Atmega Fusebit Doctor (HVPP + HVSP) - fix fusebits

Hello,
I would like to know if 12V is necessary and does it have to be well filtered? 100mA is enough?

Best regards!

A voltage of 12V is necessary, 100mA should be enough

I made this layout from http://mdiy.pl/atmega-fusebit-doctor-hvpp/, and downloaded the updates "UPDATE # 11, 04/30/2011:" I uploaded soft, everything is nice, but when I put some processor into the system without a shorted jumper EESAVE, the system is not repaired, only the green LED will flash for a while and the red LED will start to light up, but when I put the jumper on, the green LED will light up once, not more often when the system is pressed to the socket, and when it lights up, the system is not only his signature fixed

avrdude: AVR device initialized and ready to accept instructions

Reading | ################################################### | 100% 0.02s

avrdude: Device signature = 0xc7c7c7
avrdude: Expected signature for ATMEGA8 is 1E 93 07
Double check chip, or use -F to override this check.

avrdude done. Thank you.

What can you advise me about this?

What you showed with AVRDude was reading Atmega8 as doctor or patient? Because, unfortunately, it is impossible to repair a damaged signature. Do you have access to RS232?

what I showed was the reading from the patient, but this atmega was quite functional, only after putting it in the doctor it broke no, unfortunately I can't access RS232

These broken signatures are black magic. If it was good before, it is possible that it could have broken down while playing with the doctor, solder / track breaks, and the aforementioned tightening of the chip while working. Check the board carefully because if there is an error somewhere, it's only on it. Short circuits, breaks, assembly. If you read what is spit out by the uart output, you would know something more.

In my opinion, somewhere a colleague has a cold February ...

I will check the whole plate tomorrow, I will measure everything from a to z, because today you know, Sunday and it does not fit a bit and I will improve February if they are any cold on the plate

EEVblog showed a plate for this project in

in the 22nd minute. There are minimal differences because someone converted the PCB into a double-sided version, but you can clearly see that it is the same project. He made it to Australia with Maiami.

I know, I've seen this tile somewhere on the internet before, but I haven't found anything else, not even a contact with the author of this tile. The tile is not converted into a double-sided one, just the print with the top jumpers was used as the top layer and that's it. You can see that the boy added the OSH logo (open source hardware / open hardware) but he doesn't know what it means. He just recently saw all the tiles have it and added. The fact that I shared the eagle project files does not mean that the board is open hardware license. This leads to serious misunderstandings.

I watch EEVblog from the very beginning and I was able to send even the assembled layout to it myself, but I don't have any pressure on glass or stupid surreptitious advertising - that's not the point.

Witam,

Staram się uruchomić ten układ. Mam wrrsje 2.11. Podczas komunikacji z komputerem mogę wymazać zawartość flash, odczytać fusebity oraz sygnaturę procesora, lecz nie mogę zapisać nowych fusebitów. Oto co pokazuje mi terminal:

Quote:

[1B][2J
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... 1E 93 07
Searching chip... no names in 8kB ver
Read fusebits... L:E3 H 9 E:00
Should be... L:E1 H 9 E:00
Lockbits... DISABLED (FF)

What to do?...
1 - write fusebits
2 - modify fusebits
3 - set lockbits
4 - chip erase
5 - end

1
Writing E1 D9 00... DONE
Verifying... L:E3 H 9 E:00- FAIL!
Please try again...

What to do?...
1 - write fusebits
2 - modify fusebits
3 - set lockbits
4 - chip erase
5 - end


What to do?...
1 - write fusebits
2 - modify fusebits
3 - set lockbits
4 - chip erase
5 - end

4
Chip erase... DONE

What to do?...
1 - write fusebits
2 - modify fusebits
3 - set lockbits
4 - chip erase
5 - end


What to do?...
1 - write fusebits
2 - modify fusebits
3 - set lockbits
4 - chip erase
5 - end

5
Thank You

Co może być źle?

First of all, the question is, after assembly, did you check the operation of the system with another prock?

If we assume that it works correctly (correct reading of the signature), then the fuss reads correctly. L: E3 H: D9 - These fuses don't block anything at all, they just change the internal oscillator from 1MHz to 4MHz - and the ISP programmer should read it nicely.

But why does not save tea leaves, you have two options:
1. Damaged slingshot
2. Doctor is not working properly (assembly)

As for wiping the memory, only the command is sent, after which the processor clears everything itself, and it is not checked in this case that the system is actually clean.

PS paste the code into [code] tags, it will not insert these smileys.

Welcome back,
I checked the system with several processors. However, the effect is always the same. The processors seem to be working, because I can read and change fusebits with the programmer without any problems. Tracks, connections, I checked February a million times. Though there is probably the cause of the problem somewhere.
What lines, possibly short circuits, can sneak into me in a situation such as described above? I mean, I can send a memory wipe command, read the signature and fusebits, I can communicate with the system, and the only thing and probably the most important thing that I am not able to do is to enter new fusebit values. Is it possible on the basis of these observations to bring the suspicious area closer?

The WR line is used for writing and not for reading. Normally it should be up and the write pulse is a brief pull to ground.

Added after 17 [minutes]:

Ie this pulse on WR lasts 1 us so you won't get it without an oscilloscope, just make sure it's not short-circuited.

I put together the fusebit doctor's plate with the adapter for ATMEGA8515, version from the first page of the topic (V2d). I set EESAVE, the internal oscillator to 8MHz, threw flash and eeprom (2.03) through mkAVR Calculator (the program probably knows in what order what to upload?). I want to unlock one ATMEGA8 and one ATMEGA8515.
After inserting the M8 into the socket and activating the system, the red LED flashes, then turns green. I press the button, the red blinks and after about 1 second the green one turns on. Same thing with M8515.
After connecting the programmer, I get messages like this on both bones:

I had the same messages before I submitted the doctor.
I was afraid that the dice were blocked with an amen (invalid signature), so I put a functional M8515 into the doctor. It behaved in the same way as its predecessors.
After inserting it into the system in which it worked and connecting the usbasp programmer, I get a message like the first damaged one.

With M8515 inserted in the female goldpin connector, between GND and 5V is 3.43V, when the button is pressed it jumps to 5V, and between GND and RES it is 0V and jumps to 12V.

I checked the board, there are no cold solders, no short circuits, all elements are new, not recycled.
I do not have the RS232 terminal yet, first I need to find out how to assemble it ;)

Can I ask for help? What did I do wrong?

Why did you make an old version of the layout with bugs? After all, in the first post there is a great color link to the current version. Dude, improve the board to the 2H version and load batch 2.11 - everything will be spinning :)

And the described behavior is a badly loaded eeprom, which in newer versions does not need to be loaded.

Actually, I don't know why I made the old version of the board
I understand that I have to add resistors R25 100 Ohm, R26 1k Ohm and R27 10k Ohm?
EDIT:
I added these resistors, cleaned the M8 and uploaded the latest soft.
After inserting the blocked ATMEGs into the doctor and pressing the button, the red LED lights up. The ATMEGS themselves are already seen by the programmer, so they probably have unlocked. But why is the green LED off?

It's hard for me to tell. When you plug the terminal, something more will be known - but probably some error on the board. The patient's supply voltage should drop to 0V, not 3.4V as before. Check if the board is correct after updating. Also improve the fusebits because the new load is 1MHz, it matters because now everything is happening 8 times too fast.

As part of self-improvement, I etched the latest plate and soldered. Now it works as it should On the basis of the new tile, I will look for errors in the old one and correct it.
R2-R23 resistors are to be in the range from 300 to 1k? I put it on as you suggested 330.

Yes, 330 is ok.

On the other hand, with 1k resistors, it is sometimes not surprising

I started to wonder whether to put in 1k as in the description of the board and the diagram, or 330 as the author suggested. There was an update of the board and the schematic and the 1k resistors remained
The most important thing is that the doctor is working. The author deserves a lot of applause, a lot of ATMEG has been saved
PS Can you recommend some DIY RS232 converter => USB for this project?

In fact, I could change the PCB and the schematic to something smaller than 1K, but it's not a big problem, maybe 3-4 people have had trouble with it.

More like UART- USB. You will find such projects on the author's website, I recommend one on the FT232 chip

Yes, yes, if USB UART is a mandatory FT232 - let go of other systems.

Siemano.
I've blocked myself Attiny85. I don't have an overlay for tinny. How should I connect to the trap?

According to the HVSP adapter diagram.

Don't forget to short the XTAL1 and OE pins to force serial operation.

Hello, I have put the doctor's fusebit on the breadboard and I have a problem, because after connecting 2 microcontrollers and pressing the "start" button, the red LED lights up. I just don't know if it is because some connections are incorrect or because the damaged atmega is not going to be repaired.

I am also wondering if it is not the fault of the power supply, because I bought 12v 100mA regulated by the following: http://lispol.pl/product,3283,zasilacz_antenowy_stabilizowany_regulowany.html " target="_blank" rel="nofollow noopener ugc" class="postlink inline" title="" > http://lispol.pl/product,3283,zasilacz_antenowy_stabilizowany_regulowany.html
and when everything is connected, the voltage drops to about 10.5V. Is this tension enough?

There may be a problem with such voltage ...

I checked again and the system consumes about 55mA (before pressing the start button) and the voltage is then about 11V. Oddly enough, the no-load voltage is 15.10V, so it drops by more than 4V. Hence the question, is it possible to somehow reduce the current consumption, e.g. by increasing the value of some resistors?

Hello, I will say there that the equipment works great, I folded it a week ago and saved Atmege 48 and Atmege 16, the voltage I gave was 9V, the red LED was flashing but the STK200 can see it without any problems, I will add that the batch from the first post of the author's first page. really good job, congratulations
Regards Paweł