91 Civic, 197k miles
I replaced the timing belt and tensioner the other day. I
cookbooked my way through setting the tension, not knowing
how things worked, and am pretty sure the rat-a-tat-tat I
heard from the engine is the belt, somewhat loose, hitting
the upper timing belt cover, for one.I just re-did the
tensioning step, following especially Tegger's careful
(post of Oct. 4, 2005), and the engine still makes the
rat-a-tat-tat noise. I can feel the vibrations in synch with
the rat-a-tat-tat on the when I put my hand on the upper
timing belt cover.
I watched for the TB's forward side (closest to the car's
front) going tense, while the aft side went somewhat slack.
In one attempt I did hear a little zing from the tensioner
spring. I checked and re-checked this as I tightened the
tensioner adjusting bolt. On my last attempt, I do not know
if the "zing" and so tightening of the spring occurred or
I checked the timing with a timing light, and there are no
indications a tooth has jumped. I will continue to be very
aware that this could happen, especially when the tensioner
is loose or not set right.
On my third attempt, I want to
(1) loosen the tensioner screw more this time.
(2) make sure I hear the "zing" from the tensioner spring
(3) do as Jim Beam said in the above thread: (a) Turn the
engine over three times; (b) watch the tension in the belt;
and (c) watch that TDC on Cyl #1 aligns with the camshaft
sprocket being in the correct "up" orientation. I will take
the spark plugs out to make life easier this time.
Is it true for step (3) that the TB aft side should always
be pretty slack compared to the TB fwd side?
Any other clues on how to get this right?
Struggling here. Prompt responses are appreciated.
(Posted and mailed)
I'm not sure where the "turn the engine over three times" came from.
Doesn't sound right.
When the belt was off, the tensioner should have been pushed in as far
as it would go against its spring, then bolted so it would stay there.
Once the belt is on again, you
1) rotate the engine counterclockwise so as to put all the slack on the
tensioner side, which should only be one or two teeth. 2) with a quick
motion, loosen the tensioner bolt, which at that point should ZING out
to take up the slack. 3) tighten the tensioner bolt BEFORE releasing the
The important things are that the tensioner has to be pushed ALL the way
back, so it can take a good run at the belt when it's released, and that
the slack has to be entirely on the tensioner side of the belt.
If the belt seems a bit loose even after that, you can tap the bolt head
with your ratchet (while making sure the slack is in the right place).
This will shock the tensioner into pushing out just a bit more.
If you don't want to remove the covers again, you can loosen the
tensioner, turn the engine CLOCKwise THREE TEETH, then snug the
tensioner up again. This pulls the tensioner in a bit so you can try
retensioning the belt. The tensioner can't get a good run at the belt in
this case, so you may have to help it with taps from the ratchet.
Good luck and hope this helps.
Jim noted that this was a good idea after the final
tightening of the tensioner took place.
Do you mean so that the spring is applying as little force
as possible to the main body of the tensioner?
I was loosening the bolt prior to rotating the three teeth,
watching so that a tooth did not skip. I will try the above
instead, listening for the ZING.
Is this the same as you describe at [***] above?
I will try this first.
I bear in mind that the engine has to be cold during this
Unfortunately I find it too much of a battle to torque the
tensioner bolt by going in with a socket from the top, so I
have been removing the PS belt and alternator belt, then the
crankshaft pulley, and going in from the bottom. I am
rotating the left front wheel hub yada to rotate the
crankshaft the necessary three teeth or so.
1991 Civic, 197k miles here, manual transmission
Darn, the timing belt slipped, and the car would not start.
The engine is a bit flooded from three or so attempts. So
yes, the doomsday scenario has descended. Any insight for
the most efficient way to find the proper orientation of the
TB is welcome. Here is my approach so far:
Set camshaft so sprocket indicates up, marks aligned with
the top of they cylinder head, per manual's direction to get
all valves set "right," namely, such that Cyl #1 is at TDC.
All spark plugs out. Stick rod down Cyl #1 spark plug hole,
rotate crankshaft, and watch rod rise and fall. Turn
crankshaft so that Piston #1 is at TDC. Immobilize
Install timing belt carefully. I am bound to be off a tooth
or two. (Or possibly 180 degrees plus or minus?).
Reassemble, this time hopefully setting the tensioner
Attempt to run engine. If it starts, then I'm off at worst a
tooth or so. If not, then what?
Check timing with timing light. If timing cannot be set
correctly, then I'm likely off a tooth or so.
So far I found I do not have to completely disconnect the PS
pump nor remove the side engine mount to do this much.
Thanks for the elaboration on the tensioner in the other
post, Tegger. This helps. Please comment on the above as it
Don't panic yet.
If I were you, though, I would not at this point try to set the pistons
to TDC. If the cam/crank alignment is off, you need to start from
scratch, being mindful of damage to the valves..
First: Get the pistons all to half-way.
Remove belt, remove spark plugs. Stick a rod down each cylinder until
you find the pair that is closest to the top. Now carefully find out
which way turning the crank *with a socket* results in those going DOWN.
Turn crank that way until the pistons are all at the same height. This
way they are out of range of the valves.
Second: Replace the crank bolt.
Buzz it down to approximately 50 ft lbs. Use washers if necessary to
mimic the spacing of the crank pulley.
LEAVE THE PULLEY AND TIMING COVER OFF!
Third: Turn the camshaft until the #1 cylinder's valves are all closed,
and so the marks on the cam pulley and head are matched perfectly.
Fourth: Turn the crank *with a socket* so the #1 piston climbs UP, until
it reaches TDC. At this point the TDC mark on the crankshaft should
perfectly match the mark on the block. At this point it does not matter
which way you turn the crank.
Fifth: Push the tensioner so the spring is stretched as far as you can
Sixth: Replace the timing belt.
Make ABSOLUTELY certain the cam, crank and head/block marks are aligned
PERFECTLY. It is distressingly easy to get it off by one tooth. You will
likely have to turn the cam pulley slightly to get the teeth to line up.
When the belt goes on, make sure all the slack is on the tensioner side.
Seventh: Double check the cam and crank alignment marks.
Eighth: With the wrench on the crank bolt (or your hand on the cam
pulley), make certain all the slack is on the tensioner side, then
loosen the tensioner bolt with a quick motion. It ought to ZING up then
stop. Still maintaining pressure on the crank bolt, tighten the
Ninth: Triple check the marks and the tension. If OK, remove crank bolt,
reinstall timing cover, crank pulley, crank bolt, etc.
Success! After thinking most of the afternoon this was going
to be the one maintenance job where I surrendered and had
the car towed to a shop, my Civic and I are sitting pretty.
No more rat-a-tat-tat; the timing is perfect (or is near as
a person can see with a timing light and those eye-straining
crankshaft pulley marks!).
I examined that tensioner on and off the car, read and
re-read your notes, Tegger, read and re-read a few other
descriptions for setting the tension, and somehow stumbled
into the right tension setting. I was halfway through when I
read your post a few hours ago, Tegger, and was happy to see
some corroboration for what I was doing. E.g., um,
discovering/remembering this is an interference engine, so
the cam shaft is going to be obstructed unless the pistons
are all about mid-way.
I checked and re-checked the alignment of cam and crank.
When I finally felt I had the tension set, I rotated all a
few revolutions by hand to see that (1) the belt was not
slipping or slapping (it was doing that earlier; now I know
this is a sign of the wrong tension, even though the belt
was holding while I drove it around the other day); and (2)
to re-check my piston #1 TDC with the camshaft's "up" marks.
As for getting the cam and crank aligned correctly, with
belt on, I can understand if no one believes me, because I
am still in shock, but I got it right on the first try. I
thought sure there would be two to five more tries.
I shorted the service check connector and checked the
timing. The distributor needed a tiny bit of rotation to put
the center red mark (on the crankshaft pulley) yada where it
is supposed to be.
I left the PS pump and cruise control actuator disconnected
and off to the side while I troubleshot all this.
Tegger, I still need to explore that tensioner's operation.
This might be worth a write-up, because the darn thing's
operation is still pretty confusing to me. Still, with your
comments and some experimenting with it installed and nearly
all the interference down near the crank removed, I made
progress. I still feel stupid, but less so.
In any event, I very much appreciate your prompt response.
It got me going down the right paths. I did indeed panic a
bit when the belt slipped with the car running.
I think I should go to the casino now, since I surely was
so lucky today. :-)
Two thumbs up! Glad you got it going.
The tensioner can be a bit of a puzzle. I did not mention it before because
I forgot, but it is easy to get the tensioner and spring assembled
incorrectly, in which case it will not zing up properly when asked to do
I myself somehow got the tensioner and spring improperly assembled the
first time I did my belt (had replaced the tensioner). When it did not
respond correctly to my attempts at "zinging" it, I investigated further
and discovered my mistake.
My belt is not due to be replaced for a few years yet, and I did my wife's
just last year, so it may be a while before I manage to get some photos,
unless somebody has some to contribute.
I checked the orientation of the hooks at the end of the
spring on the old setup first, so I think I got those right.
I wish I'd made other observations, though. I think I will
go to my local "U-Pull-It" auto junkyard and see if I can
find some old tensioners still installed.
I wish I understood how the tensioner works well enough to
write this experience up, with photos. I am holding onto the
old tensioner and spring for awhile to see if I become
OTOH, my 91's design is so old, I am not sure it will be of
much value for very long.
operation is very straight forward - need to rotate the engine to
ensure the belt is sufficiently seated on all pulleys and that tension
on all the runs is in equilibrium. then, the tensioner can take up on
the "slackest" run of the belt. the spring is perfect for this.
except your Tercel is non-interference design... :)
Also, is there really a huge danger to damage the valves if you are only hand
turning the engine? I have seen several instances on the group where valves
did not get damaged when the engine was at low speeds. I guess your theory
is 'why take a chance?', and it does make some sense.
Theoretically, yes. But the clearance is on the order of a few
thousandths of an inch. A bit of carbon buildup and suddenly it's an
Absolutely. At all times there will be a few valves that are open. If
the belt is off and one valve is sticking way down and you crank the
engine just a bit too vigorously, that valve goes V-shaped. Serious
Even at high speeds. My boss's CR-V's belt snapped on him on the
freeway. No valve damage. (He was at almost double the recommended
replacement mileage, so it's his fault.)
That's because both cams and cranks are moving of their own inertia.
They will tend to stay more closely in time than the example above, and
will tend to stop more-or-less together as well.
I also suspect some Honda engines are just barely interference designs,
sort of like the Tercel's engine. Or maybe Honda is being conservative
in its labeling practices.
See http://home.earthlink.net/~honda.lioness/id21.html for
photos, along with an explanation of where I am confused. I
started a discussion at honda-tech.com as well. So far, one
person has commented that the "hole" has to go on a certain
peg. Which makes some sense, since I was wondering what the
hole opposite the one where the tensioner attaches is for.
Newer Civic tensioner adjusting instructions make reference
to pegging down the tensioner..., though the design is a bit
I think I may have to go back in there and at least check
everything. All is fine after a few days of driving around
100 miles, highway and suburban. But...
I had a look at the photos on your site.
Something does not look right.
See how the tensioner has a kidney-shaped hole in it? You'll notice the
curve of the hole is axial to the pivot hole shown to the right of the
The spring is shown extending radially from the tensioner. This is surely
incorrect: the spring cannot operate in this orientation. The spring should
be rotated 90 degrees either one way or the other from its current
position. In other words, it should be TANGENT to the tensioner.
The purpose of the spring is to pull the tensioner away from the
crankshaft, pulling the belt with it. The spring must be AXIAL to the
crank, but TANGENT to the tensioner.
See this pic:
I marked it up kinda quickly, so you may have to spend a bit of time
figuring out what goes where.
If this is not clear (and I suspect it will not be), let me know and I will
redo in more dramatic fashion.
Re the photos at
My photo-taking bad: The new tensioner actually now
installed on my Honda does have the spring "acting"
tangentially. What I photographed was a "mock-up" using the
old tensioner and old spring, thrown together hastily with
no attention to detail.
I checked several Hondas at the junkyard yesterday. As far
as the spring orientation is concerned, these Hondas'
tensioner installations look like my Honda's.
I can follow this drawing very easily. Thank you. What you
labeled the "pivot point" is where I may have messed up.
While I wondered about that "pivot point" hole opposite
where the spring attaches, I did not figure out that it was
a pivot point.
At this point I have been searching for instructions
specifically on "tensioner installation," because none of my
FS manuals say anything about looking for this peg and
hanging the tensioner on it... It's not like this peg is
easy to notice, given the tightness of this space. Nothing
in the newsgroup archives mentions this, either. I guess it
should have been obvious that the other, larger hole does
have a function.
I think the only question now is whether I should go back in
there, check everything, and follow the steps for tensioning
again, this time paying more attention to the peg etc.
Before starting the car a few days ago, I rotate the
crankshaft about three times and watched the TB. Also, it's
been 150 miles of problem-free driving since then.
Admittedly I can probably get back in there and out again in
a morning at this point.
Or maybe I can just take the upper timing cover off and feel
to see if the spring is in tension? I think I was getting
away with this at the junkyard yesterday.
Thanks again for your help.
you can drive with a very slack belt for some time, but it could skip at
random. a slack belt also gives flutter on the timing/sensor signals.
best to check it's assembled right and re-tension per the book. ensures
another 100k of trouble free mileage and peak performance.
Motorsforum.com is a website by car enthusiasts for car enthusiasts. It is not affiliated with any of the car or spare part manufacturers or car dealers discussed here.
All logos and trade names are the property of their respective owners.