ALFAGTV6.COM

zar162b4 · Post by **zar162b4** » Thu Apr 03, 2008 10:32 am

Hi,
I am going to install 4 clutches on the LSD of my car and I need your help. Probably it is a stupid question but I would like to ask you a confirmation on the correct assembly order.
Considering the picture I have attached, on the stock LSD differential the correct assembly sequence, as picture shows is:
1-2-2-3-2-4-gears-4-2-3-2-2-1

If I install 4 clutches, I guess the sequence would be:
1-2-3-2-3-4-gears-4-3-2-3-2-1

May I ask you to confirm this and, if this is wrong, to tell the correct sequence?
Also, I would like to ask if the clearance to be read when applying a load of 10kg (22lbs) still must have a value between 0,1 and 0,2 mm.
Thank you very much
Regards
Oscar

Jim K · Post by **Jim K** » Thu Apr 03, 2008 10:12 pm

Your proposed order is correct; it couldn't be otherwise, as there's no room to fit another '2' type between the '3' and '4'-this would be the proper thing to do instead of using the internal cage as a thrust/friction surface.
Going by load and measuring packed clearance is usually wrong even if its the makers theoretical procedure. By following it you will probably never see the specified preload. I suggest you clean everything meticulously, assemble dry, torque 2 cover bolts at 2.7-3kgm (21ftlbs). Insert stub axles and if correct, you should measure from 5-6kgm (40ftlbs) trying to turn one with the other in a vise. If you see less, you need thicker plates, if more, you need thinner ones. When coated with proper LSD oil, the torque value will be ~half of what you measured dry. With 2 friction plates, your final (wet) torque is 1.5kgm (11ftlbs) and with 4 plates, 3kgm (22ftlbs). A gradual degradation of these values is expected over time, not involving any abnormal component wear. Track time accelerates this effect and re-adjustment will be required to restore values. Quicker degradation must be expected with new friction plates vs old ones, as surface wear is initially faster with new parts.
Jim K.

zar162b4 · Post by **zar162b4** » Thu Apr 03, 2008 11:36 pm

Hi Jim,
Thank you for your reply. The oil I would use is probably going to be the Castrol Hipoy LS(please see here http://www.rgracing.co.uk/Castrol_Transmission_Oils.htm). Any experience with it? Would you suggest a different one?
Thanks and Regards
Oscar

Mats · Post by **Mats** » Fri Apr 04, 2008 1:39 am

This is where me and Jim do it differently...

I set up my LSDs with the measuring of axial lash only. There must be no interference (preload) when bolted up. Clearence up to a tenth of a millimeter is good.

The only VERY small issue with the clearance is some added backlash in the gearbox but it is dwarfed compared to the rest of the lash contributors and is nothing you will ever notice when driving.

The (IMO) very large issue with setting up the LSD with zero clearance or preload that it is very difficult to control how much initial Torque Bias Ratio (TBR) you will get and if you have preload you will get TBR and probably quite a lot because there is nothing that yields, preload builds up very fast with small interference. To have preload in the LSD is very bad unless you have serious problems with oversteer, TBR will always act yaw-damping, meaning that it will always try to keep the direction the car travels in the direction of the rear wheels (i.e. understeer).
There is an exception to the rule of course and that is when you are cornering hard and accelerates hard so that you break traction on the inner wheel, then the TBR will induce oversteer. This normally only happen in very tight radius corners or if you have a lot of power in the engine.
Naturally this is already the nature of the LSD, when you accelerate it creates TBR so it's already covered. You have also created a bigger ratio by adding twice the friction surfaces so you will have plenty of TBR even without having preload (when accelerating or engine breaking that is, not when coasting which is the way it should be)
The only thing you do by adding preload is moving the baseline of the TBR up so that it can never be fully opened. This will also create lots of heat and wear.

It's your call.

zar162b4 · Post by **zar162b4** » Fri Apr 04, 2008 8:28 am

Mats,
If I understand correctly, you follow the Alfa/ZF procedure to shim the differential? You apply a load of 10kg?
Thanks

Jim K · Post by **Jim K** » Fri Apr 04, 2008 8:40 am

I've set up LSD's this way for the historics guys here and they just love it (power levels from 140-190hp). I also did 3 or 4 for Ron a couple of years ago when I was up there and he said they were great in every respect. Having driven mine in the track before and after this óperation' I can say I wouldn't have it any other way. Almost zero clearance (but zero preload) was just not as good, even in the 'feel' area. But, like you said, everyone makes their choice.

Different opinions aside, the most important change a serious driver can make to the LSD is altering the ramp angles to achieve two things: increase lockup on acceleration and decrease understeer off-pedal. Std ramp angles are 45-45* and changing them to 70-30* for example will transform cornering. The machine work involved is precise (milling machine) and proper component hardening must follow.
On the oil issue, I use Valvoline GL-5 80-90LS and have no plans to change it. I've opened up my diffs with it after a year or so and there is zero parts wear (except the lower preload)
Jim K.

zar162b4 · Post by **zar162b4** » Fri Apr 04, 2008 1:06 pm

Thanks for your suggestions Jim

Regards
Oscar

Mats · Post by **Mats** » Mon Apr 07, 2008 1:27 am

zar162b4 wrote:Mats,
If I understand correctly, you follow the Alfa/ZF procedure to shim the differential? You apply a load of 10kg?
Thanks

I have never had to shim anything as they have always been within spec, the 10Kg preload is there only to make sure the components are properly seated and usually does not change the measurement at all.

As you may see from Jim and my replies we have completely different design filosofies. I tend to add as little lockup as possible to only keep the inner wheel from spinning but if I read Jims post correctly he alters the vehicle dynamics of the chassis by playing with diff preload and abount of TBR, this will be effective if you have no problems with the side effects but to me it's not the way to go.

Jim K · Post by **Jim K** » Mon Apr 07, 2008 7:19 am

As a last point, I will note that some years ago, BMW used ZF LSD's in their larger models, which were identical to ours only everything was a bit larger in diameter. These LSD's had a strong Belleville washer in one end (same diameter as the plates), keeping the pack under constant preload. When I opened one up, I was surprised this was not also done in Alfas. However -for the curious- Bacci in Italy sells these spring washers and they will fit our units. I got two and they are pretty strong! They can only be used though in 25% LSD's as there's no room in units modded for 47%. The price is very interesting also...20Euros each, ~$31, not bad!!
Jim K.

Mats · Post by **Mats** » Mon Apr 07, 2008 8:20 am

Yes, and all the BMW guys racing that I know of throws these preload washers in the bin the first time they open the diff.

They are there to minimize backlash in the diff wheels, i.e. make the ride more comfy...

Jim K · Post by **Jim K** » Mon Apr 07, 2008 10:55 am

You go east...I go west...''never the twain shall meet!'' as very wisely spoken so long ago!

Jim K.

Micke · Post by **Micke** » Mon Apr 07, 2008 12:46 pm

Having tried both ways I could make a short comment.
Earlier I always used the Mz route but Jimbo convinced me he's way is the path to light so I tried.

Conclusion. Lap times did not change. Driving didn't really improve but definitely didn't get worse either. Counting the placebo effect it might have improved just slightly. Then the mod was done one winter so count half a year between feelings.

Next I'll make a compromise and use close to zero slack

x-rad · Post by **x-rad** » Thu Apr 10, 2008 5:05 pm

could go with a Quaife.......

Jim K · Post by **Jim K** » Fri Apr 11, 2008 6:15 am

Quaife huh?? Here we go again...if you succeed in getting one for the transaxle, post here right away so that we may line up after you. I remember we sort of tried a group buy a few years ago and it went up like a lead balloon!!

Jim K.

Mats · Post by **Mats** » Fri Apr 11, 2008 6:27 am

The only problem with a quaife is that it's not adjustable in any way, once that geometry is set you need to buy another with a different angle if you want more or less...

LSD clutches correct assembly order