Out of the ICD circuit PIC?

Hi,
I wanted to try a PIC outside of its ICD circuit (connected to the Inchworm) after programming it. I programmed it with the Flash a LED code from William for the 16F628A.

The program works perfectly in the ICD circuit. If I remove the PIC and put it in another circuit, powering it from a regulated 5V supply, it doesn't work. I have a constant ~2V on RB0 where I should normally see it switch from 5V to 0V every ~1.2 seconds according to the code.

Do I have to connect more than just Vss and Vdd to have a functional PIC?
If so, where could I find this information by myself? The Microchip datasheet for the 16F628A is not very helpful for getting started, it is too dense with advanced technical information.

One more thing: When I try to use the Inchworm as a programmer only, there is no problem. If I try to use it as a debugger, though, I get this error when connecting from MPLAB to the Inchworm: " Invalid target device id (expected=0x89, read=0x83)". I use a 16F628A and when I go to Configure -> Select device, there's a green circle beside MPLAB ICD2 Debugger, so it should be supported.
Do the connections to the target PIC have to be different for debugging from just programming?

You need to connect the reset pin (MCLR) to 5V as well, or set it to be done internally (but this causes problems with the ICD2 programmer).

That works, thanks.

Use a resistor 10K or more between MCLR & power, bad practice to tie directly to power.

dknguyen said:

Well it's not bad practice- it's necessary or else the ICD2 can never reset the PIC (or manually either).

Click to expand...

True, it's not bad practice UNLESS you want to reset externally - obviously you have to be able to do this when using an ICD2. Adding a pullup resistor gives no advantage other than that, if you don't need an external reset, then connect it directly - the PIC is designed to be used like that.

I deleted my post as you replied because I realized what was going on (it wasn't running in the ICD circuit).