Which reminds me, do you (or anyone) know or can measure the degree offset if any from from the ring gear missing tooth to the hall effect rotor gap?

I'm supposed to figure this out for someone making a prototype tester to detect the CPS vs Hall sensor variance but I'm not sure I can.

 

I have not done that, but I have heard of it having been done. Does the hall sensor give two edges per camshaft rotation and the crank position sensor four edges per camsaft rotation? If so, just thinking out loud, one needs to count the time between the edges and compare them.

 

Having done some critical circuit timing work in the past using data acquisition equipment there are some pitfalls.
I will first start with I’m not an expert in this arena and I don’t understand exactly what you are attempting to do, but I will assume you are trying to see variations in timing signals in relation to one another.
The first big hurtle is understanding the slew rates of all the circuits involved which would include the crank/cam detection sensors and the associated output circuits along with any measuring devices used.

Basic premise of slew rate is all sensor/measurement electronics require a certain amount of time to detect and produce a usable signal indicating the physical condition has been detected.
The reason for understanding slew rates is because the results will be reported in degrees of rotation and degrees of rotation are typically calculated by measuring time. Therefore having an understanding of circuit timing/slew rates will give you an indication of how much inherent error, accuracy and resolution is in your measuring system.

At 928 engine speeds the timing signals are approximately in the 100Hz range which are very slow by today’s standards so just about any good DAQ or oscilloscope will be able to do this. Understanding the 928 factory timing circuit slew rates may be the difficult part so my suggestion would be to add an off the shelf optical sensor to sense off the flywheel for the measurement standard. Then compare the other two signals to this standard.
A little math compensation to account for the RPM change based on sample rate to report out in crank degrees.

And yes I did stay at Holiday Inn last night....

Brian.

 

Hall senors are very fast, right? And one could probably compare the dampener timing marks with the strobe light to see if the crank position sensor is also fast. And if the crank position sensor is slow, then EZ-K code probably has done correction for it. Just thinking out loud.

Manufacturer’s spec documents would be very helpful, without a doubt. The 928 MAF spec documents clarified to me why the stock MAF element can’t be used with high powered cars, for example.

 

I think the angle question might be better solved by the tester itself by programming it to automatically baseline the CPS/Hall relationship below ~1000 rpm.


It pains me to do busy work but I figure I should make videos of a Gates Racing and Continental belt as well as test them with a timing light since I have the covers off. Ugh.

 

Initially was a bit surprised to see the belt flutter on the drive side, but thinking back I recall seeing engines with chain driven OHC were fitted with rubbing fences on the driven side , apparently to prevent this. MB SOHC V8s come to mind , Rover P6 4cyl. Does the belt run from oil pump to LHS cam do this also?
I guess since the failure rate is very low that the engines spend little time in the rpm band that produces the flutter.
jp 83 Euro S AT 57k

 

As I recall, the belt only flutters after the 5-8 gear, IE. 5-8 > water pump > 1-4 (> stock tensioner pulley). I will look again in the next test.


More trivia - the Racing belt weighs 239g, the regular Gates weighs in at 246g, and the Conti is a plump 259g.

 

gates racing v. conti belt video comparison will be interesting

 

Still there but much less flutter with the Racing belt.


All spans flutter including oil pump to 5-8. (1-4 to tensioner pulley is damped by PKT system.)

 

Racing belt: with the engine hot, cam timing does not change between high rpm running and hand cranked to T|0 engine off.

The offset to match the leading and trailing cam tooth edges is the same (as close as I can eyeball it) as the regular Gates (+2, first post in thread).

 

Conti seems to flutter pretty much like regular Gates.


Continental belt: running cam timing does not change by rpm, hot, running engine vs. off hand cranked to T 0.

Tooth offset is less by about one degree retard advance vs. the other belts though.

 

As an aside, who makes the belts badged Porsche? Has the supplier changed over the years?
jp 83 Euro S AT 57k

 

can't remember if there is any downside to the gates racing belt?

 

The price and hard to source I believe.
Åke

 

So what’s your conclusion?

Is the conclusion that none of the belts retard valve timing with rpm, at least when new?

Is another conclusion that Gates and Continental belts are identical in terms of elasticity, while the Gates racing belt is stiffer? Does that observation square with direct measures of belt elasticity?

What belt solution is the best when using stiffer valve springs and higher rpms? The basic timing belt engineering prescribes more slack side preload, but how to accomplish that?

I think the elasticity is different from yield or permanent deformation over time, stretch in terms of the belt becoming permanent longer over time.