Mark944na86

So: What is the correct tension for a used timing belt? Should it more or less than the tension on a balance belt?

Since we're on the topic of timing belts, I thought we should try to get this sorted out once and for all.

Here's some discussion in the "S2 timing belt job" thread. It discusses the inconsistencies between the values specified on Bruce Arnn's "arnnworx" site, and the "clarks-garage" site:

Bruce (F18rep) replies:

I reply:

So what's the answer?

Yummybud924

not sure but a balance belt is definately much looser feeling than a timing belt.

also the kricket is totally useless for teh balance belt. the kricket won't read anything below 30 pounds.


also for the timing belt I'm a bit confused about how to tension it with the kricket. from what I remember from last time you move it backwards 1.5 teeth so the long span gets loose and then you set it to 40 pounds? if so when you turn it back 1.5 teeth the 40 pounds becomes 50 pounds.

anyways the whole tensioning thing drives me crazy so I tension it to what I think feels right which is probably too tight and then take it to my mechanice shop and pay the 100 bucks to have it checked.

Mark944na86

Well, if you read the specified values on Clark's table you will see that this should not be so.

OTOH, if you believe Arnn's values, then this should be so.

And that is exactly the discrepancy I'm trying to get resolved.

Yummybud924

I had the tension of my belts adjusted by a porsche shop and I'm saying feeling them by hand they balance shaft belt (after teh 1500 km retension) feels much looser than the t-belt.

but from what I've heard the balance shaft belt also feels looser because of it's design (teeth on both sides) but still it's noticably looser.

I will check my timing belt tension and balance shaft tension (if possible) with my kricket this weekend and post the values I get. they should be at the correct tension for used belts.

the kricket is pretty much useless for the balance belt anyways so I was never able to set it to 27lbs. the krickets lowest reading is 30 pounds and anything udner that it won't read or it will give you a really inaccurate reading.

Mark944na86

Just out of interest, how long ago did the retension get done? This is the same car that you are about to replace the T-belt on because it looks worn and steel is showing through?

Do you think the belt appeared this way when the shop did the retension? A bit of a worry if so... you'd think they should have picked that up and mentioned it...

Yummybud924

the nylond or fabric "facing" that covers the belt teeth so its not just bare rubber has worn off on a couple of teeth on the belt.

the gates belt that I got was like this when I bought it and has gotten worse now. so a couple of the teeth on the belt are not fully covered by that nylon fabric and are just bare rubber teeth.

I assume that fabric layer on the teeth also provides strength to the teeth so they don't rip off.

I've gotten timing belts like this before where there are imperfections and all the teeth are not covered by that fabric layer and some of the teeth are bare rubber or almost bare rubber.

the new belt I have is also gates but this one looks much better.

I don't blame it on the mechanic who just tensioned the belt. I don't think any mechanic will look at every tooth on a belt anyways before installing it but I do, if all the teeth are not covered by that fabric I'll return it next time and get another belt.

some timing belt failures are probably due to faulty belts.

FRporscheman

I have found some typos and inconsistencies on clark's garage. We can hardly complain though - his site is a treasured gift to all Rennkind.

The timing belt should feel taught, while the BS belt should be surprisingly loose. The hydraulic TB tensioner on the 968 will set the TB tension to something in the neighborhood of "40" on the kricket. Setting a BS belt any tighter than, say, "30" on the kricket will cause quick failure of the belt. Don't ask me how I know that.

Mark944na86

Not that simple, I'm afraid. The values on Clark's site correspond to values published in my .pdf version of the Porsche fsm -- page 15-4 for t-belt values, and page 13-9 for the b-belt values.

Now, the .pdf version of the fsm I have is dated 1985 -- this would correspond to the early b-belt (15mm) as opposed the later "new-style" b-belt (18mm).

So Clark's values are not simply typos -- they correspond to at least the 1985 version of the fsm manual that says the used t-belt values (after 3000km) should be 2.7 +- 0.3, which is the _same_ tension value as for the b-belt (new and used) at 2.7 +- 0.3. Further, he cites a value for the "new-style" (18mm) b-belt at higher value of 3.5 +- 0.5 than for the 15mm belt -- I assume he didn't pull this value out of thin air, either.

The question is, have these values been revised by Porsche since Clark posted these values -- if so, when, and what is the reference?

Or have people been routinely running their t-belts at a much higher tension after 3000kms than the Porsche recommended values for some reason?

So: Anyone actually have an up-to-date authoritative reference for these values?

tod84944

One thing not to forget...... the quailty of timing belts has improved a whole lot since the shop manual were written. I have done many timing belt swaps over the past 10yrs. On 87 and later cars with the spring loaded tensioner, it can get it pretty close to the factory recommened setting. I usually pull up on the tensioner after I loosen both bolts to let it extend, to make sure it is fully extended. I have done this many times, and when checked with the Porsche tool, it is dead up or soo close it does not make sense to re do it.
The newer cars, foreing or domestic, that have just as much or a lot more tech than a 944, do not use a special tool to set tenison, many just have a spring loaded tensioner like the later 944 cars. On the newer cars, the timing belt change on many of them is around 60K miles foreign or domestic.
A very experienced Porsche trained mechanic told me that the newer belts are good enough that you can do the thumb and index finger twist test on the longest run of the belt. Turn the belt with your thumb and index finger and if you can twist it more than a quarter turn, it is too loose. If you can not get it close to a quarter twist, and it too tight.

tod84944

Back to the original question, the balance shaft belt does have to be quite a bit looser than the timing belt. If it is too tight, it will wine really bad. Also the balance shafts bearings are pretty loose, so if you get it too tight, it will pull on the ends of the balance shafts and cause bearing wear on the balance shafts. The main thing, make sure it is tight enough that when the car is running, it is not flopping around so much that it is rubbing on the nose of the water pump.

austin944

With respect to the differing used balance belt tensions of 2.7 and 3.5, please see this post:
https://rennlist.com/forums/638498-post16.html

I think the balance belt feels looser by touch because it has a longer unsupported span than the timing belt, so it can deflect more.

It sounds like you're assuming a linear relationship between the values of the various tools for measuring the timing belt tension. I'm not sure that's a legitimate assumption; there could be a non-linear relationship. The tools aren't constructed the same, so they're not necessarily going to measure the same force at the same deflection points, even assuming they deflect the belt in the same manner.

AFAIK, the Porsche 9201 tool measures the tension in dimensionless units; I've never seen anyone explain what they mean.

joonas

About krikit. If the engine is out it is really easy to use
http://www.blaszakprecision.com/Belts.html

Like austin944 said, Porsche tool is in scale units. Tool is calibrated with testing gage to 4.0 scale units.
Arnnworx uses torque wrench so it has to mention the lbs values. I think he got them by comparing factory tool to his tool.

Mark944na86

I'm not assuming linearity, but I am assuming _monotonicity_. That is, a higher reading of 3.5 on the 9201 corresponds to a higher tension than a lower reading of 2.7 on the 9201 (whatever those values actually correspond to in standard units, e.g. lb force).

That's the issue: Regardless of units of measurement, Clark's site (and at least some versions of the Porsche FSM) say the used T-Belt tension should be _same_ as the B-Belt tension (2.7 units as measured on the 9201 tool.)

That's for the "old-style" 15mm B-Belt. For the "new-style" 18mm B-Belt (which is what I have in my S2), Clark cites the value as being 3.5 units.

That is, the reading on the 9201 should show a _higher_ value for the B-Belt (3.5) than for the T-Belt (2.7).

Now, assuming linearity is one thing, but unless you are saying that the 9201 reading of 3.5 corresponds to a _lower_ tension than a reading of 2.7 on the same device, I don't think it is plausible to say that the T-Belt is supposed to be set at a higher tension than the 18mm B-Belt.

Unless of course these FSM values have been updated at some stage by Porsche (anyone have a FSM complete with all the bulletins?)

The alternative would seem to be that, for whatever reason, many people have been routinely ignoring the Porsche FSM recommendations and resetting the T-Belt after 3000kms or so to a value much closer to the "new" belt settings than to the recommended used belt settings. That may or may not be a bad thing, but I'd like to get sorted out in my own mind what the actual most recent Porsche recommendations are.

In the FSM (page 15-4), Porsche state that the acceptable range of tension values for the T-Belt during the first 3000kms is 2.4 - 4.3 units on the 9201.

That's a very large range. Maybe some of the mythology is that the belts have to be set within a very narrow tolerance?

Mark944na86

Interesting; it supports the idea that

a) people are routinely (mis)applying the guidelines for the 15mm belt to the 18mm belt, presumably out of habit (and contrary to Porsches own TSB guidelines), and

b) updated info in TSBs doesn't necessarily displace the group consensus of what constitutes "best practice".

It also confirms that indeed Clark didn't get the 3.5 value for the "new style" 18mm belts out of "thin air"; on the contrary, it indicates he's actually uptodate regarding the info in the Tech Bulletins -- perhaps moreso than other sources.

Landseer

Don't let me take you too far off track, but lets inject a tangential idea.

Better practice would have been a better fricken design from the factory, IMO.

If I do the 944 TB again, I'm going to make a cut in the back cover, just below the water hose, to allow utilization of the Kempf tool instead. (I'll find a picture for ref). The idea is to twist the belt 90 degrees and determine if the edge of the top plate is within the notch/window. Kempf is apparently a real-world adaptation of the $650 over-engineered factory tool with same type of mechanism. It was essentially validated by 928 guys against the factory tool before they accepted it. Was before my time, just recounting the lore.

Balance shaft belt needs to be much looser or it stretches too much and whines. How loose? Not sure. Glad you are delving into it. I just did it by hand so it twisted >180. But it that corresponds to a reasonable tension, maybe that's the way to set it.