I believe that one of the critical flaws with using traps in old school games is the lack of clarity around being forewarned of them or clear guidelines for detecting them, outside of a few powers. This causes traps to often feel unfair, but also to be somewhat uninteresting, because there is almost no player agency involved, with the exception of if it is a big set-piece trap-puzzle combination. The later are fine, but are comparatively difficult and time-consuming to design, so traps tend to occur rarely.
I would propose that an adaptation of my wandering monster table design may assist. The main advantage of using such a table would be to populate it with forewarning of the traps ahead. While there would still be the occasional risk of a trap with no forewarning, this would be considerably reduced and would now exist on a spectrum. One could either use such a table dynamically, or to populate the dungeon with traps.
Here's an example modification of the table to become a trap table. Click on it to see a larger version.
Once again, one rolls 2d6 to determine the trap, and 1d6 to determine which element of the trap is discovered.
Triggers mean PCs have either immediately triggered the trap, while Danger Zone means that they are in or about to enter the area where the trap can affect them. Depending on the trap, Danger Zone may mean it is activated as well, or merely that it if it is activated, anyone within it will be harmed. A result of "Corpses" mean the PCs encounter people who have been slain by the trap, while Detritus is various bits of suggestive dungeon scenery that imply a trap without necessarily making clear what kind. Active Warning means the PCs receive a bit of forewarning if they are actively searching, or begin actively searching the area, while Passive Warnings are elements that are immediately obvious, but not necessarily clear enough to act on without further information.
One side effect of using tables is that PCs will often encounter, through Detritus and Corpses, evidence of traps whose actual location is nowhere near them. I prefer to handle these situations by treating them as indicating the presence of a disabled or previously-activated trap, or one might alternately treating them as where the final evidence of the trap ended up and generate ideas related to how and why the severed half-foot is two corridors down in a pile of rat dung rather than next to the trap itself.
The above table doesn't include the stats of the traps, but any version I use has their stats, relevant saves, etc. to the right, organised by relevant row.