Changed from 18x 0.5mm GND vias to 16x 0.6mm GND vias for spacing reasons.
Pulled GND via circle in a lot. The entire minimum annular ring for each via is now spaced the minimum trace distance from the board edge. This should be fully compliant with OSH Park’s design rules.
Swapped out lame LFPAK56 outline. It was nowhere near correct and just wasting all kinds of space. I used a Power-SO8 outline that came with Eagle and filled in the blanks by hand for now. I reduced the amount of solder paste laid down under the tab vs the old outline, somewhat like what NXP actually recommends. I’ll fix a real part later.
Tilted the squiggly line of vias at a rakish, dashing angle more fitting with today’s styles and trends.
Minor silkscreen changes.
I think what’s left to do at this point is pin down a name and then clean up the sloppy silkscreen on the bottom.
Wight, have you had a chance to play with any of the LFPAK33 FETs yet? I ordered some and have a few boards almost done with them, but haven't had a chance to try any yet. I put one on the 15mm version of this board and according to NXP it looks like they are pretty much equivalent to the LFPAK56 FETs.
Heh, frankly I haven’t thought any about them, although as you probably realize I have a couple on hand. I haven’t really shopped to see what’s available. Where it fits I feel like Power-SO8 is the ideal footprint since there are so many options available. It’s a very popular package and no one is likely to be unable to source appropriate parts.
That said, when I re-attack buck circuits I’ll be looking hard LFPAK33 and the other 3.3x3.3 MOSFETs. I can see why it’s a good choice for the 15mm layout.
I forgot to mention this: when I ran the CAM job for v009 I ran both the old and the new job and uploaded both to OSH Stencils. Sure enough, both were off. I think it’s just a flawed setup on their end.
AFAIK the buzzing is something resonating. Tuning your PWM should help eliminate it. Change duty cycles (mode levels), PWM freq, or PWM type (fast/phase correct). The Dual-PWM functionality can split PWM types right in the middle of your mode group if you’d like, that may be able to eliminate the noise at the top end. That’s what’s going on with the “#define FAST_PWM_START”.
Looking forward to pics of your next driver build.
9.6Mhz is right on the ragged edge with a protection diode in place. Atmel’s datasheet shows that up to 10Mhz is OK for 2.7-5.5v, but the diode drops around 0.2v or more. So at 2.9v the ATtiny13A sees 2.7v. There may be other considerations that push you over the edge in certain circumstances.
In other words if it’s possible I’d try and stick to 4.8Mhz rather than 9.6Mhz, just to have all your bases covered all the time.
What offtime values are you using and how much time is it taking to save?