3.2 valve seal confusion

[Mark Henry]

I'm doing the top end and studs on two 3.2 engines, one is an '87 (64Hxxxxx) and the other is an '88 (64Jxxxxx). One engine (IIRC the '87) had valve seals only on the top guides and the other only had seals on the bottom guides.
The new kit came with 12 valve seals.

I'm trying to figure out which is correct?
Thanks

[john walker's workshop]

Use all 12. Racers like to leave some off. I like white teflon. Easy to install, unlike the brown ones that can distort and possibly not center perfectly. YMMV.

[Henry Schmidt]

Two streams of thought about Teflon vs Viton.
Old school Teflon seems to allow more oil to pass. This extra lubrication is thought to reduce valve to guide friction extending the life of both guide and stem. Down side is slightly more smoke [on trailing throttle decel] and oil consumption.
With the more modern Viton seal we tend to see a better oil seal. This reduces oil flow to the valve stem and in turn creates more friction and slightly more wear. Less oil consumption and smoke but shorter valve guide life.
After nearly 40 years of building air-cooled 911 engines, it's hard to teach an old dog new tricks. Teflon is our go to for most rebuilds.
As for customer valve jobs where we are not installing the heads, we defer to the customer's wishes.

[PeteKz]

Just a data point: On my most recent build I decided to try the "racer's trick" and opted to leave the exhaust valve seals off completely, but used the brown seals on the intake. Rationale is that the exhaust valves get a lot hotter and would benefit from increased oil for heat transfer to the heads, and lubrication. Downside is potentially increased oil consumption. However, the pressure pulses of the exhaust and gravity tend to push excess oil back down the valve stem into the head instead of sucking oil into the port, as would be the case on the intake valves.

After 5000 miles, the oil consumption is pretty minimal. I don't know exactly how much, because I was losing oil from leaks around the rocker shafts that I recently fixed with the RS O-rings, so I haven't had enough time to accurately measure oil consumption. But, after a couple tanks of gas, I'm pretty sure that I am using less than a quart per 1000 miles. I think that's an acceptable trade off.

I'll try to update occasionally with my experience, but even so, my experience is very much more limited than the pro's here like Henry, Neil, Ivan, et al. So if they say it will burn another quart per 1000 miles, I'd believe them before my data point of one.

[Walt Fricke]

I'm with Pete, at least for a race/track only engine. I never could figure out how much, if any, oil could travel up the exhaust valve stem against the pressure pushing outward (back to the cam carrier) of the exhaust on the valve guide. Maybe a little bit of capillary action when the car is parked on a side sloped angle? For sure not anything one would worry about. As to oil consumption on a track car, well I've never tried to measure that, as it gets changed based on use and not consumption.

Hard to imagine it matters one way or the other on a street car.

[Henry Schmidt]

Originally Posted by Walt Fricke

I'm with Pete, at least for a race/track only engine. I never could figure out how much, if any, oil could travel up the exhaust valve stem against the pressure pushing outward (back to the cam carrier) of the exhaust on the valve guide. Maybe a little bit of capillary action when the car is parked on a side sloped angle? For sure not anything one would worry about. As to oil consumption on a track car, well I've never tried to measure that, as it gets changed based on use and not consumption.

Hard to imagine it matters one way or the other on a street car.

Oil is sucked through the exhaust guide after the exhaust valve is closed if your exhaust system is working properly. Tuned exhaust/headers [most 911 have them] create a low pressure area behind the valve to reduce back pressure and increase combustion gas exchange.

[Neil Harvey]

Originally Posted by PeteKz

Just a data point: On my most recent build I decided to try the "racer's trick" and opted to leave the exhaust valve seals off completely, but used the brown seals on the intake. Rationale is that the exhaust valves get a lot hotter and would benefit from increased oil for heat transfer to the heads, and lubrication. Downside is potentially increased oil consumption. However, the pressure pulses of the exhaust and gravity tend to push excess oil back down the valve stem into the head instead of sucking oil into the port, as would be the case on the intake valves.

After 5000 miles, the oil consumption is pretty minimal. I don't know exactly how much, because I was losing oil from leaks around the rocker shafts that I recently fixed with the RS O-rings, so I haven't had enough time to accurately measure oil consumption. But, after a couple tanks of gas, I'm pretty sure that I am using less than a quart per 1000 miles. I think that's an acceptable trade off.

I'll try to update occasionally with my experience, but even so, my experience is very much more limited than the pro's here like Henry, Neil, Ivan, et al. So if they say it will burn another quart per 1000 miles, I'd believe them before my data point of one.

I have never checked or even given it any thought. Growing up in this business I always built engines as directed by those that were more experienced. If the build spec asked for seals that were installed. I remember building the Porsche 962C engines and they did not use seals.

Today, guide material is a lot better and tighter clearances are the norm. Valve stem accuracy is to the micron so maybe what was done years ago is unnecessary now.

I have always "wondered" about the way Porsche suggest 911 valve guide clearances should be checked. With the allowable clearances they suggest, most modern engines would never be able to pull much crank case vacuum.

The only reason we leave guide seals off today is due to valve lift interference issues. Once the valve lift goes above 0.500" with the 911 rocker geometry, seal clearance becomes an issue.