Go Down

Topic: mega 2560 r3 only works without one of 22p caps of crystal of atmega16u2 (Read 1 time) previous topic - next topic

Jcolombo

A few days ago I got a mega 2560 r3 ( SainSmart ) which did not work, it was not detected when plugging usb to pc. Although I have very limited knowledge of electronics I tried to detect its failure. I started with a multimeter at least corroborate some volts and continuity observing schematic. Suddenly , when measurements the board was plugged was detected by the computer as UNKNOW (in windows 7).
The ends of the tester were between the ends of R2 (in parallel to 16Mhz crystal). Encouraged, I tried to inspect C14 and C15 (22p caps) , I disconnected the board, tested parallel to GND and were ok. I thought it was a problem welding and pressing too took off C14 involuntarily. Despite the accident , I plugged the board and good news ! ( to my surprise ! ) windows recognized the board as " Arduino Mega 2560 " voila!!  .
Checked (w/capacimeter) C14 and was ok, re-weld it .... but again it stopped working the board !   :~

Anyone know what other defect may be causing this anomaly?
Someone knows where I can obtain documents about measures/tests the standard values ??(voltages, continuity, etc.)?
Do you think I should do?

I'm afraid to use the board and finish being damaged further.

Thanks in advance!
Jorge

jremington

I can imagine that the capacitor was not properly soldered to the board or is somehow short-circuited. They aren't very critical, so if the board works without the capacitor, you can probably go ahead and use it. The board may be unreliable in operation (e.g. it could just stop working) but there is very little chance you would damage anything by using it.

Can you return the board under warranty?

CrossRoads

The cap may be the wrong value and is preventing oscillation.
Folks have mistakenly used 22uF, 1uF, 0.1uF, instead of 22pF and kept their boards from working.
Measure the one you took off, see what it is. Many meters can measure capacitance.
Designing & building electrical circuits for over 25 years.  Screw Shield for Mega/Due/Uno,  Bobuino with ATMega1284P, & other '328P & '1284P creations & offerings at  my website.

Docedison

That looks like a bad crystal more than anything else. The 'loading' capacitors are required to make the Crystal 'happy'... the technical reasons (several) are a bit more complicated and require some intimate knowledge of how crystals are matched in circuit to the driving electronics... IMO a lot more technical than the scope of this post.

Doc

Crossroads suggestion has great merit however with small value capacitor values accurate measurements are difficult due to the effects o f the meter leads... if you have the correct value caps replace them. If not then certainly measuring the part value is going to point out gross errors as he mentioned...  and probably fix your issue.
--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

Jcolombo

Thanks to all for quick response!

Just tried it with another new 21-22 pf cap which I checked with a precision multimeter (and with REL button)... but the board stopped working again... then so I decided to remove it again and use the board without it !  :)

Anyway, I left undisturbed with Crossroads's opinion...
my concern was could damage the microprocessor or same fault but apparently work even though the cap values ??are very different.

thanks
Jorge

retrolefty

There have been cases of people wiring up a 328P to a solderless breadboard with a 16Mhz crystal but with no padding capacitors and having it work fine.

There is some small of input pin capacitance for the clock pins (which adds to the external padding capacitance) and the biggest effect is most likely that the 16MHz will be off frequency by a amount then it would otherwise be without the 'offending cap', assuming the crystal/oscillator starts to begin with.

On the mega2560 board the 16 MHz crystal and padding caps are for the 32U2 USB serial converter chip, not the avr mega2560 chip which uses a ceramic resonator with internal padding caps. So as long as your PC recognizes the board when you plug it in
and you can upload sketches, I wouldn't worry about it too much. But Like Bob my guess is the cap is the wrong value and preventing the crystal/oscillator from starting up.

Lefty

mart256


Thanks to all for quick response!

Just tried it with another new 21-22 pf cap which I checked with a precision multimeter (and with REL button)... but the board stopped working again... then so I decided to remove it again and use the board without it !  :)

Anyway, I left undisturbed with Crossroads's opinion...
my concern was could damage the microprocessor or same fault but apparently work even though the cap values ??are very different.

thanks
Jorge



Sometimes there is parasitic capacitance between the crystal and GND, and this added to your 22nf caps prevents the microcontroller from starting.

It happened to me once, with a homemade Arduino. I removed one of the 22nf caps and now it works like a charm (with just 1 cap 22nf). Anyway, check the delay function is working properly (measure the time and check it has correct timing), because less capacitance on the crystal may acelerate (or slow? not sure) the clock.

retrolefty



Thanks to all for quick response!

Just tried it with another new 21-22 pf cap which I checked with a precision multimeter (and with REL button)... but the board stopped working again... then so I decided to remove it again and use the board without it !  :)

Anyway, I left undisturbed with Crossroads's opinion...
my concern was could damage the microprocessor or same fault but apparently work even though the cap values ??are very different.

thanks
Jorge



Sometimes there is parasitic capacitance between the crystal and GND, and this added to your 22nf caps prevents the microcontroller from starting.

It happened to me once, with a homemade Arduino. I removed one of the 22nf caps and now it works like a charm (with just 1 cap 22nf). Anyway, check the delay function is working properly (measure the time and check it has correct timing), because less capacitance on the crystal may acelerate (or slow? not sure) the clock.


Can't check that was as the crystal is used for the USB serial converter not the mega chip one programs.


mart256

:smiley-surprise:  you are right, there are 2 crystals in your board. On mine I used just 1.

Jcolombo

Quote
.... which uses a ceramic resonator with internal padding caps....


retrolefty, good info!

Quote
Sometimes there is parasitic capacitance between the crystal and GND, and this added to your 22nf caps prevents the microcontroller from starting.


mart256: I measured capacitance with the tester at the board, from one side to GND pin header, and with bridge to GND very close to each  caps, that was 80 to 82p. It should be 44p (!?)... I think that is confirmed your theory!


Paul__B


I measured capacitance with the tester at the board, from one side to GND pin header, and with bridge to GND very close to each  caps, that was 80 to 82p. It should be 44p (!?)... I think that is confirmed your theory!


Ah, now that is tricky, because if you measure it with the power off, then the characteristics of the MCU chip are quite different and it effectively has back-to-back ("varactor") diodes connected across each pin.

If you measure with the power on, then it is attempting - with varying degrees of success - to oscillate which is very likely to interfere with the measuring function of your meter.  To actually measure the capacitances, you would probably need to reprogram the chip to disable this crystal oscillator and remove the crystal itself.

Go Up