Hi, I have recently been working on a robotic project that uses two atmel at89c51 40 pin chips with a Vcc of 6v. I have been having some serious reliability problems and through searching the message forum I found some suggestions as to how I may go about fixing them. I still have some questions however...

1. Switch bounce is not an issue, software handles the bounce. Should I still put in a cap or schmitt trigger? If so, where and is there a standard cap value? I've also never used a schmitt before??

2. All of my "used" outputs are pulled high (+6v) via a 1k resistor. From there, 470 ohm resistors turn on a 2n3904 transistor, which then feeds into my h-bridge. The h-bridge is fairly robust (uses TIP31/32's for switching the motor direction and 3904's to turn on the TIP's). The motors use 12v and draw about .2Amp when no load and 1Amp loaded. I have a .47uF cap across the motor inputs. (sorry for so much detail I just want to give a good description) The question is are the 1k pullups too low, I'm only seeing about 2.5 into the base of the initial 3904 and I'm using 3 pullups per port (eventually there will be 6 though) Is it possible for the H-bridge to send noisy, inductive, 12volt kicks back through three layers of transistors and cause the chip to "misbehave"? If so, how could I alleviate this?

I can't seem to get the chip to reliably turn on and off my motors. I realize the circuit is complex, but I want to make sure that the chip is not at fault by protecting it's I/O as much as possible. The circuit seems to work for a while and then, under severe loads or through the operation of ancillary components (ie. a 24volt robotic arm unrelated to the uP) the chip begins to show invalid outputs (ie. stuck at '1' or '0', reset does not always yield recovery)
Again, sorry for all the details. If anyone needs more, or has any kind of solution, I would greatly appreciate the help! (I've been working on this for two months and am totally frustrated)
Thanks in advance,
Andy Oxford