A bit more or traps generally...
The best traps put the whole party into some sort of dilemma, where the party members must extricate themselves from the trap and rescue their fellow party members. That is to say, the best traps go off reliably, take multiple rounds to kill the party, and effect the party as a whole either directly by putting them in danger or indirectly by forcing them to rescue whomever has been put in danger.
Good traps therefore do some or all of the following:
a) Move or separate party members. Wind traps, pit traps, anti-gravity traps, harpoon traps, vacuum traps, rushing water traps, falling portcullis traps, and so forth separate and confuse the party and force them to set priorities about what to do next. You don't want the party members off in separate rooms. You want them dealing with problems in different parts of the same connected space.
b) Trap party members. Impaling traps, pits with locking covers (preferably grates), bear traps, cages and so forth put party members in need of rescue and extrication.
c) Have an obvious timer that the party must fight against. You want to do small amounts of damage over time, not lots of damage up front. Or in the case of a trap that is conveying the party to some horrible end (a meat grinder, a pit of lava, a lethal fall, a vat of acid, a pit filled with venomous creatures), you want it to be a struggle for several rounds to avoid being conveyed to the horrible end with different party members in different levels of predicament trying to figure out how to help each other. So traps that cause players to bleed (such as being impaled on a barbed spike) or traps that fill up rooms with dangerous substances, or perhaps which release a tide of poisonous scorpions that are now slowly advancing on the players position, or otherwise take time to become lethal are preferred for game purposes over instant death traps.
d) Go off in stages and become an interacting puzzle more lethal than the sum of its parts. Serious traps start off bad and things just get worse. It's cliché, but you really do want the players going, "It can't get worse than this... it just did." By working on multiple levels, you can party members working on various problems - light sources, damage mitigation or resuscitation of commission party members, breaking parts of the trap, trying to disarm parts of the trap, fighting off the monsters that arrive while all of this is going on, trying to solve a puzzle while this is going on, or whatever.
Finally, I want to caution you against two mental 'traps' that trap designers can get themselves into. The first is trying to hard to play 'gotcha' with your players. Don't make it a contest you are trying to win. You have infinite resources as the GM. There is no way you can lose if you don't want to lose. The intention is for the trap to lose against a clever party and for the party to have fun doing it. The intention is not for them to be awed by how devious you are. They still might be, but don't lose sight of the real goal. And the second is related, and that is to avoid the temptation to use a bunch of reverse logic. You do NOT want to punish clever planning. If you start introducing reverse logic, then there is no way to plan for problems since a good plan might be a bad plan because the 'adversary' is counting on the adventurers to be clever. It's perfectly OK to give good clues that something is trapped. It's not OK to give good clues that something is trapped in order to get the party to react to that clue and blunder into the thing that is really trapped. If you are doing that, then the you've turned the whole affair into a crap shoot. At the very least, be consistent. If you really want to have the 'adversary' employ reverse logic, then he should have done so consistently through the whole dungeon and the first trap should serve as a hint to this thinking (without being overly lethal).