For the most part, the remaining work that was done in 2020 to get Gretchen on the road is pretty standard 944 stuff, like seals, belts & rollers. Since that information is everywhere on the internet, I'll be going through it as briefly as possible, so I can move on to other, more interesting topics, while bringing up a few things you may not have heard about before - icb.
Mar/Apr 2020 - It was finally time to start putting the engine back together. Spring was almost here and I was getting impatient to drive the car. My fixed budget was running low, and my approach to the belts, rollers, water pump and oil cooler seals was to evaluate and only replace what really had to be replaced. When I first started disassembly in the winter it seemed like a couple of the rollers were in great shape. But it turned out that they all had to be replaced; the bearings were either dry or rough. The water pump was fine though, the pulley turned freely but with a bit of smooth resistance, indicating that the seal was OK. I also knew it wasn't weeping or leaking anywhere, so I decided to leave it in place, since I had plans to pull the engine the following winter and replace it then. The same applied to the oil cooler seals; in fact the water pump and the oil cooler seals had been done just a few years back and the coolant had no trace of oil. So all of the belts, idler/tension rollers, air-oil separator and front engine seals, but not the pump or the oil cooler seals were replaced.
Reaching the Top
I had discovered during disassembly that it was impossible to see the TDC mark on the flywheel through the window at the top of the bellhousing, even with a light. Not only was the marking too faint to make out, but unless the intake system was off the engine, you pretty much couldn't see the window. So my process for setting the engine to TDC involves resting a 12" socket extension on top of the #1 piston (through the spark plug hole), and then rotating the engine with a ratchet and 24mm socket on the crank bolt while watching the socket extension rise up all the way to get it close to TDC, then getting under the car to do the final line-up on the flywheel via the sight hole at the bottom of the bellhousing. In the future I'll dial-indicate the #1 piston crown through the spark plug hole instead of using the socket extension/eyeball method, this is the standard most-precise way to find TDC on any engine, and as a bonus I won't have to get under the car to use the flywheel mark at all.
Black Belt
For the most part, installing the new parts was uneventful. The stud for the balance belt idler had come out when I loosened the nut, so it was re-installed with Loc-Tite 243. The auto-tensioner can be tricky to install, as the mounting studs are hard to reach and completely invisible unless you're a contortionist. When you do get the tensioner on the studs and it's time to put on the washers and nuts, it pretty well has to be done by feel only. The tensioner has to be installed at the same time as the timing belt.
As I finished installing the belts and began to set the tension, things started to get more challenging. I had watched the Edredas video on YouTube showing how to advance the pulleys by a tooth to compensate for belt slack, and this worked great; the timing marks lined up perfectly. I found it interesting when a year later (I had the belts off for something else), I noticed the timing marks didn't line up as well as they had when this belt was first installed. Belt stretch? Sure, but maybe more than that? Sounds like an interesting future blog post!
 |
| The tensioner has three very hard to access mounting studs |
|
|
||||
|
|
||||
|
|
Past, Present, and Future Tense
Belt tensioning was something I struggled with at first, as everyone seems to. There's a lot of misleading information out there concerning proper belt tensioning, but luckily there's also some accurate and helpful info, and successfully navigating through and weeding out the BS is a challenge. On these engines, there is always going to be some amount of belt whine. How much is OK and how much is too much? Different brands of belt and different tooth profiles all produce different levels of noise. When I got the car it made very little noise, but then the belts were oily (no doubt that's why they were so quiet!) and had to go. I had also changed brands, so between that and the oil I couldn't expect the same sound from the new ones. Audio clips on YouTube are not much help, because there are so many acoustical and electrical variables at work in an online recording and playback, that you just aren't going to hear an accurate reproduction of the sound. Many people out there swear by this or that method. For the first three hundred miles or so, poor Gretchen sounded like the Computer Car from the Gumball Rally, you could hear her coming from a block away. But after a while I had filtered out all the internet myths and hype about 944 belt tensioning, and came away with three helpful pieces of information:
- It's mainly the balance shaft belt that produces the noise. At least a small amount of belt whine is expected and normal. Continental brand belts make more noise than other brands due to their tooth profile.
- The right tension on the balance shaft belt will be alarmingly loose, especially on a very cold engine!
- Accurate tensioning by feel is possible: For the timing belt you should be able to just turn the water pump pulley on a cold engine by hand. The balance belt is a bit trickier; if you pull up on the upper run , it should almost be possible to make it touch the waterpump pulley, but not quite. It will be just short of "flappy".
The Porsche factory service manual provides another helpful piece of information, and that is their recommendation to tension the belts at room temperature, or 20°C/68°F. I try to stick as close to that as I can.
I had purchased the Arnworks tensioning tool, but due to inexperience with the 944S timing and balance belts, I found it to be pretty useless. It's not that it isn't a good tool, it's fine, but without having some personal frame of reference, it was useless. For one thing, the recommended tension values supplied with the tool produced ridiculously tight belts. I'm pretty sure the wider timing belt of the 16-valve engine is going to require a different tension value than the 8-valve belt does, and the included chart only covers the 8-valve engines. This tool's strength, however, is that it's very consistent and repeatable. I learned to tension the belts correctly by feel, and then I measured the tension with the tool. That gave me real numbers to use in the future if I wanted to. Now I tension initially by feel, and then fine-tune the tension with the Arnworks tool using my own numbers.
See Me, Feel Me
Learning to tension the belts by feel is not difficult, but at first it can seem like a risky thing to do. Check out the Edredas videos. I guess it boils down to how well you trust yourself to get it right! I've found that there are a couple of things you have to watch out for:
You can't check or set belt tension (no matter which tensioning technique you use) on a warm or hot engine, because the expansion of the block & head will itself tension the belt quite a bit. If you do, then once the engine cools the timing belt will be way too loose and may skip a tooth, which is much like breaking a belt - you could easily wipe out all the exhaust valves. So if the engine has been running, you must let it cool all the way down to around 68°F.
I've found that (at least with new, clean belts) the timing belt can seem to stick to the water pump pulley after running the engine, making it seem like the belt's really tight when it's not. It isn't actually sticking, you just have to move the tension and the slack to the proper areas, like it is when the engine is running and the crankshaft is rotating the belt. To do that, rotate the motor just a hair, maybe one or at most two cam pulley teeth, via a 24mm socket/ratchet on the crankshaft bolt (in the normal direction of engine rotation), then let it rest. This will take the excess tension off the side of the belt path where there should be the least tension, allowing you to break the pulley free of the belt (unless the belt tension is way too tight). Keep in mind we're only talking about a small degree of slippage; you should only be barely able to turn the pulley by hand. Much more than that and you risk the water pump slipping when it needs to be moving coolant.
What's happening here is when the engine is running and the crankshaft is driving the timing belt, the belt is mainly pulling on the camshaft - almost all the tension on the belt is between the cam & crank pulleys while the engine is running. This is due to the high rotational resistance caused by the lifters riding up the cam lobes against the valve springs. While this is going on, whatever belt slack exists is on the other side of the crank pulley, where the water pump is, and this is why the tensioner is located there, too. Now when the engine is stopped, most of that tension that was between the cam and crank pulley equalizes around the belt, and so the water pump pulley will be under more tension than it was when the engine was running. If the engine kicks back slightly before coming to a stop and tensions the water pump side of the belt, it will be impossible to turn the pulley, even with the correct belt tension, unless you relieve it slightly by rotating the crank a degree or two as above.
More on (Moron?) temperature
The "tension only at 20°C/68°F" concept is an interesting one, when you think about it. Consider a new belt, lets say it's a timing belt, installed and tensioned to spec at 20°C/68°F. For argument's sake, lets also say that you used Porsche factory tool #9201, which none of us can afford or even find. The spec for a new belt is 4.0 on the dial. This is going to result in a much higher tension, X, when the engine / head combination have reached their normal operating temperature of somewhere around 80°C. Tension X is really what the engine designers determined to be the correct belt tension for a runing M44.XX engine. They would have taken several design goals into account while arriving at this spec, not the least of which would have been keeping the belt on the engine at the maximum rpm, maintaining an acceptable bearing load, and ensuring a reasonable belt life. Now, if you go and tension the belt at some temperature other than 20°C/68°F, even using the same mythical tool #9201, then your belt isn't going to be at that "optimum" tension X at the operating temp.
Stop! Right now what you're probably thinking is the point I'm about to bring up. For almost 40 years now, many people will have been tensioning belts at temperatures other than 20°C/68°F, and if it was critical to do it only at that temp, we would have heard about it, and you can be sure it would be all over the internet, right up there along with "change belts, rollers, and seals of unknown age before you drive the car". And it's not, is it? But, while it's clearly not super-critical to be right at 20°C/68°F, just how much leeway is there, really? Surely if you "correctly" tension a belt at sub-freezing temps it will be way too tight once the engine is up to temp, and if you do that on a hot engine it would end up being way too loose. My guess, and that's all it is, is that anywhere from 65°F to 75°F is reasonable, but is there a better way? Is it even necessary? Some (most?) would say no, but I'm way too gifted with OCD for that!
An interesting thought that occurred to me is that maybe it's possible to create your own temperature-compensated tension chart, using a tool like the Arnworks 920X. The idea is to start off with a correctly tensioned belt (i.e. do it right at 20°C/68°F), and then use your tension gauge and an accurate thermometer to record a range of tension vs. temperature information over time. Do it when there are some temperature extremes, like in late Summer or early Fall. Start checking early in the morning and throughout the day as it warms up, write down the temperature and the tension of the belts as you go. Don't start the engine, just check the tension. It doesn't have to be at TDC. You should be able to get tension numbers at temperatures from the low 60's to the upper 80's, depending of course on where you live. Do this on a brand new and also a run-in belt to have a complete set of data. This should be a reliable way to make pretty accurate belt tension adjustments over a wide temperature range. Just check the temperature before you tension, look up the correct tension on your chart, and tension the belt to that number. Voila. I have no idea if anyone has done this before or not. It's likely that most people just wouldn't worry or bother about it, but I think it would be fun to do, kind of like a science experiment. And it would be interesting to see just how much static belt tension does change with temperature. If I try it I'll be sure to give detailed results in a future blog entry.
Next Time, I'll be showing you an interesting approach to on-car cylinder head thread repairs. Stay tuned!
No comments:
Post a Comment