wrong voltage

[Alain BREGAND]

Hi Bryan
my 4.1 board is now completed
I made during test before flashing a mistake about the power suply and
I bled the cpu with 8 volt instead of 5volt . the cpu got a bit warmer
but nothing else
after correction all voltages are ok
I put the soft into the AT2560 without trouble(Open9x for '9X Radios
with Fr-Sky and speaker modifications¶)
the reading of the memory is ok with companion after flashing
but the display stays BLANK
I cannot see if the cpu is again alive or dead( no scope trace of the
oscillator, neither on the lcd lines ) ,
do anybody knows any test point avalaible for the cpu?
many thanks

[Cam]

Absolute maximum for the CPU is stated as 6.0v on the datasheet, so
there is a chance it is damaged :(

Was the LCD plugged in at the time of this error ? The AP7215 3.3v
regulator has an absolute max. of 5.5v. The Torex brand regulator would
be ok with an absolute max. @ 8.0v Maybe it's just the LCD that is
damaged.

Have you tried to see if a PPM signal is output ? Install a module and
Rx to check.

Although the ability to program the CPU via ISP looks like the CPU is
still alive, they do become damaged in strange ways with voltage
excursions like this.

Cam.

[Romolo Manfredini]

First I will check the voltages out of 3.3 regulator, then i will look for ppm.
You can PPM out of the trainer port just using a soundcard oscilloscope there are several apps for any op sys.
Regards, Romolo

[Alain BREGAND]

many thanks to Cam and Romolo for your responses
several precisions about my foolishness
I inverted the7805 entry and exit so the entry of the of the 3.3v was
bled with with 8 v ( 2 lithium ion cells)
instead of that at this time all the voltages are ok :5v,3.3v, etc
that means that every voltage regulator are OK
I changed the lcd panel without result and PPM signal is still
missing on the oscilloscope
now I think that the AT 2560 never worked, perhaps due to
overheat when soldering ,static discharge , .... I bought this AT2560
in China at a low price........
Alain


theOn 27 mai, 19:48, Romolo Manfredini <romolo.manfred...@gmail.com>
wrote:

[Bryan]

Personally, I'd say that if the MCU was fed with 8V instead of 5V -- then that toasted it, for sure. They are absolutely not tolerant of such treatment, I'm afraid.

Evidently, the Flash ROM and EEPROM circuitry can stand it -- which does make sense, given their internal operating nature. But I feel certain that the CPU circuitry would have been utterly destroyed by the over-voltage. (It would only take a single sub-circuit to be damaged to render the chip inoperative, after all.)

Do you have what is needed to remove and replace the chip? The safest way I know to do this is to pre-heat the board from underneath to 150ºC. Then use (maximum) 360ºC air from above to melt the the solder and remove the chip. Obviously, a suitable hot-air set-up is required.

You can get away with only 360ºC air from beneath the board IF you are very careful and VERY patient. This temperature is hot enough to damage the PCB if concentrated on a single area for too long -- and it takes at least four (4) full minutes to get enough heat through the fibreglass to melt the solder above.

I can report that I have several times moved an ATmega2561 (64-pin version) from one board to another and not had any trouble with it afterwards. So I believe it's far less likely that heat dmaage is the cause of your problems.

If the above is not going to be practical, I can offer to attempt an exchange for you, if you post the board to me. HOWEVER, it is very easy to damage the finer tracks of the 100-pin ATmega2561 and I have failed to do this without damage on one out of two occasions in the past. So I can only offer to do my best. (I will provide the replacement chip for you and return shipping, regardless of success or not.)

Good luck and do report back.

Bryan.

On 28 May 2012 18:19, Alain BREGAND <alain....@gmail.com> wrote:

many thanks to Cam and Romolo for your responses

...

[Cam]

>I inverted the7805 entry and exit so the entry of the of the 3.3v was
>bled with with 8 v ( 2 lithium ion cells)

You installed the 7805 regulator inverted ? and soldered the tab where
the 3.3v regulator tab should have been ?

Otherwise how did you get Vin and Vout reversed ? Is that what you mean
?

Or did you put the 5v and 3.3v regulators in each others places ?

Cam.

[Dmitri]

Hi Alain,
Another way to get rid of the dead chip is to cut its pins (very
carefully) after the chip is out using soldering gun remove all pins
from the board.
Take care!

[Bryan]

That is another way to remove the chip. However, extreme care is required. For example, you should not use clippers, because the squashing and twisting of the pins will almost certainly tear up some of the tracks. (I have managed 64 pin devices several times without a problem. But the 100-pin device tracks are so much more fragile, it seems.)

The safest way I know of cutting the pins for this purpose is to set up a high speed dremel cutting disc in with a clamp to hold the tool head firmly at just the right height over the PCB. It can be done free hand, but it's so easy to slip or have the cutting wheel 'grab' and spin off into some place it's not wanted.

If you have access to hot air, then it's probably the best bet. Other wise yes -- carefully cut all the pins from the chip, remove the chip, then remove all the pins. Keep the heat as as will work effectively, of course -- because it can be easier to lift a PCB track due to excessive heat than to damage a chip, as luck would have it.

Bryan.

[Cam]

I would add weight to what Bryan has said on how to remove the chip.
Heat from below and hot air from above, and clippers could end in
failure.

I would offer to have a go removing it as well but Bryan is the field
leader here (sorry Bryan).

Cam.

--
Cam

[Bryan]

On 30 May 2012 22:29, Cam <th9...@gmail.com> wrote:

I would add weight to what Bryan has said on how to remove the chip. Heat from below and hot air from above, and clippers could end in failure.

I would offer to have a go removing it as well but Bryan is the field leader here (sorry Bryan).

Yup. It's true. I wrecked more boards -- I mean, had more practice -- than Cam, so far. :P (So I have a slight edge on noticing slightly sooner when things are going wrong. Hmmm. :P)

Still willing to have a go, with 80% confidence things can be righted for you (having not seen the board -- 95%+ if all is within certain specs. :-D) Alas, the postal delay cannot be avoided.

Bryan.

[Alain BREGAND]

How to remove a chip like AT 2560?
I used a KADA hot air and soldering station ( sold in Dubai for 30
euros without bargain near the Hyatt hotel : in this neighborhood
many shops are selling almost mobile phone parts and soldering
stations at a very cheap price )
the temperature used was 335°C
I used also an aluminium paper from the kitchen to make a shield and
to avoid the unsoldering of surrounding components
My problem about the blank screen was coming from the crystal and the
AT 2560 was OK in spite of the wrong voltage and the board is now
working as scheduled
good day

[Cam]

Awesome glad to here.

Cam.

[Bryan]

And very glad to hear too! :-)

[FROMI]

I use dremel to remove :)
Another qwe regarding voltage:
got involved in a bad romance with one t9x radios. You know, polarity bricking.
Walked to the local store, get few 5v reg...but I have only two letters on that A E. Is that OK for us here?dont want to fry it again :)

On Monday, June 11, 2012 2:14:45 PM UTC+2, Gruvin wrote:

And very glad to hear too! :-)

[Bryan]

Sorry, but without a manufacturer and part number, the letters mean nothing. What were the part numbers from the store?

Bryan.

[Ferenc Roman]

have no clue, it was not marked..ok, I will run few tests... Polarity change will burn it, so, need two for a test:)

[Bryan]

Surely the store you got them from has all the required data?

[Ferenc Roman]

but that will take more time :)

[Bryan]

Oh yeah. Of course. :P