Black & Tan 332iT

[Bry5on]

Ok, we left off Friday with a bunch of the rear end work done, but we hadn't started on anything forward of the rear seats bench. Yesterday was another productive day and I was able to get PDC fully functioning in the car, button and all. We also got on with a bonus of hardwiring in an AEM CAN-based Wideband O2 sensor, for automatic datalogging with Gauge.s whenever I plug the device into the OBD2 port. No more laptop with Testo and slow datalogging speeds (CAN is roughly 10x the data rate). On to some paraphrased play by play:

Starting with wires routed to their destinations and some coffee, we get to work:


The e46 rear PC harness I bought looks like it was specific for the sedan/coupe and not the touring. I hadn't thought this one through ahead of time and was surprised in the moment that I was about 3' short of running the wires in the factory location behind the rear woofer. Fortunately, I'd grabbed some spare harnesses from an e53 X5 and I took a page out of the X5's book. The X5 has an intermediate harness between the rear PDC sensor harness and the PDC module harness. It uses the same pins as the e46 harness, so I depinned the e46 connector and swapped pins into the mating X5 connector to create an effective extension harness identical to the X5 setup without cutting any wires. Pleased with this workaround.


I'd also been building the harness and routing wires for the trailer module. I started with the X5 trunk harness and ran all the input wires to the bundle right above the battery. I also dropped the two rear fog out wires (run to the LCM pins 20 & 49) and the X5's 12 pin trailer-out plug into the same location, along with 4-pin trailer connector loose wiring, making all future junctions accessible in the same location (tail light harness is right there), as I'm sure there will be debugging to get this all working and I don't want to be tearing into a bunch of interior panels again.


I'll spare you all photos of the wire routing along the sill and up into the fuse/junction box area, but suffice to say it was all routed along factory routing paths, at the usual junctions, grounds, positives, etc. While I was there I added this AEM in-line CAN Wideband O2 and terminated that along the factory harness to a ground block and unused cabrio Fuse 13. It tucks nicely below the general module and the O2 sensor wires pass through the firewall grommet along with the front PDC harness for a clean install.



After the passenger side wiring was complete, I passed a whole bunch of wires behind the AC panel to the driver's side and terminated them in their respective locations, along the factory harnessing:
- Gong power
- Gong PDC signal
- PDC button and light signals
- Rear foglight wires (2x)
- LCM -- AHM comms wire
- AEM CAN comms wires

The result is a fully functional PDC system, front+rear with button and reverse-trigger, a functional gong (I'd already run the wire to the cluster last time I was in there), and AFR now showing on the gauge.s dataloger at 100Hz. I have Gauge.s setup to datalog any time it's powered up, and I wired it to the OBD2 port, along with bringing the car+wideband CAN bus to the OBD2 port. To run a datalog, it's now painfully simple: Plug Gauge.s into my OBD2 port, drive, download to my laptop over WiFi.


At this point, I disabled front PDC and set rear PDC to e46-ish settings with PAsoft


So everything was working great, except the rear PDC was basically always lightly triggering. I used the AVIN's I-bus connection to see which sensors were triggering and it was the side-rears, both of them. Hmm. Well, it turns out that the sensors I stole from the X5 are oriented 90 degrees apart from the e46 sensors. So even though the sensor fits into the e46 bumper strip just fine, it's flipped sideways. This matters because the PDC sensors basically throw a rectangle of signal, wider than it is tall, so if you flip it 90 degrees it will pick up the ground, constantly. Well the e46 sensors are kinda ugly, I've got a whole bunch of modern sensors here, and I've got the plastic punch piece and two self adhesive PDC mounts that I didn't use. So I measured the factory PDC bumper strips and made my own that use the modern sensors. Voila, no false triggering from the ground anymore:


And we're done until 4 more adhesive PDC mounts arrive to do the front:

[0-60motorsports]

That's awesome work!

[Bry5on]

That's awesome work!

Thank you! All in the quest for my perfect daily driver

[Bry5on]

For you nerds out there, here's a very interesting interactive datalog that contains a cold start and warmup period, along with a couple WOT pulls.
https://datazap.me/u/bry5on/aem-wide...olo=0-15-16-21

Interesting things in the data:
1) The warmup period of various parts of the engine, especially seeing the thermostat crack open and start regulating:


2) How clearly you can see the torque curve of the engine in the datalogs, inverted here because longitudinal accel is negative (dyno? who needs a dyno. We've got detailed mass information and a good datalog):


I'm also going to add a little bit more fuel between 4600 and 7000 RPM to target a smooth ramp down to 12.5:1 instead of the ~13.5:1 it's at now

[Bry5on]

Another victory claim over here. I wired the Curt trailer module in between the car light signals and the trailer module input signals using the fused (factory slot) wire I’d just run for the euro powered trailer hitch plug. This device converts the separate brake/turn signals of euro cars into the shared brake/turn signals of US cars.

I wired the park signals direct (skipped the curt module) to the license plate lights, wired the Curt left and right turn to the signals, and wired the Curt brake light to the black/yellow third brake light wire.

Then I wired the Curt left and right turn signal outputs to the left and right turn inputs of the trailer module. I followed the X5 wiring for the park lights, which gangs together left park, right park and brake light inputs and outputs that wire to the park light trailer wire.

Following that, I tested functions with a light tester and a load dummy. Both the light tester and the load dummy were detected as a trailer connected and appropriately disabled PDC upon connection. Light tester (LEDs) reported that all bulbs work correctly - park, brake, turns - but the turn signals started fast flashing as they detected a bulb out due to the low load of the LEDs. I then wired up a couple 6 ohm resistors to the turn signals ONLY and plugged that in. The turn signals and brake lights behaved normally after that and PDC was still disabled.

This was the behavior that I was looking for!
- US trailer light behavior
- Trailer detection with automatic PDC disable
- Trailer light failure detection
- Dummy load reaction for disabling PDC with my bike rack

So that’s the ticket, just gotta wire one of those Curt modules in series.



I also started work on a roller cart to hold my most used tools instead of fishing through the toolbox. I’ll take any suggestions or tips if anyone’s got some clever solutions.

[jrgatfh]

Another victory claim over here. I wired the Curt trailer module in between the car light signals and the trailer module input signals using the fused (factory slot) wire I’d just run for the euro powered trailer hitch plug. This device converts the separate brake/turn signals of euro cars into the shared brake/turn signals of US cars.

I wired the park signals direct (skipped the curt module) to the license plate lights, wired the Curt left and right turn to the signals, and wired the Curt brake light to the black/yellow third brake light wire.

Then I wired the Curt left and right turn signal outputs to the left and right turn inputs of the trailer module. I followed the X5 wiring for the park lights, which gangs together left park, right park and brake light inputs and outputs that wire to the park light trailer wire.

Following that, I tested functions with a light tester and a load dummy. Both the light tester and the load dummy were detected as a trailer connected and appropriately disabled PDC upon connection. Light tester (LEDs) reported that all bulbs work correctly - park, brake, turns - but the turn signals started fast flashing as they detected a bulb out due to the low load of the LEDs. I then wired up a couple 6 ohm resistors to the turn signals ONLY and plugged that in. The turn signals and brake lights behaved normally after that and PDC was still disabled.

This was the behavior that I was looking for!
- US trailer light behavior
- Trailer detection with automatic PDC disable
- Trailer light failure detection
- Dummy load reaction for disabling PDC with my bike rack

So that’s the ticket, just gotta wire one of those Curt modules in series.



I also started work on a roller cart to hold my most used tools instead of fishing through the toolbox. I’ll take any suggestions or tips if anyone’s got some clever solutions.

I love following your progress man, it’s like watching a rocket scientist lol

I really appreciate how skilled your are. Kudos.

[Bry5on]

I love following your progress man, it’s like watching a rocket scientist lol

I really appreciate how skilled your are. Kudos.

Thank you! I really enjoy the process of learning and this project really helps keep the skills sharp (enough) because you can dig into so many different disciplines.

And I’ve got one more victory claim before the weekend is over. The front PDC retrofit is complete. It works great, but was a bit of a bear specifically to modify the bumper beam. Other than that, everything went together as planned with the wire harness routed down the passenger side of the car. Now I can give my back some rest.

[Bry5on]

I was home with a cold this weekend, so I made myself useful by tackling a couple little projects.
1: I repaired my black 996 calipers with the same fix I did to the red ones by bonding in the loose pins




2: I disassembled my M3 mirrors and created a functional bench setup for auto-dimming, along with wire harnesses that will be trimmed to length




3: I pulled the roof rail trim off and double-stick taped it to mimic what the factory did. I used 1/8" (3.2mm) tape, where the factory looked like 2mm. It fit just fine as the land for the tape is 3-3.5mm, a wagon takes about 10 yards.

[0-60motorsports]

I have the same rear view mirror and love it. Dont even care where i threw the oval one LOL

[Bry5on]

On to the next project - does anyone have a 3D scanner in the Bay Area so I can get these dimensions right?

[Bry5on]

So I recently bought a 3D printer and I’ve been practicing by making smaller things to learn the ropes before I jump into printing the ZHP-CSL duct and isolator boot.

So far I’ve experimented with TPA, ASA and ABS. I expect ABS will be my most popular material, and to that end, here’s an ash tray insert that I designed to hold a smaller 1.54” screen, along with gauge.s circuit board and GPS receiver. A print just finished for the bigger screen version (which clears to the half millimeter basically everywhere) that might replace this one, but here’s how it looks so far:



Here’s the big one:

[Bry5on]

Progress continues..

First, I picked up the isofix cover that I had retrimmed in some e46 natural brown leather that heinzboehmer and I came across a few months back, finally:


Up next, we move back to 3D printing. I managed to fit the bigger screen into the ash tray dimensions, as above, and added some USB-C CarPlay and charge ports while I was at it. I tried to capture a bunch of photos to give a sense of the lighting match. It's actually pretty damn good, but the camera really struggles to capture it the same way as your eyes.

Open:

Closed:





After that insert as practice, I turned to the real project. A ZHP bumper brake duct to CSL airbox inlet duct and accordion flex boot. First I masked off the area to get some good measurements and a 3D scan with the iPhone faceid lidar sensor, which gave me the local area and surfaces I needed to route a duct:


Then off to the 3D printer for a first iteration. I used the entire build height of the printer, and actually needed a few mm more, so I nipped the top 5mm off of the printed part of the duct. You'll see below why this didn't really matter after installing the TPU printed flexible accordion boot. This is actually version 2:



It helped having an extra CSL flap to get dimensions, as this is an odd shape:


And after adding some mounting points to V3, we're in business. This piece uses two factory mounting points to secure in place along with the bumper opening to locate itself in 3D space. From there, the CSL airbox is installed from the top just as it is on the factory CSL. The flexible TPU piece in between has a groove that locks into the CSL flap just like stock, and is flexible enough to snap/pop in place on install. It also has double walls to locate on the duct itself to make sure it stays positioned, even as the engine vibrates at high revs. So, here it is:



You can see here how the flexible boot fully encompasses the section of print I had to notch out:


And lastly, a bit less exciting, the bracket for the PDC and trailer modules showed up from Europe, so I dropped that in:


All in all, a solid update I think.

[Projecte46m3]

Doing god's work in here!!

[Will]

Amazing work! Thanks for documenting and sharing with us.

[Bry5on]

Well well well, it appears those BMW engineers really were on to something with that flap ducting.

Background: I have both the pre-and post-flap ducting connected (ie: with the flap closed I'm drawing air from behind the left side of the bumper, but not with any sort of ram effect). You can see the termination of the post-flap intake duct in some of the photos above.

Tonight, on my drive home, I datalogged at a reasonably constant speed and rpm (~3500rpm, 5th gear, ~65mph) a drive with the flap open (briefly), then closed, then open again. At this constant speed and engine load, I saw an 8*C (14*F) in 1 minute, 15 seconds, with most of the drop happening in the first 15-20 seconds. This is a rather impactful change! Definitely more than I expected, as I already had a low temperature air source and I didn't see this behavior before doing the duct.

In the stock CSL firmware in sport mode the flap opens at 3300rpm in gears 1, 2, 3 (41mph) and at 40mph in gears 4, 5, 6. Seems logical.

Vertical yellow bars in the datalog below are when the flap opened/closed.