AVR Freaks

The AVR DA datasheet says that VDD and AVDD are internally connected.  What's the point of L200 on AVDD on the AVR128DA48 CURIOSITY NANO if they are not going to do the same on VDD?  Is this just cargo cult PCB design?  Or does it really serve a purpose?

Reference: http://ww1.microchip.com/downloa... , page 29 for schematic.

Good to know.

I measure 3.3 ohm resistance between VCC and AVCC in both directions with a multimeter on an AVR128DA28. This is the best I can do for you... this forum is made up mostly of users of AVRs, we didn't design them...

Thank you.  That's quite helpful.  And a bit disappointing for my use case.

Those look very nice. I wish they had USB though. Maybe I'm the only one doing USB on 8 bit any more...

There is still a big market for 8 bit MCUs because they have some advantages. The 32 bit ones, mostly ARM, can't match them on power consumption in many applications. They can sometimes compliment more powerful MCUs as well, e.g. you might have an ARM that has variable interrupt latency so you throw in an 8 bit MCU to handle some particular task.

They just work, they are solid and well understood, and you don't have to rely on buggy libraries. AVR assembler is much nicer than ARM too. I still prefer 8 bit where possible.

5V Cortex mcus:

Atmel SAMC20 and SAMC21 (2% internal OSC)

NXP E-series (1.5% internal OSC)

Nuvoton - well, lots, they seem to have more 5V than 3V (2%)

GD32, HK32 from China, not so common though

I also still prefer 8bitters, even some 8051 based ones. My wonder is why Microchip develops so many AVR variants. For a while I thought Mega0 would be a perfect replacement for some of my older designs (mega8, 88pb, mega644) now I'm not sure. I feel I would go for already an obsolete choice. And I don't prefer Dx series over Mega0 because their internal oscillator spec is just worse - a small step down that can put an entire cpu family off for me. 

5V Cortex mcus:

Also a bunch of Cypress parts, including PSOC4 and PSOC5.

Feels like one step forwards two steps back with Microchip. Every time you get something they take something else away.

I like these boards, basically just all the I/O on pins and a built in programmer. Of course I liked it more when the debuggers were $30 as well.

Also glad to see they will keep supporting Atmel Studio. Not a fan of MPLAB X and my worry is that one day they try to force everyone onto that, or just stop putting any effort into AS and letting it slowly die.

Thanks, I'll take a look at the Snap. Kinda odd they don't offer a case for a few pennies, in volume it wouldn't cost much. Also you need adaptor cables if you are already using the Atmel standard headers...

That's certainly true, although I standardized on a 0.1" pitch header for prog/debug so only need an off-the-shelf ribbon cable assembly, with optional pogo plug board. Only needing 3 wires to program is very nice.

"Increased Current Consumption May Occur When V_DD Drops

The device may experience increased current consumption of approximately 1.5 mA if V_DD drops below 2.1V and is held in the range of 1.9-2.1V. This will only occur if V_DD is originally at a higher level and then drops down to the mentioned voltage range."

So if you are running from 2 alkaline batteries you are in trouble.