Setting fuses to use an external crystal
47 views
Skip to first unread message
G. Andrew Stone
Mar 12, 2012, 5:14:02 PM3/12/12
to toasted-thum...@googlegroups.com
>> Andrew says:
>>
>> David, to change the 328 to use an external crystal, run avrdude with the following: "-U lfuse:w:0xff:m -U hfuse:w:0xD2:m -U efuse:w:0x05:m" (This will also change preserve the EEPROM across a full-chip >> erase). Of course, once you do this your chip will suddenly stop working on the breadboard since it will need an external crystal. It won't even respond to the programmer...
> David says:
> I tried doing this,setting the fuzes, with the crystal and 2 caps on the oscillator pins. I keep getting stk500 errors/timeouts ( programmer not responding) though it seems to get information from the 328p.
> Curious, can the blue and yellow leds give any diagnostics/ or what do they represent on the ThumbISP?
David, you've made progress! You are now at the point where "it will need an external crystal. It won't even respond to the programmer without one" Basically, it looks like your chip can't get a clock from that crystal. This circuit can be extremely sensitive both to cap type and values and to physical issues like distance (i.e. wire-created resistance, capacitance and inductance). Can you post a photo and schematic? And experiment with different cap values; use non-electrolytic and in the 10-35pF range.
The yellow/green and blue LEDs tell you stuff about the TTP -- primarily whether its in serial or program mode -- but there's nothing for it to tell about the chip -- because it can't talk to it :-(. Maybe I should add a special blink pattern if that's the case...
Cheers!
Andrew
>>
>> David, to change the 328 to use an external crystal, run avrdude with the following: "-U lfuse:w:0xff:m -U hfuse:w:0xD2:m -U efuse:w:0x05:m" (This will also change preserve the EEPROM across a full-chip >> erase). Of course, once you do this your chip will suddenly stop working on the breadboard since it will need an external crystal. It won't even respond to the programmer...
> David says:
> I tried doing this,setting the fuzes, with the crystal and 2 caps on the oscillator pins. I keep getting stk500 errors/timeouts ( programmer not responding) though it seems to get information from the 328p.
> Curious, can the blue and yellow leds give any diagnostics/ or what do they represent on the ThumbISP?
David, you've made progress! You are now at the point where "it will need an external crystal. It won't even respond to the programmer without one" Basically, it looks like your chip can't get a clock from that crystal. This circuit can be extremely sensitive both to cap type and values and to physical issues like distance (i.e. wire-created resistance, capacitance and inductance). Can you post a photo and schematic? And experiment with different cap values; use non-electrolytic and in the 10-35pF range.
The yellow/green and blue LEDs tell you stuff about the TTP -- primarily whether its in serial or program mode -- but there's nothing for it to tell about the chip -- because it can't talk to it :-(. Maybe I should add a special blink pattern if that's the case...
Cheers!
Andrew
David Lapointe
Mar 13, 2012, 6:02:15 AM3/13/12
to toasted-thum...@googlegroups.com
On Monday, March 12, 2012 5:14:02 PM UTC-4, G. Andrew Stone wrote:
> >> Andrew says:
> >>
> >> David, to change the 328 to use an external crystal, run avrdude with
> the following: "-U lfuse:w:0xff:m -U hfuse:w:0xD2:m -U efuse:w:0x05:m"
> (This will also change preserve the EEPROM across a full-chip >> erase).
> respond to the programmer...
>
>
> > David says:
> > I tried doing this,setting the fuzes, with the crystal and 2 caps
> respond to the programmer without one" Basically, it looks like your chip can't get a clock from that crystal. This circuit can be extremely sensitive both to cap type and values and to physical issues like distance (i.e. wire-created resistance, capacitance and inductance). Can you post a photo and schematic? And experiment with different cap values; use non-electrolytic and in the 10-35pF range.
>
> The yellow/green and blue LEDs tell you stuff about the TTP -- primarily whether its in serial or program mode -- but there's nothing for it to tell about the chip -- because it can't talk to it :-(. Maybe I should add a special blink pattern if that's the case...
>
> Cheers!
> Andrew
On Monday, March 12, 2012 5:14:02 PM UTC-4, G. Andrew Stone wrote:
> >> Andrew says:
> >>
> >> David, to change the 328 to use an external crystal, run avrdude with
> the following: "-U lfuse:w:0xff:m -U hfuse:w:0xD2:m -U efuse:w:0x05:m"
> (This will also change preserve the EEPROM across a full-chip >> erase).
> respond to the programmer...
>
>
> > David says:
> > I tried doing this,setting the fuzes, with the crystal and 2 caps
> respond to the programmer without one" Basically, it looks like your chip can't get a clock from that crystal. This circuit can be extremely sensitive both to cap type and values and to physical issues like distance (i.e. wire-created resistance, capacitance and inductance). Can you post a photo and schematic? And experiment with different cap values; use non-electrolytic and in the 10-35pF range.
>
> The yellow/green and blue LEDs tell you stuff about the TTP -- primarily whether its in serial or program mode -- but there's nothing for it to tell about the chip -- because it can't talk to it :-(. Maybe I should add a special blink pattern if that's the case...
>
> Cheers!
> Andrew
On Monday, March 12, 2012 5:14:02 PM UTC-4, G. Andrew Stone wrote:
> >> Andrew says:
> >>
> >> David, to change the 328 to use an external crystal, run avrdude with
> the following: "-U lfuse:w:0xff:m -U hfuse:w:0xD2:m -U efuse:w:0x05:m"
> (This will also change preserve the EEPROM across a full-chip >> erase).
> respond to the programmer...
>
>
> > David says:
> > I tried doing this,setting the fuzes, with the crystal and 2 caps
> respond to the programmer without one" Basically, it looks like your chip can't get a clock from that crystal. This circuit can be extremely sensitive both to cap type and values and to physical issues like distance (i.e. wire-created resistance, capacitance and inductance). Can you post a photo and schematic? And experiment with different cap values; use non-electrolytic and in the 10-35pF range.
>
> The yellow/green and blue LEDs tell you stuff about the TTP -- primarily whether its in serial or program mode -- but there's nothing for it to tell about the chip -- because it can't talk to it :-(. Maybe I should add a special blink pattern if that's the case...
>
> Cheers!
> Andrew
I probably had some wiring issues since it was a challenge to place the crystal on the breadboard-, but I'm using 22 pF caps, non-electrolytic, connected between ground and the crystal.
I'll try a cleaner layout.
> >> Andrew says:
> >>
> >> David, to change the 328 to use an external crystal, run avrdude with
> the following: "-U lfuse:w:0xff:m -U hfuse:w:0xD2:m -U efuse:w:0x05:m"
> (This will also change preserve the EEPROM across a full-chip >> erase).
> Of course, once you do this your chip will suddenly stop working on the
> breadboard since it will need an external crystal. It won't even
> respond to the programmer...
>
>
> > David says:
> > I tried doing this,setting the fuzes, with the crystal and 2 caps
> on the oscillator pins. I keep getting stk500 errors/timeouts (
> programmer not responding) though it seems to get information from the
> 328p.
> > Curious, can the blue and yellow leds give any diagnostics/ or
> programmer not responding) though it seems to get information from the
> 328p.
> what do they represent on the ThumbISP?
>
> David, you've made progress! You are now at the point where "it will need an external crystal. It won't even
>
> respond to the programmer without one" Basically, it looks like your chip can't get a clock from that crystal. This circuit can be extremely sensitive both to cap type and values and to physical issues like distance (i.e. wire-created resistance, capacitance and inductance). Can you post a photo and schematic? And experiment with different cap values; use non-electrolytic and in the 10-35pF range.
>
> The yellow/green and blue LEDs tell you stuff about the TTP -- primarily whether its in serial or program mode -- but there's nothing for it to tell about the chip -- because it can't talk to it :-(. Maybe I should add a special blink pattern if that's the case...
>
> Cheers!
> Andrew
On Monday, March 12, 2012 5:14:02 PM UTC-4, G. Andrew Stone wrote:
> >> Andrew says:
> >>
> >> David, to change the 328 to use an external crystal, run avrdude with
> the following: "-U lfuse:w:0xff:m -U hfuse:w:0xD2:m -U efuse:w:0x05:m"
> (This will also change preserve the EEPROM across a full-chip >> erase).
> Of course, once you do this your chip will suddenly stop working on the
> breadboard since it will need an external crystal. It won't even
> respond to the programmer...
>
>
> > David says:
> > I tried doing this,setting the fuzes, with the crystal and 2 caps
> on the oscillator pins. I keep getting stk500 errors/timeouts (
> programmer not responding) though it seems to get information from the
> 328p.
> > Curious, can the blue and yellow leds give any diagnostics/ or
> programmer not responding) though it seems to get information from the
> 328p.
> what do they represent on the ThumbISP?
>
> David, you've made progress! You are now at the point where "it will need an external crystal. It won't even
>
> respond to the programmer without one" Basically, it looks like your chip can't get a clock from that crystal. This circuit can be extremely sensitive both to cap type and values and to physical issues like distance (i.e. wire-created resistance, capacitance and inductance). Can you post a photo and schematic? And experiment with different cap values; use non-electrolytic and in the 10-35pF range.
>
> The yellow/green and blue LEDs tell you stuff about the TTP -- primarily whether its in serial or program mode -- but there's nothing for it to tell about the chip -- because it can't talk to it :-(. Maybe I should add a special blink pattern if that's the case...
>
> Cheers!
> Andrew
On Monday, March 12, 2012 5:14:02 PM UTC-4, G. Andrew Stone wrote:
> >> Andrew says:
> >>
> >> David, to change the 328 to use an external crystal, run avrdude with
> the following: "-U lfuse:w:0xff:m -U hfuse:w:0xD2:m -U efuse:w:0x05:m"
> (This will also change preserve the EEPROM across a full-chip >> erase).
> Of course, once you do this your chip will suddenly stop working on the
> breadboard since it will need an external crystal. It won't even
> respond to the programmer...
>
>
> > David says:
> > I tried doing this,setting the fuzes, with the crystal and 2 caps
> on the oscillator pins. I keep getting stk500 errors/timeouts (
> programmer not responding) though it seems to get information from the
> 328p.
> > Curious, can the blue and yellow leds give any diagnostics/ or
> programmer not responding) though it seems to get information from the
> 328p.
> what do they represent on the ThumbISP?
>
> David, you've made progress! You are now at the point where "it will need an external crystal. It won't even
>
> respond to the programmer without one" Basically, it looks like your chip can't get a clock from that crystal. This circuit can be extremely sensitive both to cap type and values and to physical issues like distance (i.e. wire-created resistance, capacitance and inductance). Can you post a photo and schematic? And experiment with different cap values; use non-electrolytic and in the 10-35pF range.
>
> The yellow/green and blue LEDs tell you stuff about the TTP -- primarily whether its in serial or program mode -- but there's nothing for it to tell about the chip -- because it can't talk to it :-(. Maybe I should add a special blink pattern if that's the case...
>
> Cheers!
> Andrew
I probably had some wiring issues since it was a challenge to place the crystal on the breadboard-, but I'm using 22 pF caps, non-electrolytic, connected between ground and the crystal.
I'll try a cleaner layout.
David Lapointe
Mar 13, 2012, 7:45:21 AM3/13/12
to toasted-thum...@googlegroups.com
Sorry about the quoted material.
On Tue, Mar 13, 2012 at 6:02 AM, David Lapointe
<david.l...@gmail.com> wrote:
..
...excess quoted material...
--
.david
David Lapointe
Questions are the new answers
Reply all
Reply to author
Forward
0 new messages