I'm laying out a test PCB with an ATTiny45 and an ISP header.
Do I need a resistor between RST and VCC, and a cap between RST and GND?
I've seen it both ways.
I'm laying out a test PCB with an ATTiny45 and an ISP header.
Do I need a resistor between RST and VCC, and a cap between RST and GND?
I've seen it both ways.
No cap on reset. You don't need the resistor between RST and vcc, but it's a good idea (better noise immunity).
You generally don't want a cap to ground on reset, that could interfere with proper programmer operation. Sometimes you do want a cap from RST to the dtr pin of a serial adapter for autoreset when serial connection is opened - but as the tiny x5 series doesn't have a hardware serial, that probably doesn't matter to you
You do need a 0.1uf ceramic cap between vcc and gnd right next to the chip (trace length matters - thats why I have a pile of non-salable project boards for x5's - the decoupling cap was too far away.
Yep, I have decoupling.
10k RST to VCC, right?
Ya.
Depending on your programmer, you could use debugwire rather than ISP programming and not need the ISP pins at all.
Just a connection to the reset pin and ground.
debugwire is quite useful on the small pin parts like the tiny.
--- bill
I know nothing about debugwire. I'll google, unless you have a great pointer.
See section 18 of the datasheet:
Debugwire is pretty nice. The drawbacks are that it is much slower than ISP and you have to use a programmer from Atmel as they have not disclosed the protocol so there are no 3rd party debugwire programmers.
--- bill
I'll add this here: The debugWire protocol
It's been under discussion recently on AVRFreaks: http://www.avrfreaks.net/forum/why-medbg-so-slow-load-and-run-xplained-328pb