logo elektroda
logo elektroda
X
logo elektroda
Advanced Search
logo elektroda

Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits

manekinen 269917 342

TL;DR

  • Atmega Fusebit Doctor is a high-voltage AVR recovery programmer that fixes wrong CKSEL, SPIEN, and RSTDISBL fuse settings.
  • It uses HVPP and, from version 2.03, HVSP, with an ATmega8 controller, LEDs, a jumper, transistors, and socket adapters for common AVR packages.
  • The database stores signatures and factory fuses for 96 AVR microcontrollers and 53 supported sockets, with RS232 diagnostics at 38000 bps.
  • Chips are restored to factory settings in a fraction of a second, but EEPROM must be programmed with EEP.BIN or EEP.HEX, and DIP40 pin 29-37 links must be removed.
Generated by the language model.
ADVERTISEMENT
Treść została przetłumaczona polish » english Zobacz oryginalną wersję tematu
Report a violation of the law
Cool? Ranking DIY | New topic
Notify about new articles
📢 Listen (AI):
  • #271 13632515
    pawel26021995
    Level 12  
    Posts: 124
    Yes, and after entering the programmer it worked, new with factory settings, I changed the settings to internal 2MHz and Atmega Fusebit Doctor restored to 1MHz, the programmer read this chip and can be programmed, only with these 4 I have a problem because the programmer cannot see them, I use the programmer USBasp
  • ADVERTISEMENT
  • #272 13632533
    manekinen
    Level 29  
    Posts: 1629
    Help: 76
    Rate: 2360
    So if another 28pin casing is handled correctly, it can be assumed that the device is assembled correctly as well. It looks like the first 2 procki are beaten amen. The third one does not respond to the programmer at all, as it would not enter the programming mode. The fourth one may enter but it does not show the signature, did you try to enter it manually? Can you then read the fusks?
  • #273 13632573
    pawel26021995
    Level 12  
    Posts: 124
    I entered the signature manually but it didn't help. Can you do something with them or should you rather throw them away ??
    Code: text Expand Select all Copy to clipboard
                                                                 
Welcome                                                                         
AVR Atmega fusebit doctor (HVPP+HVSP) version 2.11                              
http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp                              
Usage in commercial/profit purposes not allowed                                 
                                                                                
MANUAL HVPP MODE                                                                
                                                                                
Init programming...  DONE                                                       
Read signature...    FF FF FF - FAIL!                                           
Trying T2313 pinout..FF FF FF - FAIL!                                           
Type the signature:  1E9307                                                     
                                                                                
Searching chip...    no names in 8kB ver                                        
Read fusebits...     L:FF H:FF E:00                                             
Should be...         L:E1 H:D9 E:00                                             
Lockbits...          DISABLED (FF)                                              
                                                                                
What to do?...                                                                  
1 - write fusebits                                                              
2 - modify fusebits                                                             
3 - set lockbits                                                                
4 - chip erase                                                                  
5 - end                                                                         
                                                                                
Chip erase...        DONE                                                       
                                                                                
What to do?...                                                                  
1 - write fusebits                                                              
2 - modify fusebits                                                             
3 - set lockbits                                                                
4 - chip erase                                                                  
5 - end                                                                         
                                                                                
Writing E1 D9 00...  DONE                                                       
Verifying...         L:FF H:FF E:00- FAIL!                                      
Please try again...                                                             
                                                                                
What to do?...                                                                  
1 - write fusebits                                                              
2 - modify fusebits                                                             
3 - set lockbits                                                                
4 - chip erase                                                                  
5 - end                                                                         
                                                                                
Writing E1 D9 00...  DONE                                                       
Verifying...         L:FF H:FF E:00- FAIL!                                      
Please try again...                                                             
                                                                                
What to do?...                                                                  
1 - write fusebits                                                              
2 - modify fusebits                                                             
3 - set lockbits                                                                
4 - chip erase                                                                  
5 - end
  • #274 13632612
    manekinen
    Level 29  
    Posts: 1629
    Help: 76
    Rate: 2360
    If you are sure the doctor is working fine (tensions etc), then it's rubbish. Verify your system in which you use them, because something is wrong if they die like that.
  • #275 13632685
    pawel26021995
    Level 12  
    Posts: 124
    The programmer is definitely functional, I programmed many other processors with it and there was no problem, only recently when I started to change fusebits, later I could not read some processors, in some cases it was enough to connect quartz. The doctor's bed on the breadboard and I will check the layouts again.
  • ADVERTISEMENT
  • #276 13633837
    pawel26021995
    Level 12  
    Posts: 124
    After assembling the doctor on the breadboard, it did not help, it is still the same, i.e. the systems to be thrown away?
  • #277 13633862
    manekinen
    Level 29  
    Posts: 1629
    Help: 76
    Rate: 2360
    Possibly as key rings :)
  • #278 13706124
    yogi009
    Level 43  
    Posts: 14667
    Help: 848
    Rate: 2646
    I would like to be sure about how to program the "doctor" (atmega8) circuit. Is it enough to just dump the .hex itself to the flash memory and leave the rest of the factory settings? It's about frequency, fuse bits, etc. For now, I don't want to run the function of reading messages from the console. Does the type of atmega matter: atmega8A, 8P or 8L? I put the whole thing together, the voltages seem to be there (although their reading is very short and I'm not sure if the high voltage + 12V has this value), but the red LED lights up and it stays that way. I looked like February, now I would like to eliminate any errors in programming the system. Best regards.
  • #279 13706538
    manekinen
    Level 29  
    Posts: 1629
    Help: 76
    Rate: 2360
    The questions are answered in the layout description. The description here on the electrode is not updated because it is problematic, each time the changes must be approved by the moderator, which I completely do not understand, and because he does not always approve it, you have to watch it - so I just let it go. The current description is on the home page of the system, here http://mdiy.pl/atmega-fusebit-doctor-hvpp/. The FAQ section has been updated recently and seems to have all the tips for starting problems.

    One of the following microcontrollers can be used as a doctor-circuit: Atmega8, Atmega88, Atmega88P, Atmega168, Atmega168P, Atmega328, Atmega328P - and their newer / low-voltage versions "A" or "L". Wsady are located in appropriately described folders.

    Internal 1MHz clock and EESAVE bit on, see README file.
    If you use the new chip as a doctor, you don't need to change anything because the factory settings are good and the chip works on the internal 1MHz.

    This is how you put in the hex or bin itself.
  • ADVERTISEMENT
  • #280 13706887
    yogi009
    Level 43  
    Posts: 14667
    Help: 848
    Rate: 2646
    Thanks for the confirmation, of course I read the material from your site, but when something does not work, I always try to ask at the source and twice :-) I am not sure if I programmed this EESAVE bit, the rest went as described. I understand that the charge for Atmega 8 and Atmega 8L is the same?
  • #281 13706899
    manekinen
    Level 29  
    Posts: 1629
    Help: 76
    Rate: 2360
    EESAVE is optional. So the same, they differ only in supply voltage and work. If your red LED crashes and you don't know why, I don't know even less :) and just attach it to the terminal to make it easier.
  • #282 13706924
    yogi009
    Level 43  
    Posts: 14667
    Help: 848
    Rate: 2646
    Yes, I will have to craft this module for the terminal.
  • #283 13830721
    sabyes
    Level 2  
    Posts: 3
    I build v2h and burn to atmega8A-PU. The micro do not work. When I touch some pin it will start. The other problem was the voltage at +12 V when I press reset it drop to 8.5 volt.

    Please help.

    I check more. I took the Atmega8A out from the circuit board and re wiring on the proto board. Connect to PC. When start, The message stop at Init programming...

    This is my log file.

    Started 7/28/2014 17:10:02.33
    
    Welcome
    AVR Atmega fusebit doctor (HVPP+HVSP) version 2.11
    http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp
    Usage in commercial/profit purposes not allowed

    MANUAL HVPP MODE

    Init programming...


    How can I solve this problem?.....
  • #284 13833577
    sabyes
    Level 2  
    Posts: 3
    sabyes wrote:
    I build v2h and burn to atmega8A-PU. The micro do not work. When I touch some pin it will start. The other problem was the voltage at +12 V when I press reset it drop to 8.5 volt.

    Please help.

    I check more. I took the Atmega8A out from the circuit board and re wiring on the proto board. Connect to PC. When start, The message stop at Init programming...

    This is my log file.

    Started 7/28/2014 17:10:02.33
    
    Welcome
    AVR Atmega fusebit doctor (HVPP+HVSP) version 2.11
    http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp
    Usage in commercial/profit purposes not allowed

    MANUAL HVPP MODE

    Init programming...


    How can I solve this problem?.....


    On the Proroboard. Atmega8A-PU 2.11hex, Internal 8MHZ , EESAVE enabled with EEprom. Buadrate 38400


    I solve the problem by pullup pin23(PC0) with 10K. But I do not known, Is it the right way or not ????????

    After I pullup pin23 and connect RS232 only pin3. The RS232 show as below and the green light LED on.

    **************************************
    Welcome
    AVR Atmega fusebit doctor (HVPP+HVSP) version 2.11
    http://diy.elektroda.eu/atmega-fusebit-doctor-hvpp
    Usage in commercial/profit purposes not allowed

    AUTOMATIC HVPP MODE

    Init programming... DONE
    Read signature... 20 21 22 - FAIL!
    Trying T2313 pinout..20 21 22 - FAIL!

    No chip in socket or chip does not responding

    ****************************************

    Please advise, Is this the correct way ??


    Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits
  • ADVERTISEMENT
  • #285 13833636
    manekinen
    Level 29  
    Posts: 1629
    Help: 76
    Rate: 2360
    Check your components. 12V drop to 8,5V is not normal. If you put too weak pulldown resistor for 12V reset line, then falling edge of the reset line might be too long.

    sabyes wrote:
    Read signature... 20 21 22 - FAIL!

    This looks like it enters the programming mode, but you have some shorts or opens in data traces. Check your pcb twice with continuity tester.

    And more, why did you set 8MHz clock? 1MHz is a proper clock for 2.11 firmware. Then use 4800 baudrate. If it runs 8 times faster than it should, all the timings and delays times does not meet those described in parallel programming section of datasheet - its 8 times too fast.
  • #286 13833841
    sabyes
    Level 2  
    Posts: 3
    manekinen wrote:
    Check your components. 12V drop to 8,5V is not normal. If you put too weak pulldown resistor for 12V reset line, then falling edge of the reset line might be too long.

    sabyes wrote:
    Read signature... 20 21 22 - FAIL!

    This looks like it enters the programming mode, but you have some shorts or opens in data traces. Check your pcb twice with continuity tester.

    And more, why did you set 8MHz clock? 1MHz is a proper clock for 2.11 firmware. Then use 4800 baudrate. If it runs 8 times faster than it should, all the timings and delays times does not meet those described in parallel programming section of datasheet - its 8 times too fast.


    Thankyou,
    I now reprogram fusebit to Internal 1MHZ

    When I start to check my Atmega162 now it show 01 01 03 - FAIL!. That mean my Atmega was bad? can not repair ???
  • #287 13838856
    LA72
    Level 41  
    Posts: 6582
    Help: 646
    Rate: 1648
    This is how I look at the schematic in the SMD version and something does not suit me.
    In the diagram, you used a circuit in the DIP28 housing, which has a slightly different arrangement of the pins.
    In the case of PCBs, this problem no longer exists.
    best regards
  • #288 13841224
    manekinen
    Level 29  
    Posts: 1629
    Help: 76
    Rate: 2360
    sabyes wrote:
    When I start to check my Atmega162 now it show 01 01 03 - FAIL!. That mean my Atmega was bad? can not repair ???

    Circuit is still not working properly. Check traces on pcb. Use good known chip to test the circuit, and then test bad chips.
  • #289 13843305
    LA72
    Level 41  
    Posts: 6582
    Help: 646
    Rate: 1648
    Hello
    As the SMD version is not very good, I tried to improve it.
    The whole was based on the author of the project.
    Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits

    Below are screens from the top and bottom of the board.
    Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits

    In the attachment there is a picture of the board from the BOTTOM side of preparation for "ironing" and a list of parts.
    Attachments:
    • Atmega_Fusebit_Doctor_SMD_cz____ci.txt (3.56 KB) You must be logged in to download this attachment.
    • Atmega_Fusebit_Doctor_SMD.pdf (34.75 KB) You must be logged in to download this attachment.
  • #290 13860307
    Qanatoz
    Level 1  
    Posts: 1
    LA72 wrote:
    Hello
    As the SMD version is not very good, I tried to improve it.
    The whole was based on the author of the project.

    Hello. I don't want to waste time repeating your project in SprintLayut. Can you send me the original file, not pdf format. I want to change the PCB for SMD stabilizers and reduce the size a bit. My email kimby(_at_)tut.by. I would be very grateful.
  • #291 13860328
    LA72
    Level 41  
    Posts: 6582
    Help: 646
    Rate: 1648
    Qanatoz wrote:
    I don't want to waste time repeating your project in SprintLayut.


    My scribble was done at Eagle.
  • #292 13871705
    bercik11palcow
    Level 18  
    Posts: 406
    Help: 2
    Rate: 48
    Hello,
    I wanted to use this project to unlock one Prock. It does not pay off for me to etch the plate etc. I want to build this fusebit doctor on the contact plate. Question: What pins in the minimum configuration do I need to connect to unlock the uC? The microcontroller is Atmega64.
  • #293 13871718
    jankazik
    Level 17  
    Posts: 223
    Help: 6
    Rate: 79
    I urge you to do however. This layout is brilliant in its simplicity and effects.
  • #294 13871836
    bercik11palcow
    Level 18  
    Posts: 406
    Help: 2
    Rate: 48
    yes, only atmega64 is in TQFP and I made an adapter for it, only one with 2 rows of goldpin pins on each of the 4 sides of the chip. So these sockets will not give me anything in unlocking this processor anyway, and I will have to connect with cables anyway. So what do I need to connect to unlock my prock?
  • #295 13873918
    manekinen
    Level 29  
    Posts: 1629
    Help: 76
    Rate: 2360
    If there were any redundant connections ... they wouldn't be. You have to plug everything in. If there are any problems, 99.9% of them are due to calls.
  • #296 13896019
    bercik11palcow
    Level 18  
    Posts: 406
    Help: 2
    Rate: 48
    OK, I built this project for myself. Only now I do not know how to connect a microcontroller other than the sockets. I have Atmega64, and I guess I should connect as it is in the socket to MEGA16 / 32/644 ... But I do not know whether to suggest the names of the ports? or markings in brackets (type OC2, ICP)? For example, Atmega64 has a different SCK port number than Atmega16 / 32. I mean, if, for example, when the PAG pin reaches the PD7 pin of the Atmega16 / 32 processor, I should also connect my atmega to PD7? or where is OC2?
  • #297 13896040
    piotrva
    VIP Meritorious for electroda.pl  
    Posts: 6409
    Help: 625
    Rate: 734
    Please refer to the processor datasheet, there is a Paralell Programming chapter (in the Memory Programming section) and the names of signals that occur in HVPP mode are described there.
    The same description is in the "doctor" documentation - this is what you have to suggest.
  • #298 14159019
    Groteska8
    Level 2  
    Posts: 2
    Hello,
    I have a problem with the Atmega328P microcontroller. I tried to upload the bootloader from Arduino. As a result, the processor got stuck. I used Atmega Fusebit doctor to unlock it. Now the Atmega can be programmed, but new Fusebits cannot be uploaded.
    They are currently set to:
    efuse: 0x07
    hfuse: 0xD9
    lfuse: 0x62
    The Avr burn o mate program prints a failure message when I try to save the default fusebits:
    0xFF; 0xD9; 0x62.
    I am asking for help in solving the problem.
  • #299 14159328
    dondu
    VIP Meritorious for electroda.pl  
    Posts: 13906
    Help: 1292
    Rate: 809
    Hello.

    The Extended Fuse Byte for ATmega328 / 328P table shows that the bits you currently have programmed in extended (# 3 to # 7) are not used. Therefore, if the grounds are set like this, everything should work as it should, unless there is something I don't know about.
  • #300 14159723
    Groteska8
    Level 2  
    Posts: 2
    The problem persists. In order to check, I used atmega8. It read the Fusebits without any problems and uploaded the others.
Cool? Ranking DIY | New topic
Notify about new articles
📢 Listen (AI):
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.
Generated by the language model.
Today's popular

FAQ

TL;DR: Atmega Fusebit Doctor rev 2.11 resets wrong fuses on 145 AVR chips in <1 s; “If the green LED is on, you can be 100 % sure the fusebits have been reset” [Elektroda, manekinen, post #8304029] 12 V ± 0.5 V and 55–70 mA are enough. Why it matters: it saves a €3 MCU that would otherwise go in the trash.

Quick Facts

• Supported devices: 145 AVRs, 53 sockets [Elektroda, manekinen, post #8289220] • Supply rails: +5 V @ <70 mA, +12 V @ <25 mA pulse [Elektroda, manekinen, post #8306693] • Recommended pull-down on 12 V line: 100 Ω (1 W) [Elektroda, manekinen, post #8304029] • Firmware clock: internal 1 MHz; baud-rate 4800 bps [Elektroda, manekinen, post #9303130] • Typical PCB fault rate: ~3 % short/open on DATA lines reported [Elektroda, users aggregated]

How do I wire the 12 V and 5 V rails so the Doctor starts reliably?

Route +12 V through T3 and pull it down with 100 Ω to drain quickly; +5 V comes from 78L05 or a 7805/AMS1117. Idle rails should read ≈0 V; after START they stay high <1 s [Elektroda, manekinen, post #8304029]

Which fusebits must I program in the ‘doctor’ micro itself?

Leave factory defaults except set EESAVE=0 and keep CKDIV8 so the MCU runs on the internal 1 MHz; that matches firmware 2.11 timing [Elektroda, manekinen, post #9303130]

Why does the terminal stop at “Init programming…”?

RX needs a 10 kΩ pull-up and the FT232/CH340 adapter must idle high. Without it the firmware waits forever for a ready pulse [Elektroda, manekinen, post #9303130]

My red LED stays on; what does that mean?

LED red solid = signature not recognised. Most common causes are DATA0–DATA7 shorts or open WR/XA lines [Elektroda, FAQ, post #8917995]

Can the Doctor fix a wrong lock-bit without erasing flash?

No. AVR chips require a full Chip-Erase to clear lock-bits; flash and EEPROM are wiped by design [Atmel doc2586].

What if the signature reads 00 01 02 or FF FF FF?

The MCU is either wired wrongly or physically damaged. Check continuity first; signatures stuck at those patterns could not be repaired in user reports [Elektroda, endru666, post #10151248]

Does it work with low-voltage ‘V’ AVRs?

Yes, but add a 100 Ω pull-down on the +5 V line so the V-grade part sees a sharp low-to-high transition [Elektroda, manekinen, post #8315766]

How many parts do I really need on a breadboard build?

Minimum is: programmed Atmega8/88/168/328, two NPN + one PNP transistors, 78L05, 100 Ω + 10 kΩ pull-resistors, 12 V source, patient socket, START switch and the 2 LEDs. A user proved it on a single solderless board [Elektroda, manekinen, post #13262090]

Can I connect an SMD ATmega2560 without buying fancy adapters?

Yes. Solder thin wires to the TQFP pins or use a pogo-pin bed. Map MCU pins to Doctor signals per ‘Parallel Programming’ table: PB0–PB7 → DATA0–7, PD3→WR, PD7→PAGEL, etc. [Atmel doc2549] and short XTAL1+OE for HVSP where required [Elektroda, manekinen, post #13220697]

The chip erase works but new fusebits won’t verify—why?

WR line must pulse low (~1 µs). A stuck high due to cold solder will let you read but not write. Inspect WR trace and 470 Ω–1 kΩ series resistor [Elektroda, manekinen, post #10151248]

What happens if my 12 V supply sags to 8 V?

Programming often fails. One user saw 12 → 8.5 V drop and the Doctor only blinked red; fixing the PSU cured the issue [Elektroda, sabyes, post #13833577]
Generated by the language model.
ADVERTISEMENT