Would a capacitor fix this?

I have a signal (drawn in the image). I want to filter the small "spike" away and just keep the 2ms signal.
How can i do that. With a capacitor?
How do i figure out what value i need then?
The levels are 0 and 5V.

The scematic looks like this:

15V ---[1Mohm]-------------[470Kohm]------GND
................................|
................................|
GND---[Zener 5.1V]-- |-----[cap?]-----GND
................................|
................................|
.............................[PIC]


Edit:
Ignore all the "dots".

I suppose that if it is possible to use a cap. The 2ms signal will change to something like this (picture).
But that's no problem as long as it goes up to 5V it doesn't matter how it
"looks".

My question is, how do i figure out what cap to use. Or is thare a better way that to use caps?

I recommend to use a non-retriggerable monostable multivibrator, triggered by rising edge.

non-retriggerable monostable multivibrator

Click to expand...

Wow, don't think i can find a more advanced word!

What do a non-retriggerable monostable multivibrator really do, can someone please explain?

This device (e.g.74LS221) cannot start new timing cycle while the output active.

To expand on Sebi's (short) explanation:

A monostable multivibrator gives an output pulse of a fixed length, when it receives a trigger signal.

As you probably gathered, there are "retriggerable" and "non-retriggerable" types.

Retriggerable types will restart the timing of the output pulse, if it receives a second trigger, while the output pulse is still active. ie, if a certain multivibrator has an output pulse length of 10ms, and it is triggered, then receives a second trigger 5ms later, the output pulse will be (5+10)ms = 15ms long. Generally, they can retriggered as many times as you like.

Non-retriggerable types must finish the output pulse, before a new trigger signal will have any effect. (It will ignore any triggers whilst the output is active).