I still do. Get a diff at a discount? I can sometimes "test" them for long periods of time, but as far as getting someone else a discount, no, not likely. We don't sell them direct to end users, so there's no means of arranging a purchase to an individual.

As far as comparing a torque-sensing helical gear diff to a speed-sensing viscous LSD, well, they're two different animals with different behaviors. If you compared differential rate to torque transfer, a torque-sensing diff will have the ability to transfer torque side to side with zero diff rate, while the two wheels are rolling at the same speed, up to the point where the ratio of torque on the wheels exceeds the diff's torque bias ratio (which can be anywhere from 2:1 to 5:1, depending on design and model). At the point where that ratio is exceeded, a differential rate will begin (wheel spin), but the diff will continue to distribute torque at it's bias ratio. So, the ability to transfer torque without a diff rate means that the majority of wheel spin is prevented before even starting. However, it does not have the ability to lock up, so if a tire does become completely unloaded, it will spin.

A speed-sensing diff has sort of the inverse behavior. The curve of the slope of diff rate vs. torque transfer varies a bit with tuning, but it cannot transfer any torque until a differential rate has begun. Wheel spin must occur for the VCU to have something to correct. It has to be tuned to allow the normal differential rates you'd get while turning and cornering in daily driving without binding. They also have to be tuned to provide minimal resistance to differentiation during ABS events so that the ABS computer recognizes that a brake has locked up. Consequently, most street-going production VLSD's are pretty weak. But the design does have to potential to be fully locking if tuned appropriately. It's just that it's tough to do in practice because of other compromises. That's one reason why mechanical speed sensing devices such as viscous couplings have been almost entirely displaced by electronically controlled devices, like the Aerostar's center lock up clutch.

Basically, torque-sensing works before wheel spin to prevent it from happening and it's static biasing ability (the no spin required to distribute torque) gives it good dynamic benefits as well, helping to neutralize under and oversteer as required. The drawback is that it cannot lock up completely. A speed-sensing device can only react to wheel spin that's already occuring to try to correct it, but on the other hand, it can lock fully if tuned to. However, it can only have an significant effect on vehicle dynamics if its electronically governed, as mechanical devices tend to be too slow and have their own set of compromises.

As with anything, LSD's are design and requirement compromises. Which set of compromises works best really depends on the specific application.

 

I'd like to add my 2 cents here on viscous coupling units (VCU). I owned a GMC Syclone (AWD S-10 pick-up) for 10 years (132K miles). It had an AWD system very similar to the Aerostar E4WD system (this system was also used on the GM Astro van). The difference was that it used a VCU instead of an electronically controlled clutch to "lock" the center differential. All I can say is that the VCU got in the way 364 days a year, and did its job correctly one day per year (a very bad “compromise” ratio). It would always tighten up while turning in cul-de-sacs, and the worst thing was on snow covered roads. It would cause driveline wind-up while turning corners and cause all the wheels to lose their lateral stability. The vehicle would go into a slight four wheel drift on every corner! The Aerostar’s completely open center differential doesn’t cause ANY wind-up and all the tires can work to hold the vehicle stable and turn a corner. I’ve had a lot of experience with both systems and the Aerostar’s system is in a completely different league (I suppose I‘m “biased“ though, being part of the development team at Dana).

Getting back to LSD’s. I agree that the Ford unit is very poor. A Torsen or an Auburn would be much better.

 

OTOH, the VLSD in the V8 Explorer t-case (at least the earlier ones) actually worked very well. B-W probably spent a bit more time getting it right due to the relative high production volume. It just goes to illustrate my point that a VCU can work very well or be a POS, depending on how well it was set up. Was the case in the Syclone a NVG unit?

BTW, if you're at Dana, can you tell us anything about the D28 front axle in the Aerostar? It's obviously different from the common D28 twin-traction beam axle. Do r&p gears for a TTB 28 fit it? Or, would it be tough to retrofit an SLA Dana 35 from an Explorer/Ranger in its place? They appear very similar in size and layout, but the D35 is a little longer, IIRC...

 

It was the Borg-Warner unit. Probably the same part number and tuning as was used in the Astro (high volume). I'm sure it worked better in the van. The Syclone was basically put together using a bunch of off-the-shelf parts at GM.

I left Dana in '96 and don't know much about the SLA Dana 35. I can't help much on the D28 either. I was in the transfer case group. I know the Aerostar D28 is a spiral bevel and I think the TTB 28 are Hypoid. If I run into any of the guys I'll be sure to ask.

 

Thanks!! I've been looking for a set of these for a while!