that basically no place wants to sell me the tubing in lengths smaller than 50m. I only need like 50cm of the stuff, so it would be a huge waste of material and money to buy that much. Placed an order with one place whose website allowed me to order 5m, but we'll have to see if they actually fulfill my order.

The flexible vacuum hose I have on there now works fine, but its OD is much larger than the OE stuff and it rests on the E86 braces, as seen here:

And this one gives a good sense of how tight the clearances are:

The Festo stuff should be a much better fit for this application, as I can bend it out of the way and make sure it doesn't contact the braces.

[siamkazi]

This is frickin cool

[0-60motorsports]

Today I started on the painful process of recreating the stock plastics in CAD:



This is gonna take a while...

Great work brother!

[heinzboehmer]

Today I started on the painful process of recreating the stock plastics in CAD:



This is gonna take a while...

[heinzboehmer]

Speaking of packaging, I scanned the engine bay with the braces installed and then individually scanned the firewall plug, the cabin air filter housing and the cable tray. Here's how they fit up in CAD:




Interference with the firewall plug is minimal, which is nice. Very little work needed there:



As expected, there's tons of interference with the cabin air filter housing. This one will need significant modifications:




Fortunately, it looks like I can keep the stock cable tray! I was not looking forward to redesigning that part:



Of course, not all good news. Braces interfere significantly with the air filter:



Looks like if I raise the rear by 30mm, it'll clear:



Mocked up what that would look like with some spacers and I might be able to pull it off:




The rear edge doesn't really seal with it propped up and the air flow over it looks like it would be greatly reduced



I don't know if spacing up is the right solution. I need to give this more thought. Might need a more substantial redesign of this interface on the cabin air filter housing.

[heinzboehmer]

To do something like this, you’d want to modify the auxiliary fan support to tie the two frame rail attachments into the upper core support, then triangulate that to the strut towers. Then you’d need to essentially build a mini front subframe that could tie back to the frame rails in front, or maybe pick up on the sway bar mounts as that’s what’s doing the frame rail twisting. Tie that mini front subframe into both the aux fan support and the existing subframe. A lot of work.

BMW does something similar to what you describe with their modern cars. Here's a G30 for reference, since that's the simplest implementation I came across during my research:



They skip the "mini subframe" thing you're talking about and just do the frame rails to the radiator support part. Probably because the actual subframe is extremely beefy and spans a good chunk of the front end. Other cars like the G80 have a separate v brace going from the frame rails down to the subframe.

But anyway, the G30 braces go from the top of the frame rails to the top of the radiator support and then another set of braces go to the strut towers.

Forward facing strut tower pickup points on the E46 would be fairly easy to do by bonding in the E85/6 brackets (or similar):



But yeah, sounds like packaging would be pretty hard. I think there might be room to squeeze the diagonal braces in between the aux fan and the kidneys, but can't say for certain. Then you'd need to clearance a good chunk of the front shrouds to be able to snake the strut tower braces through.

[Bry5on]

Man, good question. You can see how well supported the frame rails are with the stamped and tubular braces, then that it is boxed in to the strut towers and the lower subframe and frame rails. This would be pretty effective in torsion but I’d wager less so than the things we’ve already done. Also way harder to package in the e46 engine bay I think.

To do something like this, you’d want to modify the auxiliary fan support to tie the two frame rail attachments into the upper core support, then triangulate that to the strut towers. Then you’d need to essentially build a mini front subframe that could tie back to the frame rails in front, or maybe pick up on the sway bar mounts as that’s what’s doing the frame rail twisting. Tie that mini front subframe into both the aux fan support and the existing subframe. A lot of work.

[George Hill]

As you can see the F8X has a much beefier front crossmember/radiator support that the E46 which is long spans of flimsy thin gauge steel. I wonder how much of a benefit there is in improving that.

Bry5on

[bigjae46]

It seems like it's one of those big generational jumps. E9x is way stiffer than E46 and every car of the E9x era has some version of these triangulation braces. E85/6 is a weird E46/E9x hybrid and it lucked out with the E9x chassis rigidity.

Think we're just slightly too old and missed out on that.

Also, just for fun, this is the bracing in the F80 M3. Again, way stiffer than the previous gen, but man, it essentially needed a full exoskeleton to get there:

As you can see the F8X has a much beefier front crossmember/radiator support that the E46 which is long spans of flimsy thin gauge steel. I wonder how much of a benefit there is in improving that.

[heinzboehmer]

I am curious why it didn't come from the factory. As in, the Z4M did, and wasn't that much later.

It seems like it's one of those big generational jumps. E9x is way stiffer than E46 and every car of the E9x era has some version of these triangulation braces. E85/6 is a weird E46/E9x hybrid and it lucked out with the E9x chassis rigidity.

Think we're just slightly too old and missed out on that.

Also, just for fun, this is the bracing in the F80 M3. Again, way stiffer than the previous gen, but man, it essentially needed a full exoskeleton to get there: