Black & Tan 332iT

[0-60motorsports]

Okay, let's get rid of some more NVH. Vibration this time.

I pretty much got the SS headers and section 1 to idle as well as stock by strapping the two pipes together, but after more driving I still wasn't satisfied with what I was perceiving to be some persistent exhaust natural frequency resonances. There were three frequencies that were bugging me; one at very near, but not quite, idle speed, one at about 1700rpm and one at about 3300rpm. I figured these were associated with the first, second and third natural frequency modes of the exhaust.


So let's do something about it to arrest those modes. Well, what's the best way to neutralize them? To shorten the unsuspended length of the exhaust and shift the frequency modes much higher, way up or out of the excitement of the RPM band completely. You can see using the above math that the natural frequency scales to 1/L^2, where L is the length of the unsuspended exhaust in this case. So let's use a trick from several other non-s54 BMWs used before, during and after this generation. We're going to make a bracket that ties the exhaust to the back of the transmission, using its existing but unused mounting bolts. Since the transmission and exhaust are both rigidly mounted to the engine, these parts already move together, and tying them together should not introduce any weird relative motion. And since the transmission is significantly further toward the rear of the car than the headers, this will effectively reduce the length of the exhaust left to vibrate 'freely' (note that the exhaust hangers don't do much to arrest natural frequency induced oscillations, that's what those non-m e46 cylindrical dampers do). The one thing to keep in mind is that the exhaust grows as it heats up, so the flexural orientation of the mount, and giving it enough length to bend elastically, is important.

After an iteration of carboard engineering, to borrow a term from Heinz, I designed a three bracket system, laser cut it and bolted it to the transmission. I then used one of my now favorite 2.5" exhaust clamps to attach the flexural brace directly to one of the exhaust tubes. The 'bang on the exhaust with my fist test' noted a massive reduction in vibration of the exhaust.

https://www.youtube.com/watch?v=uvz95Sx9ySc

Here are a few shots of the brackets installed:




And with the splash guard in place, showing full clearance everywhere:




What's the verdict?

Very big improvement, I was genuinely surprised at how effective this is. A bigger and broader improvement from this change than the other two I'd made to clamp the pipes together. It idles more like an m54 now than an s54, which I think is a very good thing and it's essentially imperceptible when switching the AC on and off, which wasn't the case before. The resonance near idle and at 1700rpm are absolutely completely gone, and the resonance at 3300rpm seems a bit attenuated and spread out a little, now between 3100-3300rpm. This might be the exhaust's new first excitation mode.

Here's the thing though, it affected essentially the entire rev band under 5000rpm in a big way. The engine is so much smoother when revving, and it's especially noticeable when blipping the throttle, or when engaging the clutch and pulling away from a stop. I hadn't noticed or pinpointed this previously, but there was a 1-2 oscillation change in the exhaust when doing those things. Unnecessary sloppiness. Now the engine makes rev changes and pulls away from a stop like a 330i - smoothly and without any secondary motions. Also more like my previous porsches, responsive with the only motion being the motion that matters. And to top it off, some of the resonance and vibration above 5k stayed, so it still has a bit of that raw supersprint header feeling when you're really on it and high in the rev band. Personally I think I'd rather it be smooth, but I expect most folks would appreciate keeping the feeling of those higher frequency vibrations.The whole rev band is very noticeably smooth now under 5k, even under load, with the exception of a less prominent bit of vibration between 3100-3300. And this is going to sound crazy, but I can swear I hear the lower frequencies in the stereo better.

I've got a complete 2.5" supersprint exhaust here (-20lb), as well as a 330i ZHP muffler (-10-15lb) that I'll likely swap in to drop the mass at the end of the exhaust to help raise the natural frequency even more. I'd still like to eliminate the 3100-3300rpm subtle vibration to make it non-m-smooth cruising on the highway, as that's 75-80mph for me. An added bonus with the ZHP exhaust is that it has a vibration damper already on it that I can swap out to tune out that last vibration mode.

I really want to A/B this against a stock M3 (will do soon), as I'd bet this is smoother than the M3 came from the factory now. And this is with 75k mile, 10 year old motor and trans mounts. I'll be A/Bing against a freshly built motor with all new mounts, so this should be a conservative comparison. I also have new mounts waiting for the oil pan gasket job I'm putting off.

So the ultimate daily driver gets even better. Hope you enjoyed this and I'd recommend doing the same to your car if you have 2.5" SS headers and section 1.

That's awesome!! If you have some files which we can print and cut the brackets out or if you are selling some that'll be great! Really need to do this to my M3.

[Ubaderb]

Rainy day today, so I did a little work on the car again. Today’s task was to tighten up the steering with two specific changes: 1) New lower steering column bushings and 2) an LPSR solid steering shaft

The lower column bushings made a big difference with how perfectly solid the steering wheel is now placed in my hands. The new bushings were actually a different design and shape than the loose ones that came out, and I expect these to stay tight and perform better over time. This was a surprisingly big change to the firm feel of the car in the hand.




Second was the LPSR steering shaft. Because this shaft doesn’t have a collapsible section anymore for installation, it is slightly shorter than the original, but the column has plenty of travel to accommodate this. I was also concerned that it was clocked differently so I purchased a spare plastic cover for the rack splines and trimmed the alignment fin off to allow any clocking at the rack side. To install, I locked the steering column in the center aligned position, and was then very careful to not disrupt the rack’s position as I removed the stock parts and installed the new ones. The result was a perfectly aligned steering wheel on the test drive. Here’s what it looks like installed.


Overall, this is a fantastic change to the car. It really makes the whole car feel firmer and much more direct, and actually reduced! vibration from the steering wheel. Well, it might be more accurate to say it added a slight bit of high frequency feedback, but removed the sloppiness and low frequency vibrations of the stock parts. Steering inputs are way more direct and controlled and it’s my opinion that the car should have come from the factory this way. I’d say this is in my top 3 for changes I’ve made to the car, 1 being seats, 2 being the Slon and v-brace chassis stiffening, with 3 being the more direct steering. Hope this encourages more folks to make this change!

Any tips on the install for the steering shaft besides the plastic cover? Or is it pretty straight forward? Mine should be in this week!

[Bry5on]

Any tips on the install for the steering shaft besides the plastic cover? Or is it pretty straight forward? Mine should be in this week!

Awesome. No real tricks, other than being super cautious to not turn the steering wheel while you're installing. You can re-clock the plastic piece if you'd like to keep the alignment fin, but I opted not to as I was too paranoid of accidentally steering the rack while removing/installing it. Good luck and hope you enjoy it as much as I do!

[Bry5on]

Ok, so I've been driving my car around with the stock airbox (and now the no NVH exhaust) and I love how responsive the throttle is at low revs and how smooth the tip-in is. Honestly it's sufficiently better than the CSL setup that it had me considering never going back. But I miss the noise. So here comes another idea..

What's different about the CSL engine, fundamentally from its principles of operation? Well, at low revs neither intake is going to be a restriction, so you can reasonably assume actual performance in the engine will be the same. The exhaust between and CSL and euro M3 from a flow perspective seems to be identical, so really it's just the cams, or maybe just the intake cam and some intake runner changes. The intake runners are short, they're only likely to go into effect really during high rpm operation. So let's assume that these are the big changes, then ask ourselves where in the tune this would be different between CSL and M3? You'd see it in the vanos/cam timing maps for sure, then probably some ignition timing, and definitely lastly fueling and injection timing. In a prior episode we dialed in the fuel, which was a huge improvement, but it's not to the point of the responsiveness of the M3. There's nothing in physics that says we can't achieve the same outcome, so let's get started.

I stripped out the relevant part throttle tables from the CSL and M3 maps: Fuel injection end angle, ignition timing, intake cam advance, intake cam advance during cat heating, exhaust cam advance, and exhaust cam advance during cat heating. The interpolation targets between the CSL and the M3 are all different, so you can't actually directly compare them. But you can build out a complete map, do the interpolation, color chart them, then compare directly. I haven't seen anyone do this, so I did it myself. The results are enlightening! There are lots of changes in the vanos timing that are likely due to the cam difference (and some resultant tweaks to ignition timing). This is likely why folks with CSL or hotter cams report smoother running engines than folks with M3 cams and CSL bolt-ons.

My car seemed pretty happy above 3500rpm with the CSL setup, but was never quite right at lower speeds. I'd still get pinging at WOT under 2k, the throttle response wasn't consistent and snappy under 3k, and there has always been some inconsistent odd behavior at 2700pm. I wanted to see if the maps explained this behavior, and for the most part, they do! So, I was happy with the high rpm performance of the CSL, but I want the low rpm performance of the M3. Enter the Mullet Tune(tm).

I've taken the best of both tables and hybridized them to hopefully keep the benefits of the CSL power and noise at higher rpm, but the smooth operator of the M3 at low rpm. Here is what that looks like visually:

Intake cam timing (CSL, M3, Mullet):

Not a huge change on the intake side, but there's a big change at low rpm, high load that's notable.

Exhaust cam (CSL, M3, Mullet):

Look how different the exhaust cam operation is at low RPM!! That big step change in the M3 map about 1/3 of the way to the right? That's 2700rpm! The data, it speaks to us.

Ignition timing (CSL, M3, CSLvM3 comparison, Mullet):


The next task will be to load this tune into the car, repeat the fuel map trimming process of a few months back, then see if the Mullet delivers. Business down low, party up high. Wish me luck.

[heinzboehmer]

Man that 2700 rpm weirdness is super noticable. Especially after fixing the partial throttle fueling. Excited to see what these maps make your car feel like.

[Bry5on]

Well, the Mullet Tune is an absolute resounding success! This morning, I removed the stock airbox, added some extra foam where the filter didn't seal well (there was sand behind the filter!), then reassembled with the CSL airbox. I also measured the length of the trumpets in an effort to calculate any shift in peak power resonance. The trumpets in the stock airbox and the CSL airbox are nearly identical in length, so I had even more positive reinforcement that the Mullet would be a likely success.





After that, I uploaded the Mullet Tune and went out for a drive. From v1, I only had to make a couple corrections: pull a degree of timing out from 1900-2200rpm at WOT, and tune the exhaust cam advance around 3700-3800rpm under mid-throttle. I haven't touched the fuel maps yet, but it's just about on the money as is.

It works exactly as expected, smooth like a stock M3 down low, and rips like before up top. Downshift blips are doing just what the pedal asks for, super repeatably and responsively, there's no lumpiness in the rev band under 3k anywhere, and you can go WOT at any time without pinging. I had a good long 6th gear uphill WOT pull from 1200rpm to 3100rpm that was smooth sailing the whole way. Tip-in is smooth and not jerky, making shifting a little smoother and easier. I was enjoying driving around at low revs nice and quietly, and finding excuses to downshift from every perceivable rpm just to get the satisfaction of solving the CSL tune issues. So this is the hot ticket for a bolt-on CSL conversion! This really increases the jeckyl/hyde nature of the car even more, it's so docile and smooth around town, but when that valve opens at 3300rpm, the game changes. A solid fix!

[0-60motorsports]

Thats awesome! You really need to start doing tuning LOL

[karter16]

Very cool! Funny I was thinking the other day that it's a shame we don't have an off the shelf option for CSL airbox with M3 cams - your work is proof that this could be an option!


Sent from my iPhone using Tapatalk

[Bry5on]

Very cool! Funny I was thinking the other day that it's a shame we don't have an off the shelf option for CSL airbox with M3 cams - your work is proof that this could be an option!

More than could be an option, it IS an option!

I'm a couple more tweaks in and I've dialed both cold start idle and warm idle pretty well now. I'm within ~2% of nailing warm idle by the AFR gauge, so I think it's ready for a full fledged part throttle tuning session now. It really doesn't feel like it needs it, so I'm not exactly making that a high priority.

[karter16]

More than could be an option, it IS an option!

I'm a couple more tweaks in and I've dialed both cold start idle and warm idle pretty well now. I'm within ~2% of nailing warm idle by the AFR gauge, so I think it's ready for a full fledged part throttle tuning session now. It really doesn't feel like it needs it, so I'm not exactly making that a high priority.

Very exciting,

Do I take that to mean this is something you're planning to offer once it's “finished”?


Sent from my iPhone using Tapatalk

[Bry5on]

Very exciting,

Do I take that to mean this is something you're planning to offer once it's “finished”?

If you're local, we can definitely arrange something. I've got to figure out the value of this and how I might go about that, especially if I were to do this for anyone not local. There's no way I could offer the iterative support for WOT tuning that some of the other guys do, I don't think the numbers would pencil out as worth the cost for anyone.

[maupineda]

Is the MSS70 in your scope :P. This is very cool an methodical approach as it should, and if fact, as is done in product development.

[Bry5on]

Is the MSS70 in your scope :P. This is very cool an methodical approach as it should, and if fact, as is done in product development.

Oh man, that would be a cool project - I'd love to have a Wideband O2 sensor in-the-loop conversion for the e46 chassis. But no, not in scope

The tune is pretty well dialed in now, I'm hitting 14.7 +/-.3 AFR (2%) in pretty much every cell in the map and I've got WOT holding steady around 12.5:1, trending down to 12:1 at redline. Cold starts and idling is nice and smooth. All said, it didn't take much to make it perfect after all of the other map work!

[puma1824]

Well, the Mullet Tune is an absolute resounding success! This morning, I removed the stock airbox, added some extra foam where the filter didn't seal well (there was sand behind the filter!), then reassembled with the CSL airbox. I also measured the length of the trumpets in an effort to calculate any shift in peak power resonance. The trumpets in the stock airbox and the CSL airbox are nearly identical in length, so I had even more positive reinforcement that the Mullet would be a likely success.





After that, I uploaded the Mullet Tune and went out for a drive. From v1, I only had to make a couple corrections: pull a degree of timing out from 1900-2200rpm at WOT, and tune the exhaust cam advance around 3700-3800rpm under mid-throttle. I haven't touched the fuel maps yet, but it's just about on the money as is.

It works exactly as expected, smooth like a stock M3 down low, and rips like before up top. Downshift blips are doing just what the pedal asks for, super repeatably and responsively, there's no lumpiness in the rev band under 3k anywhere, and you can go WOT at any time without pinging. I had a good long 6th gear uphill WOT pull from 1200rpm to 3100rpm that was smooth sailing the whole way. Tip-in is smooth and not jerky, making shifting a little smoother and easier. I was enjoying driving around at low revs nice and quietly, and finding excuses to downshift from every perceivable rpm just to get the satisfaction of solving the CSL tune issues. So this is the hot ticket for a bolt-on CSL conversion! This really increases the jeckyl/hyde nature of the car even more, it's so docile and smooth around town, but when that valve opens at 3300rpm, the game changes. A solid fix!

wow! Is there any way to do the tune remotely? I'm in dire need of assistance.

[Bry5on]

Well we've now got a total of 4 cars (two 332it wagons, two M3s) with the Mullet Tune and so far the feedback from the 3 with OEM CSL setup and SSv1 headers is that it's working quite well, no more funky low RPM or throttle blipping behavior. Next up we'll get another 332i sedan setup with stock M3 headers (same as the second 332it) and we can make sure the tune works just as well on those cars. Very pleased with this, it's no longer the end of the world if you catch yourself at low rpm and need some torque.

Here are a few pictures of the car being enjoyed as it should be (among friends)!



I can't stress enough how much of a blast it is to blast down some trails on the mountain bike, then hammer on the car through the mountain twisties on the way home. Having the bike on the back really helps it hook up on the throttle, a little bit of that 911 feeling. The bike and hitch are about 90lb together and give it a 46.5/53.5 F/R weight balance (49/51 without) (a 991 Porsche 911 is 42/58, and a cayman is 47/53). It's the absolute greatest.

The AVIN head unit crapped out on me, so I bit the bullet and ordered an XTRONS. The XTRONS is significantly better and easier to use, more polished (although not perfect, it's much better). Since this has a nice EQ, I broke out the UMIK-1 microphone and REW (Room EQ Wizard) to correct the frequency response. This car has BavSound speakers, so if you're looking to grab these settings, be sure to make sure you have a wagon with HK and BavSound speakers first!

On to the data:
Frequency response of the system set (both 1/3 and 1/6 smoothing shown):
- Flat with HK off (red)
- EQ applied with HK off (green)
- EQ applied with HK on (purple)
RT60 Decay response (after EQ):
Spectrogram plot that highlights in-cabin resonances (after EQ):

And the EQ settings:

Plus a little rear bass bump (not included in any of the EQ plots above) because the woofers in the wagon are on the rear channel, facing the tailgate:


Overall happy with the XTRONs unit now, and I prefer all of the response tweaks that the HK system adds even after EQing with it off - it really does appear pretty well tuned to the e46 cabin and speaker placements.