Atmega328P-PU (arduino uno) from breadboard to pcb

The reset line needs more than just a 10K pull up, it also needs a 100nF to ground to form an RC circuit

see the official schematics where it is clearly shown

http://www.arduino.cc/en/uploads/Main/Arduino_Uno_Rev3-schematic.pdf

They seem to ignore that on the breadboard version.

http://www.arduino.cc/en/Main/Standalone

I suspect its actually for debouncing the switch.

I'd be adding it anyway.

mark

I dont think it has anything to do with the reset switch.

since it works fine on my breadboard.

what i do realish just now is that i dont have any capacitors over the voltage regulator.

on my breadboard i am using a 5 volts pc supply so i dont use a regulator on my breadboard.

it might be possible that i have noise on the pcb version.

I do have the 2 capacitors from the crystal to ground.

where does the 10uf capacitor go. (or is this for the regulator in that case i dont have one at the moment)

you should always have some de-coupling capacitors close to the 328 across the supply, from AVcc to gnd and Vcc to gnd, about 100nF (0.1uF) would suffice for that, and a bigger 10 - 100uF across the supply too on the board (Yes a 10uF in parallel with 0.1uF, it is not for the combined smoothing but the way capacitors internal composition react to different frequencies. the larger capacitor takes care of local smoothing and the smaller ones take care of higher frequency switching noise etc that the larger one will not adequatly handle.

breadboards also have quite a capacitance between the metal strips that can either hide issues or create them and you always need to be aware of this. A breadboard is a great way to prototype solutions but when the system is running at 16Mhz these stray capacitances and Inductance starts to have an affect. so not having a capacitor on the reset to ground can work on a breadboard but may fail on a real pcb as the conditions are very different

I am not saing that the capacitor is absolutly necessary but as mcb1 says, it can provide de-bouncing of a reset switch and in addition it also provides for an additional delay while the power supplies stabilize etc.

My main point here is that, just because something works (Or does not work) on a bread board does not mean the same will be true for when it is transfered to a PCB. for slower analogue systems it is more lightly to work but at higher frequencies or faster pulses things are very different

breadboards also have quite a capacitance between the metal strips that can either hide issues or create them and you always need to be aware of this. A breadboard is a great way to prototype solutions but when the system is running at 16Mhz these stray capacitances and Inductance starts to have an affect. so not having a capacitor on the reset to ground can work on a breadboard but may fail on a real pcb as the conditions are very different

I am not saing that the capacitor is absolutly necessary but as mcb1 says, it can provide de-bouncing of a reset switch and in addition it also provides for an additional delay while the power supplies stabilize etc.

My main point here is that, just because something works (Or does not work) on a bread board does not mean the same will be true for when it is transfered to a PCB. for slower analogue systems it is more lightly to work but at higher frequencies or faster pulses things are very different

One other small thing to check, is the chip in the socket the correct way round? I will admit to making this mistake once, and whilst the chip survived, the circuit (obviously) did not work.

is the chip in the socket the correct way round

I struggle to see it working unless AREF got tied to +Vcc  ... but still worth a mention.

Also worth noting here is that a 328 without an oscillator runs at 8MHz.

This means it will accept an upload but runs at half the speed.

Scroll down to Arduino IDE to see the effect.

http://www.element14.com/community/community/design-challenges/forget-me-not/blog/2014/09/12/forget-me-not-eldermon-hardware-hacking-2

In this case Bob is having trouble with stray capacitance affecting the chip.

I think we need some good quality images of the board (top and bottom) along with your schematic, as the crystal ball is a little foggy.

Mark

I fixed the problem with my circuit. The crystal was fine, but the was a break in the found going to the 22pf capacitors, rendering the crystal circuitry non-functional. Try checking that both crystals are properly connected to ground and that no traces have been damaged.