Soldering ATMEGA328p without destroying the chip

[Glen Woolley]

Hi guys,

Is there some secret to soldering the ATMEGA328P chip to a board
without destroying the chip in the process? Ive tried hot air w/
solder paste at 300c twice and also with a soldering iron and really
fine lead/tin solder and i cant get anywhere...

Every time i try to load code to the chips it fails to pass the chip
code verification stage and only returns a Chip ID of 0x0000000 (looks
like an error to me)

What am i doing wrong?

Below is the circuit diagram but im pretty sure its correct.
http://imgur.com/UmzCj

Thanks
Glen W

[Madox]

Hi,

Are you sure it is being cooked?  Have you tried toning down the programmer speed?  [usual suspect #1]

[Patrick Barnes]

Adding to the basic questions;
- Any solder bridges or non-contacts?
- The electrical connection between chip and programmer is good?
- The programmer successfully programs another board?
- The schematic is correct, and there's nothing hanging off the pins
that might interfere (I don't know enough about AVR programmers to okay
your schematic)

- Have you tried switching it off and back on again? :-)

> --
> You received this message because you are subscribed to the Google
> Groups "Robots & Dinosaurs" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/sydney-hackspace/-/pCgXHXVxUHcJ.
> To post to this group, send email to sydney-h...@googlegroups.com.
> To unsubscribe from this group, send email to
> sydney-hackspa...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/sydney-hackspace?hl=en.

[Lemming]

Hi guys,

For the board I have been working on, the SNARC, I have hand soldered
about 20 through hole ATMega328p's without issue.

<pimpage> This is the board in question, Http://hsbne.org/projects/SNARC
If you want one, please feel free to email me off list.</pimpage>

What variant of the chip are you using Glen? Through hole or one of
the SMD versions?

Also are you programming it using ICSP or FTDI, if FTDI remember that
you need to put a bootloader on the chip first.

Lemming.

[Luke Weston]

I would say that destroying the chip by overheating it is actually unlikely to be what your real problem is.

Have you previously known your ISP programmer software and hardware chain to be proven to work on known working target hardware - say by reflashing the boot loader on a standard Arduino for example?

Any solder bridges or dry joints when you solder the chip?

Any other electronics, such as your LED display driver circuit, needs to be designed properly if it is sharing the AVR's SPI pins, otherwise it will load them down and stuff up ISP programming.

Or, if you can, disconnect other circuits from those pins during programming.

[Glen Woolley]

Im using ICSP using a AVRISPmkII. It works for other boards and all
the Arduinos i have. Can burn bootloaders/fuses and such.

Soldering a SMD version, Thin Quad Flat Package i believe. I got it
from little bird electronics a week or so ago.

One of the boards have the PSU soldered onto it but the others have
only the ICSP header, and the chip. Powered from an external 5v power
source. No crystal on the others but i want to use the internal 8mhz
crystal.

It will drive LED's but they are all on a separate board so i can
simply disconnect the board and the only things on the board will be a
SMPS and the MPU.

>> Have you tried toning down the programmer speed

Im not sure if i can. AVRdude does not appear to have that option.
Correct me if im wrong.

Thanks,
Glen

> --
> You received this message because you are subscribed to the Google Groups "Robots & Dinosaurs" group.

> To view this discussion on the web visit https://groups.google.com/d/msg/sydney-hackspace/-/pVyh2VSvDR8J.

[James McGill]

Do all ATmega328's come with the fuse bits set up for internal RC
oscillator? I know some of the ATMegaU series don't.

I second the opinion that it's much more likely an problem with the
circuit or a fused or missing pad, than it is to be a fried IC. Have
you connection tested all the pins involved in programming (MISO,
MOSI, RST, VCC, GND, SCK)?

Regards,
James

[Lemming]

I had all sorts of fun programming the most recent version of my board
as I had gotten MOSI and MISO mixed up on the ICSP header.

From what I have seen you can program it without a crystal, but I
think it's easier to do it with.

> > On Sat, Jul 21, 2012 at 7:39 PM, Luke Weston <reindeerfloti...@gmail.com> wrote:
> >> I would say that destroying the chip by overheating it is actually unlikely to be what your real problem is.
>
> >> Have you previously known your ISP programmer software and hardware chain to be proven to work on known working target hardware - say by reflashing the boot loader on a standard Arduino for example?
>
> >> Any solder bridges or dry joints when you solder the chip?
>
> >> Any other electronics, such as your LED display driver circuit, needs to be designed properly if it is sharing the AVR's SPI pins, otherwise it will load them down and stuff up ISP programming.
>
> >> Or, if you can, disconnect other circuits from those pins during programming.
>
> >> --
> >> You received this message because you are subscribed to the Google Groups "Robots & Dinosaurs" group.

> >> To view this discussion on the web visithttps://groups.google.com/d/msg/sydney-hackspace/-/pVyh2VSvDR8J.

> >> To post to this group, send email to sydney-h...@googlegroups.com.
> >> To unsubscribe from this group, send email to sydney-hackspa...@googlegroups.com.

> >> For more options, visit this group athttp://groups.google.com/group/sydney-hackspace?hl=en.

[Tristan Steele]

From my experience there is no need for a crystal to program the chip.  It is possible to program a chip with no components, except connections to the ICSP header.

The only thing that comes to mind has already been mentioned, the programming speed.  I have reused a number of these chips that have been subjected to multiple solder/desolder/resolder passes in both QFN and QFP packages and never had a device die from heat, so I really don't think this is the likely cause.

Tristan

[Matt Callow]

On 22 July 2012 17:00, Glen Woolley <silver...@gmail.com> wrote:
>
>>> Have you tried toning down the programmer speed
>
> Im not sure if i can. AVRdude does not appear to have that option.
> Correct me if im wrong.
>

-i option?

Matt

[Glen Woolley]

Hi all,

Worked it out and it turns out that avrdude has problems with this
chip with the version that comes with WinAVR and i grabbed a new
version of avrdude and patched WinAVR with it and it all works now!

Its nice knowing i dont have to re-order the chips :)

> --
> You received this message because you are subscribed to the Google Groups "Robots & Dinosaurs" group.

[tALSit de CoD]

Sure... blame it on the software!! Typical attitude from the hardware guys! 

;)

--
// talsit.org

[Patrick Barnes]

How many software engineers does it take to change a lightbulb?
None, it's a hardware problem.

How many hardware engineers does it take to change a lightbulb?
None, it's a software problem.

[Terry Dawson]

On Thu, Aug 2, 2012 at 10:46 AM, Patrick Barnes <mrt...@gmail.com> wrote:
> How many software engineers does it take to change a lightbulb?
> None, it's a hardware problem.
>
> How many hardware engineers does it take to change a lightbulb?
> None, it's a software problem.

How many light-bulbs does it take to change an engineer?

http://www.summers-consulting.com/PDF/lightbulb.pdf

Terry

[David Lyon]

What about all the people stuck in between.

Like John and Kean. They can do both hardware and software.

I think the lines are becoming more blurred these days, except
for web type developers of course. Just buy a new machine if
the battery goes flat...

[Jason Ball]

These days developers seem to believe machines have unlimited amounts of memory, limitless IO's on disk sub systems and international networks can easily transfer terabytes of data in minutes!

Sorry - for the most part I believe developers are even more detached from the hardware than before.  The exceptions are few and far between.

-- 

[Dr Ada Not-a-fruit]

On Aug 2, 10:46 am, Patrick Barnes <mrtr...@gmail.com> wrote:
> How many software engineers does it take to change a lightbulb?
> None, it's a hardware problem.
>
> How many hardware engineers does it take to change a lightbulb?
> None, it's a software problem.

Who are the three most dangerous people in computing?

A hardware engineer with a compiler
A software engineer with a soldering iron
A user with a clue

[Andrew Stone]

Hi,  I'm new to this excellent mailing list! :)

Jason's comment really resonated with me... (I'm a gen X programmer in my 30s  who grew up in the era of the C-64).
 

>Sorry - for the most part I believe developers are even more detached from the hardware than before.  The exceptions are few and far between.

 

Don't be sorry, this is true! The gap is now accelerating rapidly with the advent of the cloud... and now service platforms on the cloud (see: http://www.heroku.com/ for a brilliant platform... but it brings a new level of ignorance).

I think that this is part of the reason why platforms like the Raspberry Pi are important to the next generation of kids.
Platforms like the Pi put software and hardware nice and close together again, hopefully this will help?

PS: I would love to come down to R&D sometime soon. looks very cool!

Cheers,
Stonie.

[Dr Ada Not-a-fruit]

On Aug 2, 11:45 am, Andrew Stone <and...@drivenlogic.com.au> wrote:

> Don't be sorry, this is true! The gap is now accelerating rapidly with the

> advent of the cloud... and now service platforms on the cloud (see:http://www.heroku.com/for a brilliant platform... but it brings a new
> level of ignorance).

I think that IAAS actually increases awareness because you have to pay
for every resource your application uses, as opposed to general apps
programming where you can say 'everyone has a pentium 7 with six
terabytes of RAM!'

> I think that this is part of the reason why platforms like the Raspberry Pi
> are important to the next generation of kids.
> Platforms like the Pi put software and hardware nice and close together
> again, hopefully this will help?

Is the Pi not too high level? Minimal hardware hacking potential,
fully developed OS... as much as I hate basic, what do you think of
something like the maximite?

> PS: I would love to come down to R&D sometime soon. looks very cool!

Hopefully see you soon!

[Andrew Stone]

 

I think that IAAS actually increases awareness because you have to pay
for every resource your application uses

Yeah that is a good point vs. big bucket type computing.
Still it's a bit like putting petrol and oil in your car... won’t teach you much about the cars mechanics, or to change a flat tire.

Is the Pi not too high level?  Minimal hardware hacking potential,
fully developed OS...

I hear you on the hardware side, it's a compromise, however this looks pretty sweet: http://hackaday.com/2012/07/23/the-proper-way-to-put-an-arduino-in-a-raspberry-pi/
On the other hand I think the software side is the big pull for kids; they can play games surf the web etc... starting at $35 that's very accessible given it works with your average TV. 

 as much as I hate basic, what do you think of something like the maximite?

I had not heard of the maximite just had a cursory look a the website... have to agree with you about basic. Conversely the Pi leaves a lot of options open for education/coding.

Cheers,
Stonie.