New chips: ATtiny102, ATtiny104

Go To Last Post
45 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

http://blog.atmel.com/2016/02/23...

 

These seen to be the new "brain-dead" core (16 registers, no MOVW, etc) used on the 6-pin ultra-tinies, but with more pins (8 or 14) and peripherals (a UART!)

1K Flash, 32bytes RAM.  I guess pricing (which doesn't seem to have been announced) will determine whether these are interesting or not.

(avr-gcc supports this core these days, doesn't it?)

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

westfw wrote:
(avr-gcc supports this core these days, doesn't it?)

It supports the 16 register core but in the 4.9.2 build I have from Atmel I don't see 102/104 support:

avr8-gnu-toolchain-linux_x86_64/avr/include/avr$ ls iotn1*
iotn10.h  iotn11.h  iotn12.h  iotn13a.h  iotn13.h  iotn15.h  iotn1634.h  iotn167.h

Perhaps they have an add on "parts pack" for avr-gcc in AS7 to add these new ones?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I was disappointed that the 8 pin pin out did not match the ATtiny85. I need that USART in an 8 pin package.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I2C would be the real killer feature on these for me. They make useful "nodes" that can do sensing and control tasks if you can network them easily.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

richardb49 wrote:
I need that USART in an 8 pin package.
Per its datasheet (page 13, I/O Multiplexing) for the tiny102 (8 pin) the USART is on pins 5 (clock), 6 (TxD), and 7 (RxD).

http://www.atmel.com/devices/attiny102.aspx

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

clawson wrote:
Perhaps they have an add on "parts pack" for avr-gcc in AS7 to add these new ones?

http://www.atmel.com/Images/Atmel_Studio_7_0_790-readme.PDF

...

(page 28)

6. Device Support

...

Atmel Studio includes a tool to manage packs, called Pack Manager.

It is available in the Tools menu in Atmel Studio and gives the ability to install, remove, and list information related to packs.

...

An assumption is new packs are visible if Atmel Studio was installed via the web installer.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

westfw wrote:

 I guess pricing (which doesn't seem to have been announced) will determine whether these are interesting or not.

True, and the pricing showing on the web is not that sharp at 57c/2k &  62c/2k - with a very small 32 bytes of RAM.

Plenty of smarter MCUs for less....

 

Even the SAMD09 has a LOT more code & ram, for ~ 50c 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I got an email from atmel announcing these, and I clicked on the link to get the datasheet, and they wanted me to fill out a form. For a datasheet? Come on.

edit: datasheet is avail on atmel site, no hassle

 

 

Imagecraft compiler user

Last Edited: Wed. Mar 2, 2016 - 09:44 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What I have seen so far is that they are 105C and 125C parts, and that push the price a tad!

Most "cheap" parts are 85C max (some old parts even only 70C) 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Mouser placed an order for tiny104 XNANO boards; ETA 2016-Jun-29

http://www.mouser.com/search/include/aoo_popup.aspx?mouserpartnumber=556-ATTINY104-XNANO

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi, 

It looks that a unique ID is available on these chips

 

20.3. Non-Volatile Memories (NVM)

...

"Extra rows
– Flash - Unique ID needs to be added"

...

 

It not realy clear, is it unique ID per chip ? A unique ID that identify the chip familly ?

 

Are there any other AVR mcu that have this kind of spec that will help me to understand this sentence

 

Thanks

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Are there any other AVR mcu that have this kind of spec that will help me to understand this sentence

Yeah; the (some?) xmegas have it, as does at least the atmega328pb (hmm.  Differently.  IO registers on the PB, signature row stuff on XMegas)

There's been some speculation that undocumented signature bytes may contain similar (lot/wafer/coordinate) data on most AVRs, but ...

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DS says its 12MHz 8MHz 4MHz  VCC= 5V, 3V 1.8V. Guess you could have 3 or 4 batts, and shift the clkpr down as the battery volts decreases? What could one do with a thing like that? Long life temp logger?

 

 

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

Last Edited: Sat. Jul 30, 2016 - 03:18 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Can these be programmed only in assembler? How do you tell the compiler its not a 32 reg avr core?

 

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
Can these be programmed only in assembler?

 

So no high-level toolchains support these chips?  CodeVision introduced support in early May 2016, version 3.27.

 

I thought there was another thread with link to the "part pack" for Studio.

 

bobgardner wrote:
How do you tell the compiler its not a 32 reg avr core?

The same way you do for the Tiny10 family.  Which in essence is the same way you tell your toolchain which AVR model and magic happens to build for the correct architecture.

 

 

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It is hard to compare pricing on AVRs nowadays as it is jumping around.  We just quoted and prototyped a cost-sensitive app and selected Tiny24A.  Same pin count (and footprint?  Close.) as Tiny104.

 

But is the $0.60/1000 pricing going to stay?  That is what Tiny24A was quoted at before the new overlords.  And it will be tricky jamming the app into 1K, and only one timer, and no EEPROM, and similar.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Imagecraft said massaging the compiler to not use half the registers would be like writing a new compiler. I used to do a lot of moto assembler in the 70s 80s and 90s. Wonder if someone could write a sed script to subtract 16 from every R value > R16 and point it to the lis file. Crazy.

 

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
Imagecraft said massaging the compiler to not use half the registers would be like writing a new compiler.

So you have choices:

-- Don't use Reduced Core AVR models

-- Write in ASM

-- Embrace the Dark Side of a different toolchain

 

We did one Tiny10 prototype, but haven't yet designed-in any of the brain-dead models.  Yes, pricing is (at least was) attractive.  And yes, some of the features mentioned above are interesting.  But consider e.g. Tiny13/Tiny25/Tiny24/Tiny1634 and then for each app decide whether it is really a good thing.

 

(aren't there Cortex M0 models for less than a buck?)

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

Last Edited: Fri. Jul 29, 2016 - 08:43 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch wrote:
I thought there was another thread with link to the "part pack" for Studio.

http://packs.download.atmel.com/#collapse-Atmel-ATtiny-DFP-pdsc

1.0.78

Added ATtiny102, ATtiny104, ATtiny80 and ATtiny840. Updated documentation links.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch wrote:
So no high-level toolchains support these chips?  CodeVision introduced support in early May 2016, version 3.27.
The beta for tiny10 was Atmel AVR 8-bit GNU Toolchain 3.3.2.485 (GCC 4.5.1) and apparently full-up for 3.4.0.663 (GCC 4.6.2).

http://www.atmel.com/tools/STUDIOARCHIVE.aspx

At least several issues have been corrected since then; current issue :

http://www.atmel.com/Images/avr8-gnu-toolchain-3.5.3.1700-readme.pdf

...

Issue #AVRTC-731:
For AVRTINY architecture, libgcc implementation has some known limitations. Standard C / Math library
implementation is very limited or not present.

...

http://www.atmel.com/tools/atmelavrtoolchainforwindows.aspx 

Atmel AT12489: Getting Started with Atmel ATtiny102 and ATtiny104

http://www.atmel.com/Images/Atmel-42678-Getting-Started-with-Atmel-ATtiny102-104_ApplicationNote_AT12489.pdf

(go to page 10 for "2. Creating an Example Application in Atmel Studio" for a button blinky in C)

http://www.atmel.com/devices/ATTINY104.aspx?tab=documents 

 

IAR Embedded Workbench

https://www.iar.com/iar-embedded-workbench/#!?device=ATtiny104&architecture=AVR

 

Edit : IAR

 

"Dare to be naïve." - Buckminster Fuller

Last Edited: Sat. Jul 30, 2016 - 01:18 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
Can these be programmed only in assembler?
Only in assembler for GCC 4; C was added for GCC 5.

Appears that Atmel backported C for minimal AVR into GCC 4.

Appears C on minimal AVR went upstream into GCC 5.

https://gcc.gnu.org/onlinedocs/gcc-4.9.3/gcc/AVR-Options.html#AVR-Options (the ISA is avr1)

https://gcc.gnu.org/onlinedocs/gcc-5.3.0/gcc/AVR-Options.html#AVR-Options (the ISA is avrtiny for mcu = attiny10, ...)

An alternate AVR GCC toolchain on Windows :

Sysprogs

Prebuilt GNU toolchain for avr

http://gnutoolchains.com/avr/

...

GCC 5.3.0

...

bobgardner wrote:
How do you tell the compiler its not a 32 reg avr core?
For AVR GCC it's "-mmcu=avrtiny".

https://gcc.gnu.org/wiki/avr-gcc#Using_avr-gcc

 

Edit : typos, edit

 

"Dare to be naïve." - Buckminster Fuller

Last Edited: Sat. Jul 30, 2016 - 02:19 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch wrote:
It is hard to compare pricing on AVRs nowadays as it is jumping around.
Especially for a part that's new to stock.

https://octopart.com/search?q=ATtiny104

theusch wrote:
But is the $0.60/1000 pricing going to stay?
No due to vagaries of supply, demand, profit, economics (the study of scarcity).

I get off my a**, pull out the wallet, open the wallet, up end the wallet, pour money frown

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Wonder if someone could write a sed script to subtract 16 from every R value > R16 and point it to the lis file. Crazy.

The so-called brain-dead tinies omit r0-r15 and keep r16-r31.  In this way the existing non-orthogonality of the remaining registers is preserved w.r.t. their special functions in the standard and enhanced cores (e.g. r26/27 -> X, etc., and LDI working on registers r16 and above).

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

theusch wrote:
... and no EEPROM ...
10microfarad X5R MLCC across the coin cell, enter power-down (110nA typical at ambient room temperature, no watchdog, no BOD), pull coin cell at 2.7V (minimum for 8MHz), BOR is at 1.6V maximum.

C = dQ/dV

C = (dQ/dt) / (dV/dt)

dV/dt = (dQ/dt) / C

dV/dt = i / C

dt = CdV/i

dt = 10uF * (2.7V - 1.6V) / 110nA

    Assumption : Icc-power-down >> capacitor's leakage

dt = 100s

Hope the operator doesn't screw the hatch (battery door) and/or fumble the new coin cell.

Less risky to use a rechargeable coin cell though that increases the number of parts (connector, diode, CC/CV external power).


IMDb

The Right Stuff (1983)

Quotes

http://www.imdb.com/title/tt0086197/trivia?tab=qt&ref_=tt_trv_qu (search for pooch)

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

mojo-chan wrote:
I2C would be the real killer feature on these for me.
Can get some range with I2C bus buffers and bridges.

mojo-chan wrote:
They make useful "nodes" that can do sensing and control tasks if you can network them easily.
There are some low power or micro power RS-485 transceivers.

tiny102/tiny104 have USART Multi-Processor Communication Mode (MPCM) that's on a number of megaAVR, XMEGA AVR, and apparently Multidrop Mode for SAM S70 and SAM A5.

A terminal server can link and bridge to Ethernet.

Could bridge to USB then into a USB hub.


Maxim Integrated

Maxim

MAX3471 1.6µA, RS-485/RS-422, Half Duplex, Differential Transceiver for Battery-Powered Systems

https://www.maximintegrated.com/en/products/interface/transceivers/MAX3471.html

 

ChiliPeppr - Hardware Fiddle

http://chilipeppr.com/

...

Serial Port JSON Server

The heart of ChiliPeppr is a server that links your serial port devices and your host controller like a Raspberry Pi via websockets to your browser.

...

 

Edit : MPCM

 

"Dare to be naïve." - Buckminster Fuller

Last Edited: Tue. Aug 2, 2016 - 05:41 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
theusch wrote: ... and no EEPROM ... 10microfarad X5R MLCC across the coin cell, enter power-down ...

???  What does low power have to do with EEPROM?

 

The app I have in mind has a few setpoints much easier placed in EEPROM than flash.  And a cal value for the internal bandgap.  It would be nice if there is an updatable "user" 'configuration page'.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Use of SRAM in lieu of EEPROM if the operator can quickly swap a coin cell.

Agree that EEPROM is easier and it's also reliable.

If there's a spare pin, there are single-wire EEPROM and single-wire crypto authenticator with EEPROM.

http://www.atmel.com/products/memories/serial/single-wire.aspx

http://www.atmel.com/devices/ATECC508A.aspx (10Kb EEPROM, high-speed single pin I/O)

Not certain C for an external EEPROM would leave much for an application on a 1kB minimal tinyAVR; might use assembly language.

tiny40 (minimal AVR) does come in SOIC-20 and TSSOP-20 (temperature 85C max for both).

http://www.atmel.com/devices/ATTINY40.aspx

theusch wrote:
It would be nice if there is an updatable "user" 'configuration page'.
ATtiny104 does not have that but does have self-programming and the flash is accessible as data (starts at data address 0x4000).

There's a short assembly language snippet for self-programming :

http://www.atmel.com/Images/Atmel-42505-8-bit-AVR-Microcontroller-ATtiny102-ATtiny104_Datasheet.pdf (go to page 202 for "20.5. Self programming")

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
Use of SRAM in lieu of EEPROM if the operator can quickly swap a coin cell.

I figured out your point later. ;) 

 

Still, I don't want a field unit essentially inoperable due to static zap or whatever.

 

gchapman wrote:
If there's a spare pin, there are single-wire EEPROM and single-wire crypto authenticator with EEPROM.

But then I might as well use a Tiny or Mega with ~same total cost and less board space and more flash space and ...

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
Imagecraft said massaging the compiler to not use half the registers would be like writing a new compiler.
Minimal AVR did make it into the assembler though it's not in the compiler documentation.

https://imagecraft.com/pub/readmeAVR.txt

...

V8.24.00 - 2016/05/26

...

  Assembler
  - Added ".CPU AVR_reduced_core" directive to support the reduced core
    TINY (e.g. ATTiny10) where there are only 16 registers and lds/sts
    are 16 bits instead of 32 bits. When the directive is specified,
    the assembler generates code appropriate for these TINYs.

...

Imagecraft

Build Options - Target

https://imagecraft.com/help/ICCV8AVR/iccavr/3-ide_codeblocks/build_options_-_target.htm#IX_Target

...

  • Instruction Set - Select instruction set supported by the device: Classic, Enhanced, Enhanced (no MUL), or XMega. Only changeable if the device is set to Custom.

...

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch wrote:
CodeVision introduced support in early May 2016, version 3.27.
http://hpinfotech.ro/cvavr_revision3.html

http://hpinfotech.ro/cvavr_chips.html

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
How do you tell the compiler its not a 32 reg avr core?

 

I really cannot fathom this device family.

What does removing 16 registers really save  ??

 

theusch wrote:

But then I might as well use a Tiny or Mega with ~same total cost and less board space and more flash space and ...

 

Exactly, if the part was sub 25c, I could understand better, but  56c/1k for a 8 Pin, minimal 1k / 32 byte device, is way off the 2016 price curve ?

 

SiLabs EFM8BB1 is 32c /1k for 2x Flash,  8x RAM, and better ADC etc, in 20 pins

 

The Nuvoton N79E715AS16 is SO16N, and has a massive 16k Flash, 512 Ram, and similar ADC, for 28c/1k on their webstore.

 

ie you can literally buy two N79E715AS16 for the price of one ATtiny102 ??

 

sub 56c can buy you 20KF/496R from Cypress, 8KF/256R from ST, 3.5KF and 256R from Microchip, 16KF/2.25k/50MHz from SiLabs & TI & Renesas have parts too...

 

Does rather look like Microchip want to 'niche' the AVR to those with design-inertia.

 

 

Last Edited: Sun. Aug 7, 2016 - 06:34 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

sparrow2 wrote:
What I have seen so far is that they are 105C and 125C parts, and that push the price a tad!
A use case for a 125C part is in a surgical instrument undergoing sterilization.

Battery University

Charging Batteries at High and Low Temperatures

http://batteryuniversity.com/learn/article/charging_at_high_and_low_temperatures

...

(about mid-page)

Figure 4: Capacity loss at room temperature (RT) and 130°C for 90 minutes
Sterilization of batteries for surgical power tools should be done at low SoC.

Test: LiCoO2/Graphite cells were exposed to 130°C for 90 min.at different SoC between each cycle.

...

Autoclave Time Temperature Pressure Chart

http://www.sterilizers.com/autoclave-time-temperature-pressure-chart.html

 

Another use case is for steam wash of food processing equipment.

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Digi-Key now has stock for the 125C version.

No change in the Mouser ETA (2016-Oct-18).

https://octopart.com/search?q=attiny104&avg_avail=(1__*)&start=0

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

Last Edited: Thu. Mar 2, 2017 - 03:25 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:

Sale (-25% until 2017-Mar-31) though no stock until 2017-Mar-29 :

 

So now the same price as the ATTINY817-XMINI  ? 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

Last Edited: Thu. Mar 2, 2017 - 03:24 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The two first links on this page :

 

http://www.microchip.com/Develop...

 

are dead ended!

 

so I guess that microchip have the same problems as Atmel :(

 

Where to repport? 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Microchip support probably (i.e I don't know of the top of my head)

:: Morten

 

(yes, I work for Atmel, yes, I do this in my spare time, now stop sending PMs)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Microchip Technology Inc

Microchip Technology Inc.

Contact Us

http://www.microchip.com/about-us/contact-us

...

Website Corrections
Occasionally a typo may slip through and be displayed online. We strive for quality and will make every effort to correct any mistake pointed out to us. Please send details on all typos and any other problems you encounter on the website to Web.Issues@microchip.com.

...

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Who-me wrote:
Exactly, if the part was sub 25c, I could understand better, but  56c/1k for a 8 Pin, minimal 1k / 32 byte device, is way off the 2016 price curve ?
Reached 0.25USD each for 100 via the Digi-Key tiny102 sale.

https://www.digikey.com/product-detail/en/ATTINY102-SSFR/1611-ATTINY102-SSFRCT-ND/6833015

via

https://octopart.com/attiny102-ssfr-microchip-77761926

 

"Dare to be naïve." - Buckminster Fuller