Hi We use a '147 MSP controlling a front panel (LCD and Keypad). There is also Blackfin DSP on this carrier card (CC). Another module plugs onto the CC which has a `147 MSP (doing power monitoring and watchdog control) and a TI C64 DSP. We sometimes found that the MSP's (mostly the one on the CC) `latch- up', i.e. it does not startup and not even a RESET gets it out of that state? Only a `proper' power cycle (removing the power from the board for a while) will recover the MSP. The CC is powered by a 5V AC-DC supply. 3.3V is generated form 5V with a DC-DC and supplied to the MSP and other devices. The MSP's reset signal is supplied from a LDO's Reset pin (1.8 volt). This reset is delayed by app 220ms. There are also an push button (with RC de-bounce) on the MSP reset. It takes app 1ms for Vcc to rise to 1.1V (V_POR). From there it takes 700us to get to 1.8V (V_min). This problem is more prominent when switching on a few units at the same time. The devices are powered from mains. What can the problem be? Thanks for your help, Christo Coetzee
147 POR Problems
Started by ●February 18, 2005
Reply by ●February 18, 20052005-02-18
The 14x and other early MSP's are notorious for having a really bad
POR.
The POR is supposed to operate at 0.4V, but, in fact the decoupling caps
normally keep this up around 0.8V for long enough that the micro
latches, and can only be recovered, as you describe, with a long power
disconnection. Since uyou're using a Blackifin processor you obviously
aren't bothered about low power apps, so the simplest solution is a 1K
bleed resistor across power and ground, close to the micro. When I first
discovered this pronblem years ago I luckily had a 0.1 and a 10uF
decoupling power close to the mciro and just doubled up with the
resistor on top of the cap. Not great design, but it works like a charm,
and is easier than redisning in a separate power controller.
Al
christocoetzee2003 wrote:
>
>Hi
>We use a '147 MSP controlling a front panel (LCD and Keypad). There
>is also Blackfin DSP on this carrier card (CC). Another module
>plugs onto the CC which has a `147 MSP (doing power monitoring
>and
>watchdog control) and a TI C64 DSP.
>
>We sometimes found that the MSP's (mostly the one on the CC)
>`latch-
>up', i.e. it does not startup and not even a RESET gets it out of
>that state? Only a `proper' power cycle (removing the power
>from
>the board for a while) will recover the MSP.
>
>The CC is powered by a 5V AC-DC supply. 3.3V is generated form 5V
>with a DC-DC and supplied to the MSP and other devices. The
>MSP's
>reset signal is supplied from a LDO's Reset pin (1.8 volt). This
>reset is delayed by app 220ms. There are also an push button (with
>RC de-bounce) on the MSP reset.
>It takes app 1ms for Vcc to rise to 1.1V (V_POR). From there it
>takes 700us to get to 1.8V (V_min).
>This problem is more prominent when switching on a few units at the
>same time. The devices are powered from mains.
>
>What can the problem be?
>
>Thanks for your help,
>
>Christo Coetzee
>
>
>
>
>
>
>
>
>
>.
>
>
>Yahoo! Groups Links
>
>
>
>
>
>
>
>
>
>
>
Reply by ●February 18, 20052005-02-18
Christo, I would verify the JTAG connector/signals. I have seen MSP430F14x going to lockup because JTAG state machine moves to a state where holds CPU... and JTAG resets only when VCC goes below 0.8V. -Augusto -----Original Message----- From: christocoetzee2003 [mailto:christo@chri...] Sent: Friday, February 18, 2005 12:16 PM To: msp430@msp4... Subject: [msp430] 147 POR Problems Hi We use a '147 MSP controlling a front panel (LCD and Keypad). There is also Blackfin DSP on this carrier card (CC). Another module plugs onto the CC which has a `147 MSP (doing power monitoring and watchdog control) and a TI C64 DSP. We sometimes found that the MSP's (mostly the one on the CC) `latch- up', i.e. it does not startup and not even a RESET gets it out of that state? Only a `proper' power cycle (removing the power from the board for a while) will recover the MSP. The CC is powered by a 5V AC-DC supply. 3.3V is generated form 5V with a DC-DC and supplied to the MSP and other devices. The MSP's reset signal is supplied from a LDO's Reset pin (1.8 volt). This reset is delayed by app 220ms. There are also an push button (with RC de-bounce) on the MSP reset. It takes app 1ms for Vcc to rise to 1.1V (V_POR). From there it takes 700us to get to 1.8V (V_min). This problem is more prominent when switching on a few units at the same time. The devices are powered from mains. What can the problem be? Thanks for your help, Christo Coetzee . Yahoo! Groups Links -- Internal Virus Database is out-of-date. Checked by AVG Anti-Virus. Version: 7.0.300 / Virus Database: 265.8.7 - Release Date: 10/2/2005
Reply by ●February 21, 20052005-02-21
I've used the 430 parts in a couple designs and seen this affect. I have followed the evolution of th3 430 line so I was wondering if anyone has any information as to which parts are effected and which ones are ok. Thanks, Adam --- In msp430@msp4..., Onestone <onestone@b...> wrote: > The 14x and other early MSP's are notorious for having a really bad POR. > The POR is supposed to operate at 0.4V, but, in fact the decoupling caps > normally keep this up around 0.8V for long enough that the micro > latches, and can only be recovered, as you describe, with a long power > disconnection. Since uyou're using a Blackifin processor you obviously > aren't bothered about low power apps, so the simplest solution is a 1K > bleed resistor across power and ground, close to the micro. When I first > discovered this pronblem years ago I luckily had a 0.1 and a 10uF > decoupling power close to the mciro and just doubled up with the > resistor on top of the cap. Not great design, but it works like a charm, > and is easier than redisning in a separate power controller. > > Al > > christocoetzee2003 wrote: > > > > >Hi > >We use a '147 MSP controlling a front panel (LCD and Keypad). There > >is also Blackfin DSP on this carrier card (CC). Another module > >plugs onto the CC which has a `147 MSP (doing power monitoring > >and > >watchdog control) and a TI C64 DSP. > > > >We sometimes found that the MSP's (mostly the one on the CC) > >`latch- > >up', i.e. it does not startup and not even a RESET gets it out of > >that state? Only a `proper' power cycle (removing the power > >from > >the board for a while) will recover the MSP. > > > >The CC is powered by a 5V AC-DC supply. 3.3V is generated form 5V > >with a DC-DC and supplied to the MSP and other devices. The > >MSP's > >reset signal is supplied from a LDO's Reset pin (1.8 volt). This > >reset is delayed by app 220ms. There are also an push button (with > >RC de-bounce) on the MSP reset. > >It takes app 1ms for Vcc to rise to 1.1V (V_POR). From there it > >takes 700us to get to 1.8V (V_min). > >This problem is more prominent when switching on a few units at the > >same time. The devices are powered from mains. > > > >What can the problem be? > > > >Thanks for your help, > > > >Christo Coetzee > > > > > > > > > > > > > > > > > > > >. > > > > > >Yahoo! Groups Links > > > > > > > > > > > > > > > > > > > > > >
Reply by ●February 22, 20052005-02-22
Al, Thanks, this was the problem indeed. Vcc falls to about 0.6 volt with power off and then takes 8 sec to reach 0.3 V. In total it takes app 30 sec to reach 0V. We added the 1k and the problem seem to have disappeared!!! Will see when the client gets hold of it;-) Thanks!! --- In msp430@msp4..., Onestone <onestone@b...> wrote: > The 14x and other early MSP's are notorious for having a really bad POR. > The POR is supposed to operate at 0.4V, but, in fact the decoupling caps > normally keep this up around 0.8V for long enough that the micro > latches, and can only be recovered, as you describe, with a long power > disconnection. Since uyou're using a Blackifin processor you obviously > aren't bothered about low power apps, so the simplest solution is a 1K > bleed resistor across power and ground, close to the micro. When I first > discovered this pronblem years ago I luckily had a 0.1 and a 10uF > decoupling power close to the mciro and just doubled up with the > resistor on top of the cap. Not great design, but it works like a charm, > and is easier than redisning in a separate power controller. > > Al > > christocoetzee2003 wrote: > > > > >Hi > >We use a '147 MSP controlling a front panel (LCD and Keypad). There > >is also Blackfin DSP on this carrier card (CC). Another module > >plugs onto the CC which has a `147 MSP (doing power monitoring > >and > >watchdog control) and a TI C64 DSP. > > > >We sometimes found that the MSP's (mostly the one on the CC) > >`latch- > >up', i.e. it does not startup and not even a RESET gets it out of > >that state? Only a `proper' power cycle (removing the power > >from > >the board for a while) will recover the MSP. > > > >The CC is powered by a 5V AC-DC supply. 3.3V is generated form 5V > >with a DC-DC and supplied to the MSP and other devices. The > >MSP's > >reset signal is supplied from a LDO's Reset pin (1.8 volt). This > >reset is delayed by app 220ms. There are also an push button (with > >RC de-bounce) on the MSP reset. > >It takes app 1ms for Vcc to rise to 1.1V (V_POR). From there it > >takes 700us to get to 1.8V (V_min). > >This problem is more prominent when switching on a few units at the > >same time. The devices are powered from mains. > > > >What can the problem be? > > > >Thanks for your help, > > > >Christo Coetzee > > > > > > > > > > > > > > > > > > > >. > > > > > >Yahoo! Groups Links > > > > > > > > > > > > > > > > > > > > > >
Reply by ●February 22, 20052005-02-22
It should be fine now, I have a few thousand '149 based boards out
there
like this, and the only returns I get are when some idiot drops them and
breaks the displays. usually.
Al
christocoetzee2003 wrote:
>Al,
>
>Thanks, this was the problem indeed. Vcc falls to about 0.6 volt
>with power off and then takes 8 sec to reach 0.3 V. In total it
>takes app 30 sec to reach 0V.
>
>We added the 1k and the problem seem to have disappeared!!! Will
>see when the client gets hold of it;-)
>
>
>Thanks!!
>
>
>
>--- In msp430@msp4..., Onestone <onestone@b...> wrote:
>
>
>>The 14x and other early MSP's are notorious for having a really
>>
>>
>bad POR.
>
>
>>The POR is supposed to operate at 0.4V, but, in fact the
>>
>>
>decoupling caps
>
>
>>normally keep this up around 0.8V for long enough that the micro
>>latches, and can only be recovered, as you describe, with a long
>>
>>
>power
>
>
>>disconnection. Since uyou're using a Blackifin processor you
>>
>>
>obviously
>
>
>>aren't bothered about low power apps, so the simplest solution is
>>
>>
>a 1K
>
>
>>bleed resistor across power and ground, close to the micro. When I
>>
>>
>first
>
>
>>discovered this pronblem years ago I luckily had a 0.1 and a 10uF
>>decoupling power close to the mciro and just doubled up with the
>>resistor on top of the cap. Not great design, but it works like a
>>
>>
>charm,
>
>
>>and is easier than redisning in a separate power controller.
>>
>>Al
>>
>>christocoetzee2003 wrote:
>>
>>
>>
>>>Hi
>>>We use a '147 MSP controlling a front panel (LCD and Keypad).
>>>
>>>
>There
>
>
>>>is also Blackfin DSP on this carrier card (CC). Another module
>>>plugs onto the CC which has a `147 MSP (doing power monitoring
>>>and
>>>watchdog control) and a TI C64 DSP.
>>>
>>>We sometimes found that the MSP's (mostly the one on the CC)
>>>`latch-
>>>up', i.e. it does not startup and not even a RESET gets it out
of
>>>that state? Only a `proper' power cycle (removing the power
>>>
>>>
>>>from
>>
>>
>>>the board for a while) will recover the MSP.
>>>
>>>The CC is powered by a 5V AC-DC supply. 3.3V is generated form 5V
>>>with a DC-DC and supplied to the MSP and other devices. The
>>>MSP's
>>>reset signal is supplied from a LDO's Reset pin (1.8 volt).
This
>>>reset is delayed by app 220ms. There are also an push button
>>>
>>>
>(with
>
>
>>>RC de-bounce) on the MSP reset.
>>>It takes app 1ms for Vcc to rise to 1.1V (V_POR). From there it
>>>takes 700us to get to 1.8V (V_min).
>>>This problem is more prominent when switching on a few units at
>>>
>>>
>the
>
>
>>>same time. The devices are powered from mains.
>>>
>>>What can the problem be?
>>>
>>>Thanks for your help,
>>>
>>>Christo Coetzee
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>.
>>>
>>>
>>>Yahoo! Groups Links
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>
>
>
>
>
>.
>
>
>Yahoo! Groups Links
>
>
>
>
>
>
>
>
>
>
>