ODD...
would you mind to open up the controller and post a pic to see what caused the failure?
Anyways, the controllers all use roughly the same schematics and parts,
so if two CT controllers fail.. most likely the YK will too!
I suspect some other major issue here, it's either a bad motor
i.e. bad efficiency rating below 70% and drawing above 40Amps in reality.
a bad gear ratio /wheel combo
i.e. the motor is under constant load and building up too much heat which then decreases efficiency
or bad electrical wiring
i.e. you have a current leak or even a short somewhere.
Many things can be wrong I only named a few.
So

I'm afraid.
I'm sorry, but it's basically simple;
the controller is a
passive unit;
as long as you do not try to draw too much currents
(or it's rated current over a too long period of time)
it should not fail... one can every now and then,
it's still a cheap chinese part...
but two in a row? very unlikely (although still possible of course)
Now any
third controller in that price range is as good as the one you had;
any more expensive controller will also burn and fail if it's -as I suspect- not the controllers fault.
Okay, since you asked... the JLStar Ultipower motor controller is stable up to 50A,
(for the 1000W model) but it's indeed expensive (~150€ or $200 US)
http://shop.mat-con.net/jlstar1kw36v (sorry, only have a german shop for that one ATM)
the same 50A rating (continuous 30A)
can be had with a kelly for a much nicer price
http://kellycontroller.com/kds36050e50a24v-36v-p-138.html
or one step up with 60A continuous
http://kellycontroller.com/kds48100e100a24v-48v-p-285.html
both need a slightly different wiring than what you got now.
But again... make sure the issue is/was indeed the controller, it may be just a victim
'sid