AVR-Dopper in HVSP propably cannot work for t11/t12/13!
Posted: Wed Sep 12, 2007 8:12 pm
Hi!
I tried to look and compare the datasheets for HVSP and I found, that the HVSP algorithm differs for ATtiny45(etc) and ATtiny11/12 and for ATtiny13! It seems, that in order to get the device's singature into the latch one must pulse at least 6x (t13) or 4x (t11/t12) the SCI clock before bringing the RESET into 12V (see the datasheets for more details).
I would be glad to help with debugging any new firmware, but ATM I only have tiny11 and tiny12 to try.
(BTW, maybe it wouldn't be bad idea to add the possibility to "debug" the 12V wiritng by extending the firmware to listen for a specific vendor command that would put the HVSP reset line into 0V/5V/12V on user request. It would be easier to see whether something with the analog circuit is wrong or there is another - software - problem..)
And Yes, Christian, do You have plans to support PP in some future version? I have seen some code in the hvprog.c. Do you think ATmega16 (or compatible) would be enough for a complete ICP/HVPP/HVSP programmer?
(And Yet another note, I have look at the specification of HVPP initialization for M8 and M16 and seems really strange: one has to "clear" some fuses if for example the RESET bin is disabled BEFORE starting HV programming :-((((( )
I tried to look and compare the datasheets for HVSP and I found, that the HVSP algorithm differs for ATtiny45(etc) and ATtiny11/12 and for ATtiny13! It seems, that in order to get the device's singature into the latch one must pulse at least 6x (t13) or 4x (t11/t12) the SCI clock before bringing the RESET into 12V (see the datasheets for more details).
I would be glad to help with debugging any new firmware, but ATM I only have tiny11 and tiny12 to try.
(BTW, maybe it wouldn't be bad idea to add the possibility to "debug" the 12V wiritng by extending the firmware to listen for a specific vendor command that would put the HVSP reset line into 0V/5V/12V on user request. It would be easier to see whether something with the analog circuit is wrong or there is another - software - problem..)
And Yes, Christian, do You have plans to support PP in some future version? I have seen some code in the hvprog.c. Do you think ATmega16 (or compatible) would be enough for a complete ICP/HVPP/HVSP programmer?
(And Yet another note, I have look at the specification of HVPP initialization for M8 and M16 and seems really strange: one has to "clear" some fuses if for example the RESET bin is disabled BEFORE starting HV programming :-((((( )