When installing a new T-belt I set the crank at 30* to 45* BTDC and then move the cams into time. Often one or both cams are turned up to and beyond 360* in the process and there has never been a valve bent in the process.
And in those engines that I have repaired with timing interference damage the vast majority had minor, barely visible dings in the pistons where the valves hit. On a few the valves broke and dropped out destroying pistons as the engine continued to turn.
The Hyundai and Mitsubishi engines were the most common to show up with bent valves for me. Many were driven just a few miles after having a new timing belt installed by a seemingly competant mechanic. The procedures for pretensioning the belt and then releasing the spring tensioner seemed to be confusing to a great many mechanics.
What’s your technique for pre-tensioning and setting, and releasing the idler-tensioner gadget @RodKnox? For my Corolla the FSM suggests to remove the spring and make some measurements on it, like the length of the spring. It is supposed to be 41.03 mm or something like that. Do they expect me to measure the spring to .01 mm? Then to test how strong the spring is, 22 pounds with so many mm of extension. I couldn’t figure out how to do all those tests so, since it was spinning freely, wasn’t making any weird noises and seemed to be working ok as is, I just reinstalled it like it was, then tested the belt tension on the new belt using the ol’ “press-on-it-at-a-midway-point” deflection test. No complaints.
Edit: The most confusing thing for me was which direction that round gadget which sort of guides the belt and goes on the same place as the crankshaft, which orientation that went on.
My experience with the tensioning system used on Hyundais and Mitsubishis involved moving the tensioner pulley into tension against the belt by way of the 1"+/- long slot in the pulley’s center mount. While holding tension against the belt the mounting bolt is tightened securely and after turning the crankshaft through 2 complete revolutions and confirming the belts tension and position on the marks the tensioner spring would be released. It appeared that some mechanics would just install the tensioner loosely and expect that the spring tensioner would take up all the slack. The spring tensioner used on those engines was somewhat common on several Asian imports for many years as I recall but the specific models escapes me. The Isuzu/Honda 3.0 V-6 comes to mind. As for spring tension and length, if that were in question I would recommend replacing the part.
The Geo Metro/Suzuki Swift line of cars use a spring tension type setup You just put the belt on and let the spring do the work, then tighten the bolts on the tensioner. This is also a non-interference engine type unless you have a DOHC variant of it which is kinda rare.
There have been some peculiar set ups for tensioning T-belts in recent years. MoPar had a tensioner pulley that was pre-loaded using a weighted wrench that was hung horizontally while the tensioner bolt was tightened. The Chevette’s water pump was off centered and rotating it tightened the belt. The Escort’s tensioner was held on a sliding rail that required some awkward efforts to deal with. But the Mazda 929 DOHC V-6 was one of the easiest T-belts to replace when the complete package deal was being done.
It’s odd that I recall several peculiar T-belt set ups but can’t picture any Toyotas other than the 3.0 V-6.
The most common T-belt failure was by far stripped teeth, though. And like Ford chains the most common failure situation was at start up.
The Suzuki Reno (same as Forenza) I did this weekend had an eccentric inside the bearing that would rotate with a 6mm hex key. You had to get the bolt tightened down almost all the way, and set the tension based on a little pointer and scale behind the tension roller. The scale had a mark for a new belt and a used belt install. I rolled the engine over a few times, checked the timing marks and the scale to make sure tension hadn’t been pulled out of the belt, and went on with business.
It wasn’t the worst thing in the world but it was certainly something that took a little more time to do correctly. I saw something online about using vise grips between the two cam timing sprockets to hold them in place so I did this with a very light grip of the pliers. I basically had them only tight enough so that they wouldn’t fall off. This worked quite well.
If a piece of the old belt is doubled over with the teeth facing out and positioned between the 2 cam pulleys while they are on ther marks a wooden carpenter’s wedge can be forced in the gap of the belt to jam it in place. I have several factory tools to hold the pulleys but I can use the old belt tricker faster than I can identify which spacer to use. Nothing beats marking the new belt to mate with the pulleys though.