Cam timing and chain tension

[CharlesJones]

Hi,

I've been searching through the forum to try to find the answer to this question. I see the question a lot but with no difinitive answer, so any advice any of you could give would be greatly appreciated.

My question is how tight do I need to hold the timing chains to be able to accurately set the cam timing on my '84 Carrera? I'm using the C-shaped vice grip on the left side and the neat little chain tensioner tool for the right side. With these tools I find I can put a lot of tension on the chains, and I'm worried that I might be put too much tension on them. Should I back off the tension a little, or should I set the timing with a lot of tension on them? At the moment I could put more tension on the chains but they're already tight enough so you can't flex them with your finger. It seems changing the chain tension slightly makes a big difference to the cam timing.

Setting the cam timing is pretty scary stuff - I'm terrified I'm going to stuff it up and destroy all my valves on the first turn of the engine. What a nightmare that would be! So thanks for any help and advice you can give.

[ChrisBennet]

Here's a couple of observations:

The tension on the chains changes (retard/advances) the cam timing.

The tighter the tension the more consistent the readings will be - too a point.

It's more important to get both cams timed the same than to the actual number. When the motor is running the chain will have varied tension on it and the cam timing will change depending on whether the motor is trying to accellerate the cam or slow it down.


I've given this some thought and come to the conclusion that the chain tension only has to be tight enough to get consistent readings. I make them tight enough to get consistent reading but no tighter. My reasoning is that way I'll have about the same tension on each side.

(The following numbers are made up but probably in the ballpark if you using a 3.2 cam.)

For example, if your tension wasn't tight enough you might get measurements with a range of as much as .06mm in the case of your 3.2 cams (1.21, 1.25, 1.24, 1.23, 1.27) => 1.24 +/- .03

Increase the tension and you'll see the measurements get more consistent say within a .02mm range (1.26, 1.26, 1.25, 1.27, 1.26) => 1.26 +/- .01

With a new chains and sprockets, tightening it some might get you measurements like this: 1.27, 1.28, 1.27, 1.28, 1.28, 1.28 => ~1.275

Notice how the cam timing advanced whenever the tension increased? By my reasoning, once I tighten the chains past the point where I'm getting consistent numbers, I'm changing the cam timing without any feed back i.e. I can't tell how far past "tight enough" I've gone.

Of course my reasoning may be specious. Like I tell my friends, thinkings not really my strong suit.

I use the little tool to tension the side that I'm currently setting. For the other side I just do something to keep the chain from getting too slack. (A collapsed hydraulic tensioner with the pin in it, a clamp or a little tool I have.)

Tip: I always put the fan housing on for cam timing. I find it impossible to judge the pulley position accurately just using the case seam.

-Chris

[Hald]

speaking of chain tension during cam timing ... with the benefit of having used Chris' tensioning tool when we checked the timing before disassembly, I improvised a similar tool which I could use on both sides, simultaneously, for very little cost. Seen in the following picture, they're made of small turnbuckles & they worked pretty well!



This is the best picture I have available. If anyone is interested in details, let me know.

[philippeF]

I am...

[Hald]

For those interested, here is a picture of the improvised tensioning tool.



It's made from a small aluminum turn-buckle found at a local hardware store. The turn-buckle is the type often used to straighten a warped screen door or to tighten a cable. It has both right and left-hand threads so that, when turned, the overall length increases or decreases. In this application, it increases in length to tighten the chain.

Since it has both left and right-hand threads, and it's easier to find right-hand threaded bolts & screws, I kept the left-hand threaded eye bolt and opened the eye sufficiently so that it just fit over the bolt that normally holds the cam chain tensioner. For the part that pushes against the idler, I found a wing-bolt which threaded into the turn-buckle body, cut it down a bit so that it would screw all the way in, and ground a notch in the wing which fits against the bottom of the idler where the normal cam chain tensioner makes contact.

I don't know the exact cost, but I don't think it cost me more than $5 for the parts and maybe 20 minutes to fabricate the first one. 5 minutes or less for the second one.

I hope this is of use to some of you.

[CharlesJones]

Thanks for the advice guys.

I'm still a little confused though - its so hard to quantify how much tension is being applied to the chains using the vice grips and the tensioner tool. I think what I'll do tonight is try it again but back off the tension a little, and try to see how consistant my readings are as per Chris's suggestion.

But Chris, am I correct in thinking that by consistant readings you're talking about taking a reading with each 720 degree turn of the crank? ie. set the timing so it reads 1.25 at 360 degrees from Z1, and then take readings at 720 degree turns after this? And also, if I did have too much tension on the chains when I timed the cams do you reckon this setting would be too far off for when the chains were under normal tension when the engine is running?

Thanks again for your help.

[ChrisBennet]

Originally posted by CharlesJones
Thanks for the advice guys.

I'm still a little confused though - its so hard to quantify how much tension is being applied to the chains using the vice grips and the tensioner tool. I think what I'll do tonight is try it again but back off the tension a little, and try to see how consistant my readings are as per Chris's suggestion.

That why I use the little tool to tension the side the I am measuring. I feel I can control the pressure easier than using vice grips. Mark's ("Hald") tool works great BTW.

But Chris, am I correct in thinking that by consistant readings you're talking about taking a reading with each 720 degree turn of the crank? ie. set the timing so it reads 1.25 at 360 degrees from Z1, and then take readings at 720 degree turns after this?

Yes, the multiple measurements are obtained by rotatating the crank 2 rotations (720 degrees) and measuring again. See below for the reason for taking multiple measurements.

And also, if I did have too much tension on the chains when I timed the cams do you reckon this setting would be too far off for when the chains were under normal tension when the engine is running?

Thanks again for your help.

If you had the same amount of "too much tension" on both chains you would be OK. The trouble is, you can't tell how much tension you have. If your measurement are in the middle of the range you aren't going to hurt anything.

If you think about how the chains are tensioned when the engine is running it might help. The chain tensioner+idler acts like the tensioner on the rear of 10speed bike - it has nothing to do with the actual tension on the chain. The tension on the bike chain is more or less a function of how hard you are pushing down on the pedals.

Consistency is the key to good measurements. Re-measuring multiple times lets you see if your measurements are consistent and thus reliable. It will also allow you to see how much "noise" or unavoidable error there is in your measurements.

Once you know how much "noise" there is in your measurements you can determine just how close you can time your cams side to side. For example: Suppose that you find that your cam measurements fall within a range of .02mm for a single cam (1.25-1.26mm). Given that range, you shouldn't expect to get both cams measurements to match any closer than .02mm to each other. The best you can hope for is that with multiple measurements you get an average reading that is less than .02mm different side to side.

Think of it this way, if you have 2 clocks that are only accurate to +/- one minute a day you shouldn't expect to the get their times to match each other to the second.

Having said all this, I kinda doubt you would notice the difference between timing the cams "dead on" and having them a little off. I worry about this stuff 'cause that's the way I am. Get them close, triple check, move on, and enjoy driving your car.

-Chris

[CharlesJones]

Thanks for your advice Chris - this has been very helpful to me indeed and I really appreciate it.

I've just been spent a couple more hours retiming and retesting my cams, and after taking multiple readings like you suggested I'm getting an average reading of 1.225 +/- .025 for the left cam, and 1.23 +/- .02 for the right cam. This is well within the range specified in the book (1.1 to 1.4) and is pretty similar left and right, so I guess it should be okay. I'll go and check it one more time tonight and then whack the chain tensioners on.

It was just *such* an effort torquing the cams without letting the dial gauge move - I felt like I'd had a good work out in the gym after trying that a few times! Now I'm looking forward to getting the rest of the rockers in and the sealing the lot up.

Thanks again for your help.

[ChrisBennet]

Your welcome! Next time don't worry about letting the dial guage move while torquing. Just torque the bolt and turn the crank 2 turns and remeasure.
-Chris

[viejopatron]

This is good stuff. The turnbuckle tensioners are neat in that they allow control of the tension and Chris is a great source.

I am not an expert on these Germanic Thingys, however I think that the thing to keep in mind with these pesky cam chain drives is that they drive in one direction only. Tension should remove looseness in the chain on the non-pulled run, but not so tight that the chain cannot be moved several millimeters up/down relative to chain travel. Too much tension adds unecessary load to the drive and shortens it's life. Therefore if you must (or just can't help yourself) reverse direction, then reverse well past your "STOP" point, stop and reverse again in the normal drive direction. Doing so will seat the chain and sprockets in their respective drive features and remove slack that normally is not present. This is especially true with used sprockets as the tooth profiles become asymetric with use.

How do I know this? I'll only say '27 Indian Scout, '69 Honda 160, '62 Norton Atlas, '75 Norton Commando and '90 Medici w/ Campy Gruppo.

[hcoles]

The turnbuckle thing works great. I picked up a few different sizes at OSH a few weeks ago and used them for my cam timing. As the others have said I won't worry too much about the chain tension. As you turn the crank the chains get tensioned pretty good anyway. Tension the chains until they are tight (if I can't move the chain with my fingers very easy...I call that tight enough) and always rotate the engine in the normal direction and slowly until TDC and check the reading. In any case...the cam has 17 holes and the gear has 16 holes (or vs. versa) so the amount you can move the adjustment isn't that small anyway.. maybe 1.4 degree of cam rotation...someone can check my math. More or less 8 holes moves you one half of a hole pitch. 360/16/2/8 = 1.4 I'm guessing there are other things bigger to worry about..checking the cam timing was probably the most fun I've had during the rebuild.
That's my 2.5 cents worth on that....
-h

[stevepaa]

Well I am having a little difficulty doing my cam timing on my 964. I have used wooden dowels in place of my tensioners and then just use the tops to apply pressure on the chain. Works well.

So when I get to 1.26 mm, pull the pin, move the crank Z1 to top and the reinstall the pin, I get 1.16 on each succeeding 720 turn.

If I tried to correct by moving pin one hole up, then I get ~1.50 consistently.

It seems to me that there is "give" in the combination of cam wheel, cam sprocket and pin that just won't allow me to get to 1.26 and stay there.

Should I tighten the cam bolt to full torque when I first reset pin at 1.26 position?

Any comments appreciated.

[ChrisBennet]

Originally posted by stevepaa
Well I am having a little difficulty doing my cam timing on my 964. I have used wooden dowels in place of my tensioners and then just use the tops to apply pressure on the chain. Works well.

So when I get to 1.26 mm, pull the pin, move the crank Z1 to top and the reinstall the pin, I get 1.16 on each succeeding 720 turn.

If I tried to correct by moving pin one hole up, then I get ~1.50 consistently.

It seems to me that there is "give" in the combination of cam wheel, cam sprocket and pin that just won't allow me to get to 1.26 and stay there.

Should I tighten the cam bolt to full torque when I first reset pin at 1.26 position?

Any comments appreciated.

There is nothing wrong with getting 1.16. The range is 1.16-1.36.
It's more important to get them even side to side.

You can fine tune the the cam timing by using the sprocket holding tool (9191) to twist the sprocket one way or the other before snugging it up it with an impact wrench (and then torquing). The sprocket holes can be treated as short slots that allow some adjustment.

When you say that tensioning it with a dowel works well are you saying that you get consistent readings? I would think the numbers would be all over the place that way.

-Chris

[stevepaa]

I am using the dowels as a stiff plunger in place of the chain tensioners. I can tighten down on the nuts holding the chain tensioner top and get a tight chain. Once I got 1.16 I could get it repeatably every 720 degrees.

[hcoles]

I got the same DI readings for turnbuckle tight compared to as run pressurized tensioner.

[davidppp]

One of the less desirable characteristics of chaindrives is harmonic oscillation.

The guides and tensioers are designed to minimise this..but I bet it is still very significant..

The variation induced will be much larger than the sort of error inherent in the setting method..

And, FWIW, it can be uselful to deliberately retard the cams..

Racing 911 motors seem among the most buzz-proof of motors with inclined valves..

Just ask a Lotus runner..

Kind regards
David

[AlfonsoR]

Subscribing

[boyt911sc]

I find these mechanical tensioners very helpful in getting consistent readings during cam timing set-ups. The key is to get as many readings that are consistent in values and reproducible. While it could be done using other tools like vise grips and C-clamps, using the Stomski tool or similar ones make the work very convenient.



Tony

[Walt Fricke]

Those turnbuckles are a great idea. I like my store boughten adjuster, and I have a set of solid tensioners I can use as well (in the traveling parts box in case of disasters at the track), but such a simple gadget!

There is, indeed, a fair amount of slop in the vernier pin in hole system. On the early cams, with the protruding end you can grab with the cam holder, you can use the slop to your advantage - when you are close to the number you are shooting for, put the pin in. If it won't go in, tweak a hair with the holder. Then use the holder to turn the cam a little this or that way depending, but not so far that you see the crank pulley move. See if that gives you what you want. If not, reposition the pin and try again. Eventually you get a number you can live with (easy if you accept the factory tolerance, harder if trying to be anal). At which point you hold the cam and tighten the nut with the crow's foot. Not super tight - maybe half torque. Then turn the crank through a couple of measuring cycles. If it is still to your satisfaction, you can get out the big torque wrench, and the cheater pipe to keep the holder from moving (I have a pipe which will reach the floor while about 80% overlapping the holder's rod - less than that risks bending the rod), and crank in the torque.

I found the newer, bolt end, style cams much harder to dial in - trying to pry the cam via the inner teeth through the holes. I'd hate to do that with all the rockers installed.

[AlfonsoR]

How helpful are adjustable cam sprockets?
JBRacing Image

I mean i know they should help, but it seems they're not really necessary. But still i wonder.... would the folks at JB Racing have developed this for not much of a good reason? Has anyone ever tried a set?