
Chris Holden of NerdClub shares his tips on how he successfully program an ATMega128:
Finally got an ATMega128 chip coded and programmed successfully. The great news is it doesn’t require Arduino. The even better news is, we can use Oshonsoft to write the code! Yay.

I don’t get it why people still use the AVR soft USB (like the USBasp does).
It’s unreliable, especially in noisy environments, and it fails you when you need it the most.
So why still use it when you can get $1 USB micros?
When you’re using 49 of the 51 available i/o lines and don’t want to have to solder extra components like caps and crystals and take up i/o pins for dedicated USB on the chip, that’s when some people resort to low voltage programming over serial/uart! Why the programmer doesn’t use dedicated USB hardware, I don’t know (I didn’t realise that it doesn’t). I only program these game boards in a controlled environment so noise isn’t a problem.
The 12mhz version of v-usb used with the USBasps seems to be the most reliable.
There’s actually 5 different versions, so you can’t equate results with one version to all.
Since when can you get an AVR with hardware USB for $1? And where???
You wouldn’t believe it, but there are other microcontrollers than AVR.
http://www.microchip.com/wwwproducts/Devices.aspx?product=PIC16F1454
You wouldn’t believe it, but there are other micros than PICs.
http://www.microchip.com/wwwproducts/Devices.aspx?product=PIC16F1454
p.s.:
hardware USB V-USB
differential transmitter yes yes
differential receiver yes no
controlled impedance yes somewhat
controlled slew rate yes no
noise suppression yes no
oversampling yes no
USB compliance certification yes no
Notice the difference?
Sorry, layout was messed up.
p.s.:
hardware USB V-USB
differential transmitter yes yes
differential receiver yes no
controlled impedance yes somewhat
controlled slew rate yes no
noise suppression yes no
oversampling yes no
USB compliance certification yes no
Notice the difference?
I give up. This comment system is broken.
@Cyk: some of your comparisons are red herrings, like slew rate. The v-usb software does not control the slew rate – the external RC circuit does that.
You also left out one important comparison: cost. V-usb has been done on an Attiny10 that costs 70c for qty 1.
http://cpldcpu.wordpress.com/2014/03/19/%C2%B5-wire-usb-on-an-attiny-10/
Yup, comment system isn’t working as of 15:55-ish UTC. I sent an Email to DP with my browser-shown error via the Contact Form, which does appear to be working at the moment.