That new Atmel Tiny4

Yo, what!?!? SOT-23-6 8-bit micros?? Where? I'll consider them even if they are from Atmel! I'm off to go look.....

Jon

Hmm. Never mind. I forgot. I looked at that 4MB datasheet on Nov.

Jon

For interest, what feature were you looking for?

--------------------------------------- This message was sent using the comp.arch.embedded web interface on

Digikey ???

Super, but no Eeprom so limited use. good pwm etc,

Availability at Digikey (not that I buy them there, but it is an important indicator) and Mouser. ATTiny10 is qty 0's at Digikey, non-stock at Mouser. Reliable, programmable brown-out detect and POR, internal ring osc with 2-5% over temp, small eeprom, comparator, and bandgap -- adc as an option instead of the comparator, 5.5V down to maybe 2V operation (eeprom, included), 512-1kb flash, 32+ bytes ram,

Actually, it's the BOR and POR that are more crucial. I like these for specialized power-on reset and brown-out chips for other micros which don't have the brown-out built in. I often want to include a manual switch for resetting the board, as well. The _many_ SOT-23s out there for this kind of purpose do _not_ include the programmable brown-out nor the ability to store some info in eeprom, for example, and I want it all in one at similar price and size. I don't want to buy 10 boxes of almost-the-same chips. Rather just program up what I want from a single part number.

Jon

You missed Temperature sense ?

I think Atmel chose to clone a PIC socket, and not look at the wider scope.

Yes, EE is nice, but the paddle size in SOT23 is _very_ small, and EE has a high die cost.

Perhaps in a _slightly_ larger package ?

-jg

I don't need much EE. 8 bytes? But I think you are referring to the charge pump, which may be unavoidable over such a Vcc range?

Of course, I also wanted 10mA sink, 5mA source ... or better. And that takes die space just for the fets alone. And the aluminum traces that carry that current, I suppose. But they benefit from not needing that many wire bond pads, yes?

In any case, that's the application space for which I'd like a fill. Nice BOR and POR, ability to accept and debounce a manual switch, the ability for modest communications of parameters to the main cpu it is responsible for reseting and dealing with brown out situations, and tiny and cheap and versatile and ...

Jon

There has been a lot of discussion about them on AVR Freaks. Even the Tiny10 which was touted months ago isn't officially available yet, although one or two people have been able to get hold of a few samples. There doesn't appear to be any programming or debugging support for them yet in AVR Studio, but someone has manged to build his own programmer and write some software for it.

Leon

I have some samples of the Tiny10 from Arrow, and Farnell claim they will have stock on 15th Jan 2010, otherwise everyone is quoting 7 weeks lead time for a reel of 5K.

I'm struggling to find a means of programming them in-circuit at the moment though. Anyone know of a third party programmer which will do this?

Regards

Martin

his?

You will have to build your own programmer, as I said previously. See AVR Freaks.

Leon

">

Where's the usable eeprom then ????

Microchip have had SOT23 MCUs for ages.... PIC10F

I didn't say there was any. You may have thought I was talking about the part, when I was actually stating what I'd like to have in a SOT23-6 micro. Different thing.

Jon

Yeah. Not with brown out. :(

Jon

Jon Kirwan skrev 2009-12-18 09:55:

Probably to new, but that will change

There is a POR, and a BOD that can either be programmed to generate an interrupt, or a reset, but it is not very configurable, and not fully documented in the last datasheet.

min typ max Trigger level VLM1L 1.1 1.4 1.6V Trigger level VLM1H 1.4 1.6 1.8 Trigger level VLM2 2.0 2.5 2.7 Trigger level VLM3 3.2 3.7 4.5

R/C oscillator, same as on other AVRs

No EEPROM!

Comparator yes, No bandgap!, references vs VCC

ADC

Yes 1.8V - 5.5V, Can only be programmed at 5V though.

Yes ATtiny4/5 = 512, ATtiny9/10 = 1kB

Yes

No clue

TBD, they use a reduced core with only the 16 upper registers. No instructions to access flash directly, but flash is also mapped into dataspace at 0x4000. Cannot write to flash so no EEPROM emulation. Think that may make things a little difficult for existing compilers. Codevision may be the first to support.

Yes

Yes

Yes, 16 bit timer.

Yes

Yes

When the internal BOD triggers, the CPU can be reset and the I/Os become an input.

By adding an external pulldown resistor, you can assert a reset. when the Brown-Out occurs. If you trigger an interrupt, the resistor is not needed.

Lack of EEPROM is the main issue here.

STK600 with new ATTINY10 topcard is what you can use now for development. You can obviously connect the ISP header to your target instead. Just have to make sure you have the 5V.

AVR ISP Mk II and/or AVR Dragon with upgraded firmware will be available early next year.

Lack of presence at Digikey and Mouser, as well. And the lack of a bandgap -- no, Vcc won't do. But I appreciate very much your other replies -- particularly on the BOR, which I will look at with your comments in mind. Since the flash cannot be written and therefore no eeprom emulation and no eeprom, you are right that's another issue. But... I think I can live with the problem if that were the only one.

Jon

I'm looking at 8127C-AVR-10/09 and on page 28 I find that there is "vcc monitoring" and a VLM circuit. But for gosh sake that darned thing sucks power -- page 146, curve on the page's bottom. Cripes.

Jon