Selecting the right gear oil

We are frequently asked which gear oil we recommend for the T3 Syncro. The official VW guideline from the ‘80s speaks of 75W90-GL4 for differential & transmission.

 But our recommendation is different:

Differential: 75W90-GL5

For the highly stressed hypoid teeth, in which the crown and pinion touch only along a small surface area (extreme surface pressure), the GL5 standard is the optimum. For this purpose, the base oil is provided with numerous additives.

 Manual transmission: 75W90-GL4+

Although the same hypoid gear pattern is also present in the manual transmission, consideration must be given here to the synchronizer rings of the gear wheels. Because GL5 lubricates so well, it can cause the synchronizer rings in the transmission to slip and partly lose their function – which can cause gears grinding and wear to occur when changing gear.
In addition, the additives of the GL5 oils can attack the non-ferrous metal component (Molybdenum coating) of the synchronizer rings.

But there are also oils with the unofficial name GL4 / GL5 (GL4 +), which manage the balancing act between the hypoid gearing’s ideal lubrication and a residual friction for synchronization.


API ratings

Gearbox oils are classified by the American Petroleum Institute (API) using GL ratings. The higher an oil’s GL-rating, the more pressure can be sustained without any metal-to-metal contact taking place between transmission components.

Designates the type of service characteristic of spiral-bevel and hypoid gears in automotive axles operated under moderate speeds and loads. Gear oils to API GL-4 are today typical representatives in transaxle transversely mounted gearboxes.

Designates the type of service characteristic of hypoids in automotive axles under high-speed and/or low-speed, high-torque conditions. Gear oils according to API GL-5 are today preferably used in differentials.



After a VC fails closed, it then fails open.

We suspected this issue for a long time, but now we found evidence to support our theory.
But let’s tell the story from the beginning.
A customer brought us his old Viscous Coupling (VC) in an unknown condition to our workshop, After the first visual inspection, I suggested to the customer that we could actually save the time on the test-rig, because fragments of the upper X-ring were already hanging out of the lubrication hole of the inner hub (see attached photo). I assured him that his VC had failed open (no drive) and would only reach about 50 nm friction moment.


But when the test-run started, we were all very surprised! The Viscous Coupling was not failed open, but closed (no slip at all)! It had sucked oil inside.

After testing all VCs from my own stock with obvious X-ring damage, I found one more VC in a similar condition.

This Visco clutch shows an advanced stage of failing open. The lowered moment indicates that some silicone oil has already leaked out and the Visco will soon completely be failed open.


Our Theory:

The constant hump pressure (>10 bar) in the oil-sucked Viscous Coupling causes the rubber of the X-ring to creep and slowly be pushed into the gap. As a result, the X-ring can no longer create a seal.

To cut a long story short: After a VC fails closed, it then fails open.

Testing of a “baking oven VC” #2

After having tested a so called baking oven VC months ago, I decided to test another one due to verification resasons.

I got the baking oven VC from a german guy who cooked it at 120°C with a 60K silicon fluid.

60K Visco
60K “baking oven VC”

60K VC Testing
first testrun

After the first minutes of the testrun I was absolutely sure, that the VC is missing a lot of fluid, because the chart was much to low for a high 60K viscosity. During the testrun, the air in the VC is mixing with the silion oil as tinny bubbles – which looks like super sparkling mineral water. These tinny bubbles in the fluid reduce the shear force of the silcone oil:

more air = more bubbles = lower effektive viscosity

At a temperature of 135°C the teststand stops automatically for safety resasons. After a short cooling break, we did another testrun at a higher temperature of 155°C – that’s the maximum temperature that should not do any harm to the X-ring sealings. But there was not any signs for a Hump.

Testlauf #1 (135°C)
Testrun#1 (135°C)

Teslauf #2 (155°c)
Testrun #2 (155°c)

I refilled the baking oven VC with 12,5K silicon oil to find out the amount of missing fluid. All in all 20,3g (!) fluid was necessary to get the VC in the correct range of tolerance. Two gram silicon oil increases the hump-temperature in 8°C. This means, that the baking oven VC would have had a theoretical hump-temperature of 200°C. But that’s just theortical, because practical the X-rings would have been damaged at these high temperatures.

20,3g Silikonöl nachgefüllt
+20,3g Silikonöl

Testlauf #3
Testlauf #3


It’s not always that simple, even if it seems so. Some thoughts why the baking temperature can’t be identical with the hump temperature can be found in the last section of the first baking oven test we did months ago.