1. This forum section is a read-only archive which contains old newsgroup posts. If you wish to post a query, please do so in one of our main forum sections (here). This way you will get a faster, better response from the members on Motherboard Point.

Reset on ATMEGA8535 oddity

Discussion in 'Embedded' started by Dennis Clark, Oct 5, 2003.

  1. Dennis Clark

    Dennis Clark Guest

    I've used the simple cap/resistor RC circuit on the reset line of my
    AVR boards forever quite reliably to hold the AVR in reset until the power
    was up. However I'm finding that it is highly unreliable on the newer
    ATMEGA8535. Has anyone run across an appnote or errata that specifies
    the need to do something different between the 90S8535 and the MEGA8535?
    If you have can you post a reference or your experience anyway? If this
    doesn't work I guess I'll have to break down and put a real reset device
    in there like a DS1233 or something.

    Thanks,
    DLC
     
    Dennis Clark, Oct 5, 2003
    #1
    1. Advertisements

  2. Dennis Clark wrote:
    ....
    You can always enable internal BOD.
    I know, it is not answer for your question, but maybe can help.

    Regards,
     
    Artur Lipowski, Oct 6, 2003
    #2
    1. Advertisements

  3. Dennis Clark

    Dennis Clark Guest

    : Dennis Clark wrote:
    : ...
    :> If you have can you post a reference or your experience anyway? If this
    :> doesn't work I guess I'll have to break down and put a real reset device
    :> in there like a DS1233 or something.
    : You can always enable internal BOD.
    : I know, it is not answer for your question, but maybe can help.

    Artur,

    Thanks, that was the first thing that I did, but it didn't help a bit.
    I've also found that when fiddling with the fuse settings on the AVR one
    should be rather careful. I twiddled one bit (slow rise on resonator) that
    rendered the chip non responsive. Since the device is in a board using
    ISP programming this presented a bit of a problem. I solved it by grabbing
    the clock out crystal pin signal from one board and injecting it into the
    clock in crystal pin on that board to reset the fuses. Whew.

    thanks anyways though,
    DLC

    : Regards,
    : --
    : Artur Lipowski
     
    Dennis Clark, Oct 7, 2003
    #3
  4. Dennis Clark

    Tim Mitchell Guest

    I've found that the Mega devices are much more sensitive to the fuse
    settings than the AT90S... devices were. I've also had the above problem
    of losing ISP by changing one of the startup fuse settings, I had to get
    hold of a parallel programmer to rescue the device.

    I would guess that this is causing your problem. The mega parts have an
    enormous range of power-up fuse settings which only seem to work in
    limited conditions and maybe the startup characteristics of your board
    fall in between them. While testing out the various fuse settings I
    found that devices would sometimes not start up the first time, then
    next time and all subsequent times would start, which made testing a bit
    unpredictable.
     
    Tim Mitchell, Oct 7, 2003
    #4
  5. Not nice.
    When facing this type of problem, it can help to cycle through
    a tube of (ideally spread date codes) devices. That way, you have
    devices
    of long un-powered times to check. It's surprising for how long
    unpowered CMOS can behave a little like EEPROM, especially when you
    don't want it to :)
    -jg
     
    Jim Granville, Oct 7, 2003
    #5
  6. Dennis Clark

    Hul Tytus Guest

    A bleeder resistor on the circuit board can be requisite at times.

    Hul


    : Tim Mitchell wrote:
    :>
    :> In article <3f822549$0$196$>, Dennis Clark
    :> >: Dennis Clark wrote:
    :> >: ...
    :> >:> If you have can you post a reference or your experience anyway? If this
    :> >:> doesn't work I guess I'll have to break down and put a real reset device
    :> >:> in there like a DS1233 or something.
    :> >: You can always enable internal BOD.
    :> >: I know, it is not answer for your question, but maybe can help.
    :> >
    :> >Artur,
    :> >
    :> > Thanks, that was the first thing that I did, but it didn't help a bit.
    :> >I've also found that when fiddling with the fuse settings on the AVR one
    :> >should be rather careful. I twiddled one bit (slow rise on resonator) that
    :> >rendered the chip non responsive. Since the device is in a board using
    :> >ISP programming this presented a bit of a problem. I solved it by grabbing
    :> >the clock out crystal pin signal from one board and injecting it into the
    :> >clock in crystal pin on that board to reset the fuses. Whew.
    :> >
    :> I've found that the Mega devices are much more sensitive to the fuse
    :> settings than the AT90S... devices were. I've also had the above problem
    :> of losing ISP by changing one of the startup fuse settings, I had to get
    :> hold of a parallel programmer to rescue the device.
    :>
    :> I would guess that this is causing your problem. The mega parts have an
    :> enormous range of power-up fuse settings which only seem to work in
    :> limited conditions and maybe the startup characteristics of your board
    :> fall in between them. While testing out the various fuse settings I
    :> found that devices would sometimes not start up the first time, then
    :> next time and all subsequent times would start, which made testing a bit
    :> unpredictable.
    :
    : Not nice.
    : When facing this type of problem, it can help to cycle through
    : a tube of (ideally spread date codes) devices. That way, you have
    : devices
    : of long un-powered times to check. It's surprising for how long
    : unpowered CMOS can behave a little like EEPROM, especially when you
    : don't want it to :)
    : -jg
     
    Hul Tytus, Oct 8, 2003
    #6
  7. Dennis Clark

    Dennis Clark Guest

    : In article <3f822549$0$196$>, Dennis Clark
    [snip]
    :> Thanks, that was the first thing that I did, but it didn't help a bit.
    :>I've also found that when fiddling with the fuse settings on the AVR one
    :>should be rather careful. I twiddled one bit (slow rise on resonator) that
    :>rendered the chip non responsive. Since the device is in a board using
    :>ISP programming this presented a bit of a problem. I solved it by grabbing
    :>the clock out crystal pin signal from one board and injecting it into the
    :>clock in crystal pin on that board to reset the fuses. Whew.
    :>
    : I've found that the Mega devices are much more sensitive to the fuse
    : settings than the AT90S... devices were. I've also had the above problem
    : of losing ISP by changing one of the startup fuse settings, I had to get
    : hold of a parallel programmer to rescue the device.

    I'd have been SOL there, my part is PLCC and my trusty STK500 is strictly
    DIP, which makes it useless with the new MEGA parts being in those tiny
    pitch SMT only formats...

    : I would guess that this is causing your problem. The mega parts have an
    : enormous range of power-up fuse settings which only seem to work in
    : limited conditions and maybe the startup characteristics of your board
    : fall in between them. While testing out the various fuse settings I
    : found that devices would sometimes not start up the first time, then
    : next time and all subsequent times would start, which made testing a bit
    : unpredictable.

    I've decided NOT to fiddle with the fuses any more beyond the simple
    stuff and add a 80 cent reset chip. My MEGA8535 starts fine every time
    with my cockroached DS1233 on it. That was enough to decide me - I went
    back and modded the schematics and added the part to the PCB layout, fin.

    DLC
     
    Dennis Clark, Oct 8, 2003
    #7
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.