N52 Swap Discussion

[hoveringuy]

I'm pleased with how the engine bay looks, it's come a long way from the wiring shit-show when I first had it running! This is using a Z4 harness. I highly recommend Z4 as the base harness because of the IVM, which replaces a tangle of plugs, fuse boxes and relays with one simple, svelte little box. It combines all the relays, fuses and plugs and makes things a lot easier.

Now, the drivers side of the Z4 harness is too short to reach the glovebox, but the e90 harness is perfect, so I swapped all the wires on that side from Z4 to E90. It's simple, trust me, only takes an hour or two. On the passenger side, the Z4 harness is 2 1/2 ft too LONG, but there's not nearly as many wires and they're quite simple to shorten.

Pictures with and without the M3 beauty cover.

[hoveringuy]

Small update just to show how I ended-up mounting the ECU and IVM. It will be even neater once I'm done optimizing the wiring and zip-tying. I'll have a separate post on what I used for the wiring harness later.

[nando]

Yeah, I still believe no-MAF is the way to go - but with the changes in your engine, especially the headers - I think the fuel model is too far off for it to work correctly. Normally, you just set the AF target, the DME matches it, and you carry on with your day. As you know, it's not happening like that with your car for whatever reason. But at least with the MAF, it should be easier to get running properly even if you lose a few HP.

I thought about it this afternoon, and actually I wonder if there is more from the S54 calibration of MSS70 we can copy to your DME (MSS70 and MSV70 are 98% identical - MSS70 even has valvetronic maps, although they're calibrated for full lift). I only ever tried transferring the wideband feedback loop (although, I think on your current tune - it doesn't even have that). I never tried copying the base fuel maps or any other parameters. Since you're making similar power to a stock S54. that actually might work as a better baseline.

There's no reason this won't be a hell of a lot faster, and definitely smoother, once the bugs are worked out. I really appreciate your feedback and patience on getting this sorted - because I really want to do this conversion myself!

[hoveringuy]

N52 track day #2. Good day, learned some more stuff.

1) Nando had previously raised the minimum operating speed of the water pump to ensure that sufficient coolant made it through the core for heat. Possibly related to that I found that on the drive to the track the temperature was oscillating wildly. It would get to 1/2 on the gauge, about 205F, and then plunge down to just above the blue. It would slowly creep back up to 205 and then repeat.

I think what's happening is that the MAP thermostat is being activated. The heated wax melts and expands, which opens the thermostat some more. It gets a rush of extra coolant which cools the engine down until the wax shrinks and the cycle continues.

I pulled the electrical plug from the thermostat and the motor was rock-steady at 205F/96C for the rest of the day.

2) The car definitely has more power, but it's not usable. At WOT it absolutely runs like a beast and my top speed on the straight increased from 126 to 131, but the 95% of the time that I needed part throttle modulation it didn't do very well. So maintenance throttle in the corner was hard to get right and I was timid rolling on the gas coming out of the corner because it was moist asphalt and I didn't feel confident that power would be smooth.

So, even though the M54 had less power, I put down faster lap times with it because it was smooth, sharp and completely predicable.

I'm not running a MAF because the newer N52's infer airflow from their engine model and then fine-tune with 02 sensor feedback. I no longer have a valid engine model; with the intake and the headers I haven't just bent it, I've completely broken it.

So, the plan is to put a MAF on to have measured airflow instead of inferred.

[hubcapboy]

Speaking of, are you still planning on using the Tesla booster?

It's sitting on the floor in the garage waiting for a rod length adjustment.

[hubcapboy]

Looking forward to how your lap times improve as you get everything worked out.

I did learn something new about n52 AC compressors by getting an incorrect part today. Apparently some 2011/12 e92 328's got an AC compressor with a clutch. The part number is 64 52 6 918 749, the car I was working on had the standard clutch less compressor so I do not know if there are any climate control differences.

This is something I found and came to conclusions about... but can't reproduce them. The difference is shown here: https://www.realoem.com/bmw/enUS/sho...94#64526918749



These are labeled as with magnetic clutch, and without magnetic clutch... and boy does the diagram look like that's the case. The reason I didn't pursue this is that there's no difference in the wiring on these cars, so the part that I cared about (how to activate the compressor when I wanted A/C) was still an issue. I *believe* that this was a late hardware revision to the compressor to save wear on the bearings during periods without A/C demand, which was a shortcoming of the original design, and this was done by having the pump itself activate and deactivate the freewheel clutch with its own internal (well, external) wiring.

I believe these are both variable rate compressors and thus...

The signal *to* the compressor is the same - probably PWM - signal for either unit, as is the plug. We could wire the compressor clutch directly (probably) and engage it with the e30 body harness, but the variable pump would be in whatever it's resting state is, which I believe is 0%, resulting in a very confusing installation where the system charges and the pump cycles on but there's no pressure change from the compressor.

For people who are reading this for the first time, the M73 compressor shows up on page 20 of this thread.

[hoveringuy]

Did you just bolt on the N54 intake manifold? Are you using any kind of adaptor? :-)

It's an almost perfect port match. It needs to be slightly shimmed by about .8mm to center it, but yeah, bolts right on. N54 and N52 are very, very similar in a lot of ways.

[LukeJ]

Did you just bolt on the N54 intake manifold? Are you using any kind of adaptor? :-)

[hoveringuy]

It's just cool that you can basically get to the power of a stock S54, but with the weight of an M42 (or less, actually).

Here's compared to a highly-tuned S54 (not sure exactly what that entails..). That motor is putting down around 330whp, it's just not a fair fight. On the other hand, I also weigh 800lbs less.

Also compared it to a hot M54, which peaked at 250whp at around 6700 but had a better bottom-end.

Everyone at the shop was very surprised at what an N52 could do, for sure.

[nando]

Get me a dump of the calibration in PAsoft - should be 0x8400000 - 0x85FFFF (same mode as dumping ram). I can start looking for that spot at 2500rpm.

It's pretty impressive considering the engine is entirely stock internally. I always knew these had a lot of potential. You could do MILVS and different exhaust cam rockers (basically the same thing as a cam swap), I bet that'd get to 300hp, but honestly with 288hp at 7500rpm, I doubt it would be worth the trouble.

the 3-stage is probably better suited to the bigger cars, but the torque you're getting is plenty for a little E30. The downside was getting it all in tune since the intake/exhaust are so different from stock.

It's just cool that you can basically get to the power of a stock S54, but with the weight of an M42 (or less, actually).

[hoveringuy]

So yesterday we followed the dyno to the end of the rainbow.

The car had been just "meh" at the last track session. Speeds were down from the M54 and fuel consumption was atrocious, but it ran smoothly otherwise.

Took it to 22RPD "22nd Race Product Development" and the baseline pull showed 189 whp and the A/F was pig-rich, like 9.

Now, these guys tune BMWs all the time and make some awesome high-horsepower cars. They know ECU's as well as almost anybody, but and even they had a hard time with it. There are a lot of different maps and operating modes in the MSV70 and they struggled with finding the best way to make corrections. Ultimately, after lifting the torque limits they adjusted the global fuel scalar which was effective. There's more work to be done, however.

The graph I'm showing here shows a peak of 275whp at 6900 with torque 250 at 4700. In reality, the motor keeps pulling well into the 7's and we saw up to 288 at 7500. If we had lifted the rev limiter to 8000 I think we would have seen 300. S54 headers work that well.

There's still one little rough spot at around 2400 rpm that needs some future work but I'm pretty happy with it.

[wazzu70]

that won't help you because none of the RWD suspension will work with the AWD subframe. the pickup locations and mounts are totally different. And people who have done "RWD" swaps on the ix (using all ix front suspension), have found that the handling is.. not good.

You would need to put an ix subframe in a RWD car (have to
fab mounting location to match ix) and make custom LCA that replicate the RWD balljoint location.

OR use an ix shell with custom LCAs and strut tower mounting location (if it could even fit in the confines of the ix tower). Maybe E36 stuff would be able to fit better.

Bottom line is there is no bolt in solution or anything close.

[TobyB]

Hmmm-
now I want to try cutting and shutting an E46XI subframe down to fit the E30.

I'm sure it'd be a miserable failure, but fun to try.

t
haz welder

[nando]

that won't help you because none of the RWD suspension will work with the AWD subframe. the pickup locations and mounts are totally different. And people who have done "RWD" swaps on the ix (using all ix front suspension), have found that the handling is.. not good.

[Bmwtech07]

This may be a dumb question, but are the frames between the IX and the I the same where the subframes bolt up? It appears that a lot of the work for this swap is with the subframe/oilpan interference. I ask because I can get my hands on and IX that has a body that is rotted out beyond saving, and no interior, but the drivetrain and suspension are all there. If it all bolts up and there is enough clearance, it would be worth it to me to swap the front subframe/steering/hubs/suspension from the IX to make this work until there is a pan readily available. Is there something else that I am missing? (not including all the wiring, exhaust work, ECU tuning, ect)

I love the idea of using the Tesla brake booster for clearance on the 3stage disa manifolds. I would build this for more of a summer fun car and not necessarily for track so the stock redline/powerband would be find for my needs...for now. Speaking of, are you still planning on using the Tesla booster?