Hi there! I'm having issues programming by board. The dreaded:
avrdude: stk500v2_program_enable(): bad AVRISPmkII connection status: Unknown status 0x00
I'm using the avrispmkII and have re-milled another board without much luck. No shorts. Every VCC pin is reading correct 3.3V. The programming header is attached properly – connections confirmed with a multimeter. I'm starting to doubt my board design. Any thoughts?
(The jumper pictured below is a 0 ohm resistor)
Thank you!
js 
Where is pin 1 of the chip??? You have C2, C4 and C6 in parallel and connected to pin 1 of the programming header?? Pin2 should be VCC but is connected to a resistor??
I can't follow the tracks.
As always, WHERE is your schematic? WHY did you not provide it? That is generally step 1 !!
I hope the trace from J2 - is not your ground lead going everywhere...way, way too thin...not even the width of a pad! Give a 60 to 100 mil gnd trace & tap off of that where needed.
You want SOLID. Yeah, you can probably hang out the window using a heavy shoestring if you are careful, but a 1/2" rope is much better.
Generally part numbers & component values are not labelled on the PCB, just give the designator.
Hi js and avrcandies!
Thank you both for your responses. I have since figured out that my issue was that I did not have an external oscillator, which my atmega16u4 (same would be true for at atmega32u4) requires. In my defense, I prototyped the circuit using a breadboard and an atmega168 chip. However, there is no excuse for not reading the datasheet. Below are my schematic and photo of my Frankenstein monster. Yes, I am using a resonator as an oscillator, which weirdly enough works. I since learned that a resonator has capacitors built-in so I am aware that there's a good amount of excess capacitance that will likely cause errors.
js thank you for pointing out my in series capacitors, I used them in excess for decoupling purposes. The traces are thin on purpose, I'm using a 1/64 end mill. Any additional advice would be stellar!
The schematic is a near disaster--it's a drawing to clearly show your connections...why don't you want to?
Why are these shown separated? Were you running out of room? Doesn't the power connector connect directly to the regulator?
================
To avoid mistakes, every interconnection should have a dot
any junction should connect to a wire going to the part with a dot. Only one wire should touch any part pin.
============
Since all connections should have a dot, this implies you have 2 caps connected together & pin 3 is connected to gnd. However, you ppobably want the caps grounded as well.
=================
Also, never make connections using a cross, in case the dot becomes hard to see, or you are glancing at the diagram in bad light.
Here, there can be zero doubt, where the fet is connected to and where VPLUS connects to. Also, VPLUS is an output from this section, so it is drawn on the right of this capacitor bank. If it was supplying power, it would be drawn on the left.
==========
This is a complete mess & ambiguous:
===
Why is you power connector in the middle of the page? It belongs on the left, or possibly right, unless the page was so full it could only go in the middle.
Why is your sensor nowhere near the AVR? Isn't that what it is connected to? It's as though everything is randomly floating around like balloons in the wind.
If this was a drawing for you new house, would you draw the furnace in the middle of the dining room? Put stuff where it belongs, so when you view the schematic it forms a clear picture!!
When you redraw your schematic, you will be much happier with the result
