logo elektroda
logo elektroda
X
logo elektroda
Advanced Search
logo elektroda

Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits

manekinen 260986 342
ADVERTISEMENT
Treść została przetłumaczona polish » english Zobacz oryginalną wersję tematu
Cool? Ranking DIY | New topic
Notify about new articles
  • #91 9039615
    manekinen
    Level 29  
    Is this system working properly? Because these settings mean an external 3-8MHz RC oscillator. After adding such an oscillator, does the system work and can it be programmed? Because in my opinion, the system is damaged, it sometimes happens that the coffee grounds freeze and you cannot change them - I have such systems myself ... the code is performed but neither flash nor coffee grounds can be changed :(

    Of course the fusebit doctor works fine, especially if the atmega16 has also been recognized.

    I will paste here a small upgrade to the board that I was supposed to add with the program update, but it probably won't happen too quickly, so ... an additional 100ohm resistor pulling the + 5V line improves a lot of problems with the system ... the whole thing works more stable with low-voltage systems, it allows incl. repair of RSTDISBL fusebit in attiny2313 ( MORE )

    Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits
  • ADVERTISEMENT
  • #92 9093885
    1 Marcin
    Level 17  
    Hello, I submitted a DOCTOR but unfortunately something does not really want to work, i.e. I programmed the Atmega with the last version of hex and I do not recognize a working atmega 8. I managed to unlock attiny2313 but it is probably by accident that after putting the prock back to the doctor, it does not recognize it. I noticed that during the resuscitation of Prock, when he touches any path, including the mass, he sows bushes onto the terminal. After inserting a working atmega 8, it shows a different processor signature every time
    Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits
    Maybe someone has any ideas what is wrong with the plate I made according to. the original, so theoretically it should work. Marcin
  • #93 9107240
    manekinen
    Level 29  
    FAQ https://www.elektroda.pl/rtvforum/topic1716521-60.html#8917995

    You must have some short circuits on the board, possibly cold February. Check but CAREFULLY. The rest approx.


    UPDATE # 7 - 23.01.2011


    PCB ver V2e - additional resistor pulling the line + 5V.
    Program ver 2.07 - critical fixes for 8 procks, 3 new ones added.

    After adding a resistor, the system will be more stable, this fixes the problem when some systems initialized in non-high voltage mode and the RSTDISBL fuska could not be repaired (including Attiny2313)

    CRITICAL AMENDMENTS for AT90s1200, AT90s2313, AT90s2323, AT90s2343, AT90s4414, AT90s4434, AT90s8515, AT90s8535
    Fixed factory fuska value for AT90s2313 - short start-up time is disabled by default
    New chips have been added: Atmega406, Atmega16HVB and Atmega32HB. They are not listed, they are only supported theoretically. For more information, please refer to the documentation on programming "battery magament avr's" chipsets.
    Fixed bug in RS232 messages

    There is also a version SMD this circuit, based on pcb v2d. Here http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp/#update6pl
    Attachments:
  • #94 9108009
    1 Marcin
    Level 17  
    Quote:
    You must have some short circuits on the board, possibly cold February. Check but CAREFULLY. The rest approx.
    I checked the plate and it turned out that one path was interrupted, now it is nice - I unlocked 2 pieces of attina 2313.
    I am full of respect for the author of the project. Marcin.
  • #95 9198300
    skwarag
    Level 12  
    Hello everyone.
    First of all, many thanks to the authors of the project.

    I tried to revive the atmega 644, but the doctor replies that he cannot read the signature. Does this mean that the system is completely damaged?
    I will add that atmega died during programming - mkAVR crashed, after 10 minutes I disconnected everything and that's it ...

    Maybe some suggestions.
    Best regards.
  • #96 9200595
    manekinen
    Level 29  
    What signature does the chip show? Check all connections to this atmega well, check on another processor whether the system is working properly.
  • #97 9203140
    skwarag
    Level 12  
    All right!!!

    I checked the connections as you suggested and ... during soldering I burned 3 pads under the resistors in the data line. As a result, there was no connection (visually, nothing was visible, yet ...).
    I fired up the "miracle worker" without a problem, read a good mega32 proc, and then unlocked the blocked mega644 without any problems - and I thought that PLN 25 would go to the basket.
    A revelation for me !!!
    Thanks a lot.
    When you are in Krosno, I invite you for a beer, hehe.
    best regards
  • ADVERTISEMENT
  • #98 9203253
    Zbyszek_64
    Level 24  
    Hello.
    I ran the doctor on batch 2.05 and unlocked 2 eights. The first one accepted the settings without any problems, the second one was supported by the ALLOW ERASE jumper. I have a problem with the atmega 32. The green diode blinks briefly and it lights up red. while programming it. Accidentally touched one of the pins and felt a tingling sensation. After measuring the housing against ground, it came out of 98V. Now I grounded the housing. I do not know if it makes sense to do anything with it?
    Many thanks to the author.
  • ADVERTISEMENT
  • #99 9203761
    manekinen
    Level 29  
    Oh, it's probably dead slingshot. Though it would be nice to see what the doctor sends to the terminal.

    I plan in the next update (2.08) to add dunks for the younger atmegi8 sisters, backward compatible. The systems I have associated are: Atmega88, Atmega88P, Atmega168, Atmega168P, Atmega328, Atmega328P. Are there any more? Versions with 16KB memory and above will send the full name to the terminal.

    Is there anyone who could test the operation of these inputs? It is about the above. procki. If there is anyone willing, I invite you to PW.
  • #100 9229668
    MatygaD
    Level 15  
    manekinen wrote:
    Oh, it's probably dead slingshot. Though it would be nice to see what the doctor sends to the terminal.

    I plan in the next update (2.08) to add dunks for the younger atmegi8 sisters, backward compatible. The systems I have associated are: Atmega88, Atmega88P, Atmega168, Atmega168P, Atmega328, Atmega328P. Are there any more? Versions with 16KB memory and above will send the full name to the terminal.

    Is there anyone who could test the operation of these inputs? It is about the above. procki. If there is anyone willing, I invite you to PW.


    Hello

    I can test on ATMEGA168 I have loose and a few other procks for testing from Tiny13, 24, 44, 45, to Mega 8,16,168, 32, 8535

    Regards
    Darek
  • #101 9229684
    Zbyszek_64
    Level 24  
    Another problem - Atmega 32 unused red diode turns on and on all the time. The programmer recognizes the processor. The same is with the blocked one, except that the programmer does not recognize the mc. On the Atmega 8 everything is fine. I checked all the connections on the board, no has a short circuit. What could be the reason for this doctor's behavior?
  • #102 9231153
    manekinen
    Level 29  
    And have you removed the metal connectors from the 29th to the 37th pin from the atmegi32 socket, as I wrote in the description? There is a photo. If so, and the processor is still not recognized, then there really must be a bull on the board because this model works 100% with the doc.
  • ADVERTISEMENT
  • #103 9237994
    manekinen
    Level 29  
    Zbyszek_64 you do not write back if it helped, but you gave me a few points so I understand it's probably it :)

    MatygaD you can test if you have time and feel like it :)

    UPDATE # 8 - March 5, 2011

    Fixed factory fusebit values for Attiny28, Atmega161, Atmega163.
    After the last fix, AT90s2313 was still incorrectly handled - fixed

    From now on you can use the following (m8 backwards compatible) chips as a doctor :
    Atmega88, Atmega88P, Atmega168, Atmega168P, Atmega328, Atmega328P.
    Wsady for 16kB and 32kB chips send patient names after uarcie (as in ver.2.04)

    This is probably the last update for the atmega8 / atmega88. I plan to add the TPI protocol (tiny programming interface), add prock support with such an interface, and make an adapter for them. Currently, there are literally a few bytes of memory space left, so a larger processor will be needed for the project to develop further. Also 8kB memory chips will not be supported any longer.

    Information on any updates to my layouts here - twitter.com/manekinen
    Attachments:
  • #104 9247608
    MatygaD
    Level 15  
    Hello

    ATMEGA168V found a place in the doctor instead of the current MEGA8.
    I checked the fusebit deletion from mega 16, 32, 168, Tiny 13 and 25 everything works flawlessly, and again spits out the chip type on the terminal :)

    Regards
    Darek
  • #105 9248023
    piotrva
    VIP Meritorious for electroda.pl  
    Such questions and comments:
    1.Can I replace the transistors with the bc556 and bc548 models?
    2. with new processors (168 and 328) AVCC should be connected
  • #106 9248245
    manekinen
    Level 29  
    MatygaD - thanks so much for checking, I couldn't do it myself.

    piotrva
    1. I see no contraindications, plump and so they will do their job.
    2. Shot in 10! In fact, the AVCC is not connected, and it supplies the RDY, BS1, XA0, and XA1 lines - three of them work bidirectionally, so I do not know how they can set a logical 1 so strong that the LEDs light up? Tell me if you have already started the system? And on what layout?
    AVCC to be corrected, of course, I will try to cram some connection to pcb.

    added ...

    In the attachment, a V2f plate made quickly, AVCC connected with a jumper. If there are any errors, please write :)
    Attachments:
  • #107 9252256
    piotrva
    VIP Meritorious for electroda.pl  
    I made the layout on the uC atmega168, uploaded the software (hex file via avrstudio4, then bin via bascom) and ... nothing
    I connect to rs232 to max232 and I get the following bushes:
    Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits
    could this be due to wrong value in the osccal register?
    because this mega168 once "programmed" to me as mega8, because I mistook the boxes and then resurrected her avr dragonem
    The fusebits are programmed well
    Edit:
    I don't know what was wrong, but I think my rs232 cable broke somewhere, because the messages appear correctly on the ftdi adapter :D
    EDIT2:
    I don't know what is still wrong, but the program stops at INIT programming ... and only lightly touching the pins with your finger causes the program to go on ... and no chips or sockets are connected to the processor. Nothing changes after connecting attina13 ... in total I get this message:
    Code: text Expand Select all Copy to clipboard
    
Welcome
AVR Atmega fusebit doctor (HVPP+HVSP) version 2.08
http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp
Usage in commercial/profit purposes not allowed
for firmware updates: twitter.com/manekinen

HVPP MODE

Init programming...  DONE
Read signature...    00 01 02 - FAIL!
Trying T2313 pinout..00 01 02 - FAIL!

No chip in socket or chip do not responding
Please try again...

Thank You


    EDIT3:
    probably something wrong with this m168 ... unfortunately I only have one, but I will check it out :D on the m8 it is ok for now with the attiny13 processor
    if I have time, I will block the m2560 with a dragon :D
    EDIT4:
    and yet it works with tiny13 ...
    So far I have tested:
    attiny13-OK
    attiny85-OK
    90s2313-OK (admittedly it displays FAIL at the beginning and then Trying T2313 pinnout, but then programs the fuses with default values, although this processor probably cannot be blocked :D )
    EDIT5:
    attiny2313-OK
    atmega2560-OK (I tested on 3 pieces, 1 on the board-OK, 2 in the socket-OK, 3, overheated during hot soldering, as expected it does not respond)
    Code: text Expand Select all Copy to clipboard
    
Welcome
AVR Atmega fusebit doctor (HVPP+HVSP) version 2.08
http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp
Usage in commercial/profit purposes not allowed
for firmware updates: twitter.com/manekinen

HVPP MODE

Init programming...  DONE
Read signature...    1E 98 01 
Searching chip...    Atmega2560
Chip erase...        DONE
Read fusebits...     L:E2 H:D9 E:FF
Should be...         L:62 H:99 E:FF
Lockbits...          DISABLED (FF)

Rewriting 6299FF...  DONE
Verifying...         L:62 H:99 E:FF - OK!


15 fixed chips

Thank You

    so you can confirm operation for these processors

    EDIT6:
    Photo of my equipment (mini doctor that can be used as a test board for atmega8 / 168/328 processors and tqfp-100 socket):
    Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits

    EDIT7:
    I only miss one thing (I have not yet implemented it in my version), namely the pull-up resistor on the reset line. sometimes when I touch the plate from the bottom, the processor resets and starts working, which is not desired
  • #108 9257013
    manekinen
    Level 29  
    So maybe one by one :)

    As for the tea leaves, I quietly changed the attachment on the second day. Small changes in the read.txt file, added fuses settings for m88, m168, m328 - values from m8 will not work.

    edit2 ... With attiny13 you need an hvsp adapter, on your board you need to shorten the appropriate pins so that the doctor "detects" it and works in hvsp mode. Without it, it works in hvpp mode, and after init is done, programming expects high on pin RDY - and possibly reads it when touched with a finger. The hvsp adapter bypasses this expectation, and inserting the hvpp prock will force the correct (steady) state to RDY.

    edit4 ... Prints FAIL because it is a 20-pin processor, there is a different pinout for them. If a valid signature is not received, the chip tries to pinout according to T2313 (hence the name). Normal operation. And the AT90S2313 can be blocked as much as possible, it has a SPIEN fusebit.

    So it works on the M168? (because in the log I can see the names in the terminal). Was the reset line the problem?

    Thanks for the tests, when I am at the house (Friday) I will send you a small gift at PW, for you MatygaD also :)
  • #109 9257328
    piotrva
    VIP Meritorious for electroda.pl  
    Yes, indeed I forgot to jumper hvsp the first time, but then I figured it out. It works with this mega168, but it was a problem with the rs232 cable, otherwise everything ok. Tomorrow I will do some tests of this megi168, because I still do not like the behavior with an empty stand in hvpp mode - sometimes you have to wait longer for info about the lack of a processor . Besides, everything works fine. For the sake of peace, I can still check m1280, tiny26 461 tomorrow.

    As for the reset, I write about it because I know from the autopsy that without a resistor, when touching certain connectors with your fingers, you can even damage the batch in the up.
  • #110 9260010
    manekinen
    Level 29  
    piotrva wrote:
    because something still does not like the behavior with an empty socket in hvpp mode - sometimes you have to wait longer for info about the lack of a processor.

    But I wrote above. There is a transient on the RDY line because this input is not internally pulled up. The system waits for the patient to be ready ... indefinitely, or until we touch this line with our finger. This is normal behavior without the chip in the cradle.

    All in all, I can pull it up, because the uninitialized or busy system (patient) will probably bring it to the masses anyway - I will have to check if there will be any conflicts, because this line is important when saving data and we would not want to send new data until the previous data is saved is not over.
  • #111 9260030
    piotrva
    VIP Meritorious for electroda.pl  
    Then there are no problems, I did not read the earlier post with understanding. Today I did not have some time to test these processors, which I mention in an earlier post. But by the evening maybe I will do something else.
  • #112 9262067
    IMEX
    Level 20  
    Will it work for FTDI - I have the Original GSM BOX on USB RX, TX

    Unfortunately, I have Lapka without COM
  • #113 9262278
    piotrva
    VIP Meritorious for electroda.pl  
    It will work without any problems. You connect the common ground and the output from the doctor to the rx line of the ftdi system.
  • #114 9264309
    piotrva
    VIP Meritorious for electroda.pl  
    Subsequent tests completed:
    Code: text Expand Select all Copy to clipboard
    
Welcome
AVR Atmega fusebit doctor (HVPP+HVSP) version 2.08
http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp
Registered to Piotr Rzeszut (piotrva)
for firmware updates: twitter.com/manekinen

HVPP MODE

Init programming...  DONE
Read signature...    FF FF FF - FAIL!
Trying T2313 pinout..1E 92 08
Searching chip...    Attiny461/A
Chip erase...        DONE
Read fusebits...     L:62 H:DF E:FF
Should be...         L:62 H:DF E:FF
Lockbits...          DISABLED (FF)
Rewriting 62DFFF...  DONE
Verifying...         L:62 H:DF E:FF - OK!

21 fixed chips

Thank You



Welcome
AVR Atmega fusebit doctor (HVPP+HVSP) version 2.08
http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp
Registered to Piotr Rzeszut (piotrva)
for firmware updates: twitter.com/manekinen

HVPP MODE

Init programming...  DONE
Read signature...    1E 1E 1E - FAIL!
Trying T2313 pinout..1E 91 09
Searching chip...    Attiny26
Chip erase...        DONE
Read fusebits...     L:E1 H:F7 E:00
Should be...         L:E1 H:F7 E:00
Lockbits...          DISABLED (FF)
Rewriting E1F700...  DONE
Verifying...         L:E1 H:F7 E:00 - OK!

22 fixed chips

Thank You

    So attiny26 and attiny461 (v) can be added to the list as checked
    However, I have a few comments about the description itself: files from the sockets directory are unreadable for me, e.g. the one for tiny26 / 461 / ... does not make the adapter) and I, for example, used the adapter diagrams.
  • #115 9264849
    manekinen
    Level 29  
    Thanks for the next tests! In fact, I do not keep this list, i.e. I do not update it, but maybe I will put it all together and write it down somehow :)

    Procks with 20 pins have dual functions for two pins. They are combined XA1 + BS2 and PAGEL + BS1:

    Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits

    From the female goldpin connector -> 20pin procek should connect:
    BS1 -> BS1 / PAGEL
    BS2 -> XA1 / BS2

    What about the rest can be deduced from the diagram :)
    PD4 and PD6, this is just for attina2313.

    Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits
  • #116 9264959
    piotrva
    VIP Meritorious for electroda.pl  
    I know just about this, only I wrote that I had to use the diagram, because you did not write in your descriptions what this specific derivation is called in the doc. I added a pull-up on a reset by a 3k3 resistor and it does not reset when it fails.
    A list could be compiled. Besides, once again a big plus for a great project.
  • #117 9266983
    manekinen
    Level 29  
    Well, because these descriptions in the sockets directory only apply to a specific adapter :)

    I have skimmed the comments on the site, my inbox, and posts on this topic, here are some proven layouts:

    at90s8515, at90s1200, at90s2313
    mega8535, mega8515, mega8, mega16, mega32, mega88, mega48, mega168, mega328P, mega162, mega324P, mega644P, mega644, mega128, mega2561, mega1280, mega2560
    tiny2313, tiny25, tiny45, tiny85, tiny26, tiny13, tiny461

    There is not much of it, rather the most popular layouts. But if a chip - e.g. tiny461 - works, its brothers tiny261 and tiny861 MUST work too :)

    I am very curious about the tiny15 chip, no one has checked it yet and this misfit has a separate clock line that needs to be treated differently - the program only theoretically supports it :(
  • #118 9267060
    piotrva
    VIP Meritorious for electroda.pl  
    no atmega1280 and atmega2560 are missing on the list
    and I think that's why no one has checked tiny15, because hardly anyone buys it, because this processor has no frame at all ...

    Added after 5 [minutes]:

    when I do some bigger purchases, I can buy it in total, although programming this bone is a good gymnastics due to the lack of a framework :D
    ---
    and so on the topic, but from a different barrel, my colleague did not think about writing a parallel programmer to the ATMEL 8051 family based on the AVR processor and with a USB RS232 adapter? because in my wardrobe there are some procks from this family from disassembly and I would love to play with them. Maybe when I have a lot of free time, I will try to put together something like this myself
  • #119 9288530
    ejcon
    Level 14  
    Hello
    I have a problem with unlocking Atmega 32, when I start it, I get something like this, what's wrong with this system?
    HVPP MODE

    Welcome
    AVR Atmega fusebit doctor (HVPP + HVSP) version 2.09
    http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp
    Usage in commercial / profit purposes not allowed
    for firmware updates: twitter.com/manekinen

    HVPP MODE

    Init programming ... DONE
    Read signature ... 9E 95 82 - FAIL!
    Trying T2313 pinout..9E 95 82 - FAIL!

    No chip in socket or chip does not responding
    Please try again ...

    Thank You
  • #120 9288945
    manekinen
    Level 29  
    See the FAQ in one of my previous posts.
    After the signature, I conclude short circuits on the date lines.
    For questions like 'why doesn't work', please provide the PCB and software versions :)

    A new version of the software appeared on the project website - 2.09 - unfortunately I did not have time to paste it on the forum, but I will try to do it at the weekend :)
Cool? Ranking DIY | New topic
Notify about new articles
Report a violation of the law

Topic summary

The discussion revolves around the Atmega Fusebit Doctor, a device designed to repair misconfigured fuse bits in AVR microcontrollers, particularly the Atmega series. Users share their experiences with the device, detailing issues such as incorrect fuse settings, inability to read signatures, and challenges in programming various Atmega models. Solutions include using the device to reset fuse bits, ensuring proper connections, and troubleshooting with UART communication. The conversation highlights the importance of correct voltage levels, resistor values, and the need for careful assembly to avoid short circuits. Users also discuss the potential for using the device with different AVR models and the necessity of firmware updates for compatibility.
Summary generated by the language model.
ADVERTISEMENT