|
|
|
06-17-2009, 07:41 PM | #3 | |
Second Lieutenant
3
Rep 242
Posts |
Quote:
If you want, you can PM it to me instead of posting it (to avoid potential flames). Or you can post it at your own risk =P Now, onto business. In J Tylers thread, we were discussing how exactly the E-Diff functions in this car. Based on my experience: the car detects load bias in the differential (possibly measured at the transmission, since afaik there are no sensors on the differential) and applies the brakes to the wheel with significantly greater load bias up to a point at which it cannot compensate without doing more harm then good and the diff goes to full open action. Some scenarios to clarify my hypothesis: A) The car is stopped, tranny in gear, clutch disengaged. You dump the clutch. The power will go via the drivetrain to the wheels at which point one wheel will lose traction. The load bias is very high. The vehicle detects this and does not engage brakes. This is speaking from confirmed experience. When I have botched launches, the scenario I described is precisely what happened every time without fail or deviation. B) Same launch situation but you do not dump the clutch, you let it out smoothly. The power find the wheels but one wheel does not lose traction (and when I say lose I mean completely lose... like smokey burnout lose... Im aware that you are always losing a littel bit of traction even when the car is driving perfectly straight and normal). The load bias is low (let's estimate 10%... so one side of the diff is seeing 10% greater load than the other... where the above example 'A' would be showing something like an 80% load bias). The vehicle detects this load bias is within the range neccesary to apply it's E-Diff magic (let's say that range is like 10-40% just for kicks) braking the wheel that has lost more traction than the other and reducing it's wheelspin, which in turn allows the wheel that has lower load to "catch up". Now keep in mind Im not talking about diverting power, Im simply talking about sapping power that's already traveling to the more loaded wheel. C) Same launch, except you dont even launch. You just smoothly and slowly accelerate from the light. After all, you're just dropping your grandma off at home and don't need to give her a heart attack or something. Naturally, as stated above, there is always some loss of traction and some load bias. That's how tires work, by losing pieces of themselves. Otherwise we would get infinite mileage out of one set. Here the load bias is very low, percieveably non-existent to us lowly humans. Let's say its .5-1%. If the vehicle detects it at all (which it might not, naturally solving the problem Im about to mention) it identifies it as a "normal driving condition" and does not do any braking. The reason is that you would, obviously, prematurely wear your brakes if you were applying brakes (even just a little bit) for every moment you were driving. That's my theory. It's entirely conjecture, based only on my drag practice and my daily driving. I could be way off, but it all sounds pretty good to me and no ones called me crazy yet =P So to give my opinion to your questions more specifically: 1)Yes, but it's the inside wheel, as proven here. 2) In my experience, yes. My best launches have produced alternating dashed skid marks in the road where one wheel is spinning then stops and the other wheel spins and it goes back and forth. If you were to take these two dashed lines of skidmark and put them together they would precisely form one straight line. It's REALLY weird looking. I've never seen it before this car. 3) I would say... yes. A mechanical LSD would be doing this in infinitesimal time, essentially carying out both actions simulataneously. An E-Diff would only be doing this as fast as it's microprocessor can detect the change and react to it which might be fast but never as fast as 1/infinity. |
|
Appreciate
0
|
06-17-2009, 07:52 PM | #4 | |
My ride: bimmergirl84
22
Rep 302
Posts |
Quote:
I'll start by correcting your assumptions: - Yes, a mechanical LSD is a WORLD of difference from an open diff with traction control. There is a reason why virtually every performance/racing car has one. - The "E-LSD" IS traction control. That's exactly what it is; nothing more. Calling it an "E-LSD" is BMW marketing nonsense. - Traction Control and Electronic Stability Control have nothing to do with whether or not the car has an open diff. A mechanical LSD will supplement those programs, but does not replace them. On most cars, the TC/ESC are standard "off-the-shelf" programs made by Bosch and are the same on cars with and without a mechanical LSD. - The "E-LSD" (call it traction control because that's what it is) brakes the slipping wheel in pulses, resulting in an ON-OFF-ON-OFF power application, sending the vehicle into a bucking motion. You are correct, during the time that the slipping wheel is being braked, power is sent to the outside wheel; but again, it's in an ON-OFF-ON-OFF manner. A mechanical LSD DOES NOT send power to one wheel at one specific ratio; it sends power at a variable ratio that depends totally on the instantaneous grip available to both rear tires. No mechanical LSD splits power at a specific set ratio.... they all vary based on the instantaneous grip available. - "If the electronic LSD clamps down on the inside wheels rotor in hard jitters, it would be pretty efficient." - NO, it wouldn't. It would be very inefficient. This is why, again, virtually EVERY racing car on the planet has a mechanical LSD. A mechanical LSD acts instantaneously and constantly. An E-LSD reacts to the loss of grip that already happened. On a track, that = time lost. On to your questions: 1) No. BMW's "E-LSD" allows a massive amount of inside rear wheel spin. I'll post pics later tonight. 2) No. The E-LSD has a very low threshold where it simply stops engaging and allows all-out wheelspin. 3) Every traction control system I've tracked has been massively inefficient and in no way a substitute for a proper mechanical LSD. Hope that helps clear things up... |
|
Appreciate
0
|
06-17-2009, 08:49 PM | #5 |
Major
194
Rep 1,457
Posts |
I just wanna add that: ON-OFF-ON-OFF kinda pulse braking is I believe a result of PCM - Pulse Coded Modulation and the frequency of the ons and offs in a unit time will look like a variable ratio when you sample and quantize the signal. In theory, E-LSD might deliver power like a mechanical LSD in variable fashion.
However. I believe that e-lsd can not be as efficient as regular lsd. Because of the simple fact that braking means a loss of kinetic energy. In helical lsd you do not have this loss.
__________________
|
Appreciate
0
|
06-17-2009, 10:37 PM | #6 | |
My ride: bimmergirl84
22
Rep 302
Posts |
Quote:
What you would want to vary is the amplitude of the pulses (the strength of the brake application to the slipping wheel). And this would require very very sophisticated programming to work as well as a mechanical diff... not to mention the added variable of rear brake pad fade/wear it would incur and that would have to be calculated or sensed by the program. And even if you could pull all that off, it still wouldn't be as good as a mechanical diff because it's still reacting to the loss of traction. |
|
Appreciate
0
|
06-17-2009, 11:23 PM | #7 |
Private
6
Rep 97
Posts |
this is a great thread with a lot of intelligence, so I will make the stupid comments...
what is the downside of making lsd standard? I remember, as a kid, reading those wonderful bmw brochures with hood as the front cover (1970s i guess) and learning about lsd's from there. Every bmw had them, and that's why that had that "superior handling feel" or whatever. Now they are reserved for m cars. Anyway, I am sure there is a simple engineering-based answer, but since I am not an engineer, I don't know what it is... |
Appreciate
0
|
06-18-2009, 12:05 AM | #8 | |
AutoXer
34
Rep 683
Posts |
Quote:
|
|
Appreciate
0
|
06-18-2009, 01:06 AM | #9 |
•
19
Rep 709
Posts |
The engineers are baffled, talk to the finance guys.
__________________
|
Appreciate
0
|
06-18-2009, 07:13 AM | #10 |
Major
194
Rep 1,457
Posts |
As I said I do not believe it is as good as mech. diff, but It is close enough with electronic representation and I believe bmw brake system adjusts accordingly to brake pad wear and also it has brake assist (not sure about it).
just a question: What is the relation of engine rpm 14k-18k with TC pulses? I kinda lost you there because I do not know. and I think it is banned in F1 again in '08
__________________
|
Appreciate
0
|
06-18-2009, 09:12 AM | #11 | |
Banned
7
Rep 197
Posts |
Quote:
The reasoning behind power diversion (not efficient power diversion) when you LOCK 1 wheel is that drivetrain momentum and engine torque will remain constant along an infinitely short interval. In an open differential setup, if the brake locks on one wheel, energy loss through friction stops and it is preventing further torque transmission to that wheel. If you lock a wheel in an open differential a good deal of power is diverted to the opposing wheel. I say "a good deal" because the diff is connected to the power train and will produce some amount of drain back through that route, especially during lateral force shifting. Frictional losses should not be allowed inside a true open differential, right? Anyway, I'm surprised by the great responses. It's sounds like the ELSD can't function as it was engineered because of potential differential stresses (locking a 100 mph spinning wheel to jam all the energy back through the diff to the opposing tire, which has static contact = damage). The alternating skid mark burnout comment is also interesting. You're still only getting 1 tire worth of contact at almost any position along the displacement. If the force shifting increment sped up faster and faster with time, then a two wheel skidmark could be produced (not just approached, because of the wheel momentum). The alternating burnout also sounds BAD from a non drag racing perspective! Juggling slip between 2 wheels at some mechanically convinient interval sounds like it would be inconvinient for track dynamics. I already contacted Quaife to inquire about a helical LSD for the 135i since they make one for the 130i. I also informed them that it needs to handle ~420 hp/~370 lbft (although I converted to SI for them, of course). If anybody here has suggestions for a good differential it would also be appreciated. After I get my Procede set up at around [3000 miles crossed out] 50,000 miles or 4 years a mechanical LSD will be my first safety upgrade, followed by alignment setup and a second set of seats and 6 point mounts for track weekends. My Exige handled like an open diff f2000 car and required a lot of counterinput on open roads as a daily driver. Oh, and I just read the comment about high and constant adjustment pulses. I agree that this works for interuption but you wouldn't want it for braking, there are mechanical constraints. I'm guessing the high adjustment rates/rpm relationship is that F1 engines don't have much of a flywheel and that you don't want the RPM to flutter with fuel cut intervals. Not sure though, not my comment... Last edited by The Cthulhu; 06-18-2009 at 09:53 AM.. |
|
Appreciate
0
|
06-18-2009, 12:48 PM | #12 |
Banned
7
Rep 197
Posts |
http://www.mini-madness.com/browsepr...-and-135i.HTML
mmm... Giken diff on a 135i, that's the LSD I would've put on my Exige S. Reviews look good. |
Appreciate
0
|
06-18-2009, 04:17 PM | #13 | |
My ride: bimmergirl84
22
Rep 302
Posts |
Quote:
*Edit: Click the links in my sig to see it in action. We're at around 400hp I'd guess... we have the JB3 + Intercooler + Berk Technology full race exhaust system (catless 3" downpipes, midpipes, muffler) |
|
Appreciate
0
|
06-19-2009, 04:44 AM | #14 |
Lieutenant General
1867
Rep 13,043
Posts
Drives: BMW M340I G20
Join Date: Aug 2007
Location: UK
|
Taking in mind why an open diff was developed in the first place, doesn't an LSD tend to push you straight on in the corners, which might be unsettling for the average BMW driver.
|
Appreciate
0
|
06-19-2009, 06:33 AM | #15 |
Banned
7
Rep 197
Posts |
A locking diff will push you, I'd think a helical LSD would as well. With ~370 WHP, you're going to be using the gas to combat understeer, I'd imagine. Slow-in and fast-out, and camber and toe changes would probably help, too.
Is there a reason for using a Quaife diff rather than a OS Giken diff? Does it have anything to do with driving line or the existing ELSD that's engaged when all other nannies are off? Also, are you using a 1-way, 1.5-way, or 2-way? Oh, and one more question: what's the price and process to convert a manual with welded crown ring to the Quaife? |
Appreciate
0
|
06-20-2009, 01:34 PM | #16 |
Lieutenant
38
Rep 566
Posts |
Great summary, right there - everyone read the above statement to yourself a few times.
__________________
|
Appreciate
0
|
06-20-2009, 06:30 PM | #17 | ||
My ride: bimmergirl84
22
Rep 302
Posts |
Quote:
Does an LSD "push you straight on in the corners"? - NO. A helical/Quaife-type LSD is completely OPEN on corner entry (i.e., during deceleration). It only locks up under power (acceleration). Quote:
No. It's "Fast in, fast out." I don't know what type of diff the OS Gyken is. 1/1.5/2-way refers to different styles of clutch-type LSDs; not helical LSDs. All helical LSDs are 1-way only (meaning they only lock up during acceleration) The price... I'm not sure, I think it's kinda pricey. Talk to Harold @ HP Autowerks. |
||
Appreciate
0
|
06-30-2009, 12:32 PM | #18 |
Banned
7
Rep 197
Posts |
The Giken is a clutch type, but a very nice clutch type. It would definately push and drift, since it's force locking. I'm kinda stuck wondering which is best. The main downside to the helical is if your wheel lifts it becomes an open diff, but that's fixed by the ELSD function.
The Giken is a true locking, driftable diff, and could prove to be fun. I don't know if it would be too understeery if I'm not spinning the wheels, or if the ELSD would brake a wheel or anything, thereby burning up the diff clutch. I would probably go Quafi on this car, and pull my whole case out and mail it so that I only have to spend $500 extra due to my welded crown ring. If I chose that route, would I be stuck with the 3.08 ratio? If so, that sucks. |
Appreciate
0
|
09-16-2009, 03:46 PM | #19 |
New Member
1
Rep 17
Posts |
quick question i have the stock e lsd and was wondering since this giken lsd is force locking that means that both wheels will stay spinning even during deceleration?
|
Appreciate
0
|
Post Reply |
Bookmarks |
|
|