24V AWD Swap Knowledge Base

[The Dark Side of Will]

I've had a saved search on eBay Europe for 2.35 and 2.47 BMW front diffs for liiiiiiiiiiiiiiiiiiiiiterally a few years now. They'd come up every now and then for 600-700 Euro, but mostly the E90 chassis version, which I don't think I can use. I already acquired an E46 2.47 front a while back, but had also snagged a 2.35 rear (188mm unit). The rear 2.35s & 2.47s are WAY more common. I paid more to ship mine than I did for the unit. An E46 2.35 front finally popped up recently in Vilnius, Lithuania for 200£... not sure why he priced it in £ instead of Euro, but whatevs. Shipping from the Baltics is always surprisingly cheap as well, so I jumped on it. It arrived yesterday.

So now I have the 2.35 front and rear gearsets. Yeah, that's really tall, but the 5 liter S62 makes more low RPM torque than the 6 liter LS2 and the E30 is a REALLY light car compared to an E39 M5--which I've heard also runs GREAT with 2.65 gears (210mm unit), so I think the taller gearing will work great with my application. Also, E30's use short tires.

I also have a Euro market E53 3.07 front diff. It appears actually to use the same case as the numerically higher ratios, so I that was an expensive piece of knowledge to gain.

My plan is to put the E46 2.35 gears into the E53 3.07 case with a Quaife for a Z3 1.9 in order to have the most expensive front diff ever installed into an AWD BMW.

Search string, since I actually had to read eBay's documentation to set it up:

bmw (2.35,2.47) (vorder,vorderachsgetriebe,vorderachsegetriebe,front differential), Europa

And here is an article on boolean searching on eBay, since their documentation on it sucks donkey balls:

https://community.ebay.com/t5/Buying.../td-p/29941476

[The Dark Side of Will]

mw044 has ATC700 T-case and if I'm not mistaken, an Arduino-based controller for that..



I did M50 swap in my '86 iX this year with all-E34 stuff. So far haven't had control for center lock but I've designed and built my own Arduino-based controller for it. It gathers ABS sensor signals and will communicate with ECU. I've measured the center diff lock breaking torque is more than 200 Nm at under pwm 50 % duty so it's quite capable. What I'm unsure of is the how will the lock wear if used on the limit instead of fully driving it on/off. Ideally I'd like to allow just a bit of rear wheel slip and hold it there, but the safe way for sure would be to set a threshold after which the lock is ramped fully on and held there until TPS goes down and then ramp it down to 0 %.

I wanted to have permanent awd and just control the lock but if I would do this now, I'd really consider x-drive tc again.. The total driveline ratio is short (by starting point was 3.64 / G260) and 6-speed manual would make the already long package even longer.

The car is still very much in progress and not nearly finished, but this thread needs a picture once in a while so here it goes:



No, it won't stay n/a but the plenum is also there to clear brake reservoir. I swapped E34 525iX booster/mc as well, and it's tight with stock manifold/tb even though I moved the booster ~2cm away from the engine.

This thread and especially Nisse helped the swap a lot :)

Also, have you looked at this for transfer case control?
https://www.full-race.com/store/niss...kyline-gt-r-1/

As long as the E34 T-case accepts a PWM input for the locking clutch, then generic center diff controllers for Subarus, Evos and Skylines will probably work.

[The Dark Side of Will]

mw044 has ATC700 T-case and if I'm not mistaken, an Arduino-based controller for that..



I did M50 swap in my '86 iX this year with all-E34 stuff. So far haven't had control for center lock but I've designed and built my own Arduino-based controller for it. It gathers ABS sensor signals and will communicate with ECU. I've measured the center diff lock breaking torque is more than 200 Nm at under pwm 50 % duty so it's quite capable. What I'm unsure of is the how will the lock wear if used on the limit instead of fully driving it on/off. Ideally I'd like to allow just a bit of rear wheel slip and hold it there, but the safe way for sure would be to set a threshold after which the lock is ramped fully on and held there until TPS goes down and then ramp it down to 0 %.

I wanted to have permanent awd and just control the lock but if I would do this now, I'd really consider x-drive tc again.. The total driveline ratio is short (by starting point was 3.64 / G260) and 6-speed manual would make the already long package even longer.

The car is still very much in progress and not nearly finished, but this thread needs a picture once in a while so here it goes:



No, it won't stay n/a but the plenum is also there to clear brake reservoir. I swapped E34 525iX booster/mc as well, and it's tight with stock manifold/tb even though I moved the booster ~2cm away from the engine.

This thread and especially Nisse helped the swap a lot :)

I think Nisse mentioned a controller that could do fully locked or fully open on the xDrive T-cases, but didn't have capabilities beyond that.

Have you tried hitting the E34 T-Case with 12V to see what the locking torque is? I @$$ume you measured that 200NM number measured by trying to twist the front output while holding the rear?

[nando]

It looks like he modified the firewall and pushed the engine back.

[The Dark Side of Will]

How'd you fit the oil sump and diff? Frankenpan?

[muuris]

You're right. I just remembered he used the original ix transfercase in that build. Bummer.

mw044 has ATC700 T-case and if I'm not mistaken, an Arduino-based controller for that..



I did M50 swap in my '86 iX this year with all-E34 stuff. So far haven't had control for center lock but I've designed and built my own Arduino-based controller for it. It gathers ABS sensor signals and will communicate with ECU. I've measured the center diff lock breaking torque is more than 200 Nm at under pwm 50 % duty so it's quite capable. What I'm unsure of is the how will the lock wear if used on the limit instead of fully driving it on/off. Ideally I'd like to allow just a bit of rear wheel slip and hold it there, but the safe way for sure would be to set a threshold after which the lock is ramped fully on and held there until TPS goes down and then ramp it down to 0 %.

I wanted to have permanent awd and just control the lock but if I would do this now, I'd really consider x-drive tc again.. The total driveline ratio is short (by starting point was 3.64 / G260) and 6-speed manual would make the already long package even longer.

The car is still very much in progress and not nearly finished, but this thread needs a picture once in a while so here it goes:



No, it won't stay n/a but the plenum is also there to clear brake reservoir. I swapped E34 525iX booster/mc as well, and it's tight with stock manifold/tb even though I moved the booster ~2cm away from the engine.

This thread and especially Nisse helped the swap a lot :)

[The Dark Side of Will]

Hacking through the ins and outs of using the ATC-700 or similar T-Case: http://www.corner-carvers.com/forums...ad.php?t=50342

The actuator is actually not a stepper motor. It's a DC motor with a Hall effect sensor to count revolutions. Many other "on demand" T-Cases use a similar motor arrangement.
HOWEVER, only BMW uses a worm gear drive that can not be back-driven by the load from the clutch mechanism.

In most applications, the controller PWMs the motor in order to control motor torque and thus clamp load. As the duty cycle goes up, the motor torque--and thus clamp load on the clutch--goes up. When duty cycle goes down, the motor torque goes down and the clutch relaxes, transferring less torque to the front driveshaft.
In a BMW, instead of just reducing PW and allowing the mechanism to the relax, because of the worm gear drive, the controller has to actively drive the motor the opposite direction in order to reduce front driveshaft torque.

That means that the center diff controllers on the market which output a PWM signal will not work with the BMW ATC series T-Cases.
Such aftermarket controllers could work with the E34 T-Case, however.

Example controller: https://www.full-race.com/store/niss...kyline-gt-r-1/

I may have to buy an ATC-700 to try it out, but a PWM-style motor may be adaptable to the ATC-700. There are T-Cases on the market that work pretty much EXACTLY the same as the ATC-700, but use a conventional PWM motor instead of BMW's weird setup.

[The Dark Side of Will]

Right... just need to know how many teeth per revolution.

GM actually did exactly that when they started using the F40 six speed manual from the Saab 9-3 in the Pontiac G6. The Saab used Bosch electronics and ECU got vehicle speed from the ABS/SC over the bus. The GM system in the Pontiac still needed a hardware VSS, so they just installed a reluctive sensor that read the final drive gear teeth directly.

[Eskie]

The Quaife for a Z3 1.9 bolts into the E53 front diff.

A Quaife doesn't necessarily make them both turn the same speed... it just keeps the inside from lighting up.

Ahh I see. I just re-read where you mentioned reading it off the front sprocket. I bet it wouldn't be that hard to drill a hole and weld a bung for one of those threaded hall effect sensors. I guess you could put it on either the transfer case or front diff. You wouldn't need a missing tooth to read speed right?

[The Dark Side of Will]

The Quaife for a Z3 1.9 bolts into the E53 front diff.

A Quaife doesn't necessarily make them both turn the same speed... it just keeps the inside from lighting up.

[Eskie]

In a corner... maybe? Could average the left and right front wheel speeds and multiply by the diff ratio to get the front driveshaft speed.

Driving the stepper motor wouldn't be a big problem... just have to have a quad driver that can handle enough current.

My project is to drop an S62 into an E30 iX. The car will end up front heavy. I intend to use a Quaife front diff and open rear to get the car to turn. Applying enough torque to the front driveshaft to make that Quaife useful will be very important.

Which front diff housing are you planning on running? When you say quaife, your taking about a limited slip unit? With a LSD in the front you could get a pretty good wheel speed reading from one corner. I say this because with a LSD both front axles should be turning relatively the same speed, right? I guess other options would be to run a speed sensor off the front driveshaft probably.

Regardless an S62 awd e30 is going to be crazy. The heavy ass E39 that the engine is in is already nuts.


Sent from my iPhone using Tapatalk

[The Dark Side of Will]

In a corner... maybe? Could average the left and right front wheel speeds and multiply by the diff ratio to get the front driveshaft speed.

Driving the stepper motor wouldn't be a big problem... just have to have a quad driver that can handle enough current.

My project is to drop an S62 into an E30 iX. The car will end up front heavy. I intend to use a Quaife front diff and open rear to get the car to turn. Applying enough torque to the front driveshaft to make that Quaife useful will be very important.

[Eskie]

Could you use one of the front wheel speed sensors for front input?

As far as getting it to lock and unlock I could probably figure out how to drive it. I’ve driven stepper motors with an arduino before. I have a torn down ATC transfer case that I should start messing with. I could try to build a driver and mess with it on the bench.


Sent from my iPhone using Tapatalk

[The Dark Side of Will]

(which means input & front/rear output speed sensors, chassis G-meter, maybe some reading of engine torque)

Duh moment: input and rear output speed will always be the same. This could be substituted using the VSS from the rear diff with zero additional hardware in the car. One could probably use a sensor to read the teeth on the front output sprocket to get front output speed. But after that it would still need steering angle, chassis G's and probably something to do with engine load/torque output.

[The Dark Side of Will]

That would be really sweet. I'd probably experiment with and arduino to see what I could get going but before that I'd probably just have an on/off switch--unlocked/full lock.

It has to be an ardruino or similar because the clutch is controlled by a stepper motor. Even with just a "lock/unlock" switch, there would need to be some logic to step the motor in until it stopped, then stop trying to drive it... then step the motor back out until it was at the other end of its travel.

Nisse, I think, noted some type of European controller. I asked about its capabilities, and he said it was limited to "Locked" and "Unlocked".