hi - Is there a way to debug attiny 6pin? ( ATtiny4, ATtiny5, ATtiny9, ATtiny10 )
I saw the datasheet ( ATtiny4/5/9/10 ) - it only supports "TPI" ( And "TPI" is for programming only and not for debugging )
hi - Is there a way to debug attiny 6pin? ( ATtiny4, ATtiny5, ATtiny9, ATtiny10 )
I saw the datasheet ( ATtiny4/5/9/10 ) - it only supports "TPI" ( And "TPI" is for programming only and not for debugging )
For a part that small, just look at your listing & figure out your error---you couldn't have more than a 100 lines of code.
What's the error?
I checked out all 8-bit avrs ( tiny / mega / xmega ).
And all micro have at least one debugging method.
Except this 4 micro ( ATtiny4/5/9/10 - with 6 pins ) that have no debugging methods ( I did not find, This is a better statement )
hi - Is there a way to debug attiny 6pin? ( ATtiny4, ATtiny5, ATtiny9, ATtiny10 )
you couldn't have more than a 100 lines of code.
In assembler, the 512-byte t4/5 have 256 instruction words.
But I agree. Can be done in your head.
Thanks to all the friends.
For a part that small, just look at your listing & figure out your error
The other commonly used alternative is to debug using a larger chip that DOES have some sort of on-chip debug.
The tiny tinys do have some unique limitations, so there's a certain amount of care you'll have to take not to use non-tinytiny features, and if you're using interrupts you may have to fiddle with the vector table...
ATmega8, ATmega8515, ATmega8535 have no on-chip-debug.
The ATmega8 is still commonly used.
.
I suggest that you develop on a better device. Then rebuild for the cheapest suitable target that is available in your country.
.
David.
ATmega8, ATmega8515, ATmega8535 have no on-chip-debug. The ATmega8 is still commonly used. . I suggest that you develop on a better device. Then rebuild for the cheapest suitable target that is available in your country. . David.
The full list is as follows :
ATtiny4
ATtiny5
ATtiny9
ATtiny10
ATmega8/A/L
ATmega8515/L
ATmega8535/L
I suggest that you develop on a better device. Then rebuild for the cheapest suitable target that is available in your country.
this is good idea
The full list is as follows
If you are starting out with AVRs perhaps consider the 40 pin DIP packaged 164/324/644/1284P series. They all have JTAG which is probably the easiest to use debug interface. The chips also generally have at least "one of everything" that you can find in most mainstream AVRs so they make a good place to learn.
I meant only series tiny, mega, and xmega - Not all avr with cpu 8-bit
So did I.
For one thing I can't think of any "Tiny" that has a debug interface (this is part of what makes them "tiny"/cheap!). But there are "mega"s besides the ones you listed that don't have debug either as I say anything with a model name that starts AT90S is so old as to predate debug interfaces.
I forget how many 8bit "AVR" there are these days - perhaps around 350? There's at least 50..100 of them that don't offer a debug interface.
avrcandies wrote:
you couldn't have more than a 100 lines of code.
In C, maybe. Much less in fact.
While I agree that the address space at 250 or 500 words is quite limited, can you give me any basis for this "much less than 100 lines of C"? How do you count? Are you doing Obfuscated C contest with everything on one line?
I suspect the statement could be more accurately states something like this: "The toolchain that I know and love, with infinite value, does not play well with brain-dead AVR models, and couldn't fit anything useful into the Tiny10 family -- much less Tiny4/5." That doesn't mean anything about "C". Sure, any competent or guru ASM programmer would find a way with enough time and effort to pack a bit more into the code space. But that doesn't necessarily mean that a C implementation of the same app wouldn't be feasible. BTW, how many lines of code will your ASM implementation take?
how many lines of code will your ASM implementation take?
I can't think of any "Tiny" that has a debug interface
I have an ATtiny4 app with almost 600 NCLOC, including macros, directives, labels, etc. 248 instruction words, 16 bytes of LUT. 100% flash use.
So you guys are saying that a C implementation would take how many lines of C? Or is the implicit claim that it can't be done?
In the below fragment, the ASM action code has twice as many lines as the C implementation. Is that good or bad? I could spread it out with a token on each line.
int accumulator; int increment; ... accumulator += increment; ...
;0000 000B accumulator += increment; + 000022 0f24 +ADD R18 , R20 000023 1f35 +ADC R19 , R21
I don’t think we’re dealing with absolutes here as a LOC is nebulous at best. Nevertheless, as a guestimation, the likelihood of the avg 100 lines of C fitting into a small memory device is not looking good but not impossible.