MSP430x2xx family - Power up - Problem with Brownout Reset

[dinaebnit]

Hi,

I am facing a problem with BOR on MSP430F2121. I am using a solar cell
as power supply. The power supply on my MCU rise very slowly (above
30mV/s). When the power supply goes up to 1.9V, the MCU does not start
propery. I assume that the BOR signal is not generate. I expected that
a  BOR signal reset the MCU in order to start the application when
power voltage is sufficient.

Any idea to fix that?

Thanks

[leon Heller-2]

----- Original Message -----
From: "dinaebnit" <dinaebnit@...>
To: <msp430@...>
Sent: Wednesday, December 05, 2007 2:30 PM
Subject: [msp430] MSP430x2xx family - Power up - Problem with Brownout Reset

> Hi,
>
> I am facing a problem with BOR on MSP430F2121. I am using a solar cell
> as power supply. The power supply on my MCU rise very slowly (above
> 30mV/s). When the power supply goes up to 1.9V, the MCU does not start
> propery. I assume that the BOR signal is not generate. I expected that
> a  BOR signal reset the MCU in order to start the application when
> power voltage is sufficient.
>
> Any idea to fix that?

An external reset chip?

Leon
--
Leon Heller
Amateur radio call-sign G1HSM
Yaesu FT-817ND and FT-857D transceivers
Suzuki SV1000S motorcycle
leon355@...
http://www.geocities.com/leon_heller 

[Hernan Tradatti]

What frequency is the MCLK running?.
Maybe you are trying to run at high speed before proper voltage is
stablished and that´s why your program crashes.
bye

--- In msp430@..., "Leon" <leon355@...> wrote:
>
> ----- Original Message -----
> From: "dinaebnit" <dinaebnit@...>
> To: <msp430@...>
> Sent: Wednesday, December 05, 2007 2:30 PM
> Subject: [msp430] MSP430x2xx family - Power up - Problem with
Brownout Reset
>
>
> > Hi,
> >
> > I am facing a problem with BOR on MSP430F2121. I am using a solar
cell
> > as power supply. The power supply on my MCU rise very slowly
(above
> > 30mV/s). When the power supply goes up to 1.9V, the MCU does not
start
> > propery. I assume that the BOR signal is not generate. I expected
that

> > a  BOR signal reset the MCU in order to start the application when
> > power voltage is sufficient.
> >
> > Any idea to fix that?
>
> An external reset chip?
>
> Leon
> --
> Leon Heller
> Amateur radio call-sign G1HSM
> Yaesu FT-817ND and FT-857D transceivers
> Suzuki SV1000S motorcycle
> leon355@...
> http://www.geocities.com/leon_heller
>

[old_cow_yellow]

Do you use a crystal or change the DCO frequency?
After BOR releases nRST, Vcc is still very low and the CPU will not be
able to run with fast MCLK.

Also, if you turns on peripherals or starts other activities that
requires extra power, Vcc may drop and cause another BOR again and again.

If none of the above, an external BOR chip will always help.

--- In msp430@..., "dinaebnit" <dinaebnit@...> wrote:

>
> Hi,
>
> I am facing a problem with BOR on MSP430F2121. I am using a solar cell
> as power supply. The power supply on my MCU rise very slowly (above
> 30mV/s). When the power supply goes up to 1.9V, the MCU does not start
> propery. I assume that the BOR signal is not generate. I expected that
> a  BOR signal reset the MCU in order to start the application when
> power voltage is sufficient.
>
> Any idea to fix that?
>
> Thanks
>

[dinaebnit]

--- In msp430@..., "Leon" <leon355@...> wrote:
>
> ----- Original Message -----
> From: "dinaebnit" <dinaebnit@...>
> To: <msp430@...>
> Sent: Wednesday, December 05, 2007 2:30 PM
> Subject: [msp430] MSP430x2xx family - Power up - Problem with
Brownout Reset

>
>
> > Hi,
> >
> > I am facing a problem with BOR on MSP430F2121. I am using a solar cell
> > as power supply. The power supply on my MCU rise very slowly (above
> > 30mV/s). When the power supply goes up to 1.9V, the MCU does not start
> > propery. I assume that the BOR signal is not generate. I expected that
> > a  BOR signal reset the MCU in order to start the application when
> > power voltage is sufficient.
> >
> > Any idea to fix that?
>
> An external reset chip?
>
> Leon
> --
> Leon Heller
> Amateur radio call-sign G1HSM
> Yaesu FT-817ND and FT-857D transceivers
> Suzuki SV1000S motorcycle
> leon355@...
> http://www.geocities.com/leon_heller
>

Well, I could be a good idea.
 
Hereafter is my experimentations :
Once the power supply is above 2.8V, I try to apply a hardware reset
signal on pin RST. The MCU did not start even if the supply voltage is
suffisant. The current is about 50uA (seems to be proportionnal to
supply voltage).

My program does not modifie the default MCLK & DCO settings. It only
initialise the timer resource on external 32768 crystal clock.

btw,
Can anybody reproduce the problem? It would be interesting to share
experience on such strange behaviour.

[dinaebnit]

Well, the DCO frequency is not modified, I think it is about 1MHz on
startup.
It seens that the MCU does not start even if the power supply is
suffisant.
I understand that the MCU can loop on startup routine when heavy
peripherals are switched on. I do not think this is the point.

External BOR will probably not work out. In my experience, a hardware
reset cannot solve the problem.

Here is a new point. When power supply is low (below 1.0V), I apply a
reset pulse on RST pin. The I rise up the power supply slowly. When
the supply voltage is above 1.4V, the MCU starts up; the program I
wrote togges an output at 1Hz.

Does anybody know if such external BOR chip is available for a 1.5V
threshold?

--- In msp430@..., "old_cow_yellow" <old_cow_yellow@...>
wrote:
>
> Do you use a crystal or change the DCO frequency?
> After BOR releases nRST, Vcc is still very low and the CPU will not be
> able to run with fast MCLK.
>
> Also, if you turns on peripherals or starts other activities that
> requires extra power, Vcc may drop and cause another BOR again and
again.

>
> If none of the above, an external BOR chip will always help.
>
> --- In msp430@..., "dinaebnit" <dinaebnit@> wrote:
> >
> > Hi,
> >
> > I am facing a problem with BOR on MSP430F2121. I am using a solar cell
> > as power supply. The power supply on my MCU rise very slowly (above
> > 30mV/s). When the power supply goes up to 1.9V, the MCU does not start
> > propery. I assume that the BOR signal is not generate. I expected that
> > a  BOR signal reset the MCU in order to start the application when
> > power voltage is sufficient.
> >
> > Any idea to fix that?
> >
> > Thanks
> >
>

[dinaebnit]

The DCO frequency is the default one on startup.

wtb, I post the question on TI support. Still no reply.
I will let you know when I get an answer.

Thanks

Daniel

--- In msp430@..., "htrada2" <htrada@...> wrote:

>
> What frequency is the MCLK running?.
> Maybe you are trying to run at high speed before proper voltage is
> stablished and that´s why your program crashes.
> bye
>
> --- In msp430@..., "Leon" <leon355@> wrote:
> >
> > ----- Original Message -----
> > From: "dinaebnit" <dinaebnit@>
> > To: <msp430@...>
> > Sent: Wednesday, December 05, 2007 2:30 PM
> > Subject: [msp430] MSP430x2xx family - Power up - Problem with
> Brownout Reset
> >
> >
> > > Hi,
> > >
> > > I am facing a problem with BOR on MSP430F2121. I am using a solar
> cell
> > > as power supply. The power supply on my MCU rise very slowly
> (above
> > > 30mV/s). When the power supply goes up to 1.9V, the MCU does not
> start
> > > propery. I assume that the BOR signal is not generate. I expected
> that
> > > a  BOR signal reset the MCU in order to start the application when
> > > power voltage is sufficient.
> > >
> > > Any idea to fix that?
> >
> > An external reset chip?
> >
> > Leon
> > --
> > Leon Heller
> > Amateur radio call-sign G1HSM
> > Yaesu FT-817ND and FT-857D transceivers
> > Suzuki SV1000S motorcycle
> > leon355@
> > http://www.geocities.com/leon_heller
> >
>

[old_cow_yellow]

If for some reason the WDTNMI bit in the WDTCTL register is set to 1
while NMIIE bit in IE1 register remains 0 (PUC default), the RST/NMI
pin will not generate a reset nor an interrupt.

This would explain why it does not reset. But how does a slow rising
Vcc set WDTNMI?

--- In msp430@..., "dinaebnit" <dinaebnit@...> wrote:
> ...
> Hereafter is my experimentations :
> Once the power supply is above 2.8V, I try to apply a hardware
> reset signal on pin RST. The MCU did not start even if the supply
> voltage is suffisant.
> ...

[DeeAitch]

dinaebnit wrote:

Hi,

I am facing a problem with BOR on MSP430F2121. I am using a solar cell
as power supply. The power supply on my MCU rise very slowly (above
30mV/s). When the power supply goes up to 1.9V, the MCU does not start
propery. I assume that the BOR signal is not generate. I expected that
a  BOR signal reset the MCU in order to start the application when
power voltage is sufficient.

Any idea to fix that?

Thanks

Hi 2121 user!
I have experienced the same problem, both in 2121 and 2101 devices and the problem seams to vary between batches of devices. I have recently done some measurments to really understand when and why the BOR does not fully resets the MSP.
My setup:
- MSP2121 with pullup resistor (22k) on RST pin.
- Vcc is generated by solar cell, via voltage limiter (3V).
- Software inits. external resonator (3,4MHz) on startup.
- When resonator is running, the MCLK is devided by 2 and shortly thereafter by 4.

My findings:
Normally, the CPU start execute code at ~1,5V. (reservation for measuring errors) and stops executing code at ~1.4V. (hysteresis in BOR).

1. Once the MSP is woken up properly, the Vcc may vary in any dV/ds as long as Vcc is over ~100mV. The BOR seams to work in this scenario (ofcource, I have to do further tests on this).

2. If Vcc starts from 0V and increases ~1V/s, sometimes the BOR does not seam to work (See uploaded screendump of oscilloscope were the yellow track shows CPU code execution and the red track shows Vcc).  Because the CPU never starts even if the Vcc reaches 3V. In this state, the MSP cannot be reseted with low signal on RST pin. The only way to get the MSP out of this state is to set Vcc < 100mV (or 0V).

3. If the Vcc starts from 0V and increases 2V/s or faster, the MSP starts up fine and acts according to the normal scenario described above.

My conclusion so far:
The MSP may not power up/reset itself properly when Vcc increases from 0V.

Is it not the BOR task to ensure proper reset regardless of positive/negative Vcc ramps (dV/Ds)?
So, what is the solution...?
What happends in the MSP if the RST pin is held low (with external voltage detector) until Vcc is 1.6-1.8V? Can it solve all strange Vcc scenarios like the one described above in point 2?

Kind Regards
/DeeAitch

[DeeAitch]

DeeAitch wrote:

dinaebnit wrote:

Hi,

I am facing a problem with BOR on MSP430F2121. I am using a solar cell
as power supply. The power supply on my MCU rise very slowly (above
30mV/s). When the power supply goes up to 1.9V, the MCU does not start
propery. I assume that the BOR signal is not generate. I expected that
a  BOR signal reset the MCU in order to start the application when
power voltage is sufficient.

Any idea to fix that?

Thanks

Hi 2121 user!
I have experienced the same problem, both in 2121 and 2101 devices and the problem seams to vary between batches of devices. I have recently done some measurments to really understand when and why the BOR does not fully resets the MSP.
My setup:
- MSP2121 with pullup resistor (22k) on RST pin.
- Vcc is generated by solar cell, via voltage limiter (3V).
- Software inits. external resonator (3,4MHz) on startup.
- When resonator is running, the MCLK is devided by 2 and shortly thereafter by 4.

My findings:
Normally, the CPU start execute code at ~1,5V. (reservation for measuring errors) and stops executing code at ~1.4V. (hysteresis in BOR).

1. Once the MSP is woken up properly, the Vcc may vary in any dV/ds as long as Vcc is over ~100mV. The BOR seams to work in this scenario (ofcource, I have to do further tests on this).

2. If Vcc starts from 0V and increases ~1V/s, sometimes the BOR does not seam to work (See uploaded screendump of oscilloscope were the yellow track shows CPU code execution and the red track shows Vcc).  Because the CPU never starts even if the Vcc reaches 3V. In this state, the MSP cannot be reseted with low signal on RST pin. The only way to get the MSP out of this state is to set Vcc < 100mV (or 0V).

3. If the Vcc starts from 0V and increases 2V/s or faster, the MSP starts up fine and acts according to the normal scenario described above.

My conclusion so far:
The MSP may not power up/reset itself properly when Vcc increases from 0V.

Is it not the BOR task to ensure proper reset regardless of positive/negative Vcc ramps (dV/Ds)?
So, what is the solution...?
What happends in the MSP if the RST pin is held low (with external voltage detector) until Vcc is 1.6-1.8V? Can it solve all strange Vcc scenarios like the one described above in point 2?

Kind Regards
/DeeAitch

As a reply to my own post recenly, I would like to add an answer to my own questions:
Q. What happends in the MSP if the RST pin is held low (with external voltage detector) until Vcc is 1.6-1.8V? Can it solve all strange Vcc scenarios like the one described above in point 2?
A. No. I just discovered that holding the RST low until Vcc gets over 1.7V does not help.

/DeeAitch

[redbeard4545]

--- In msp430@..., DeeAitch <sm5wpx@...> wrote:

>
>
>
>
> DeeAitch wrote:
> >
> >
> > dinaebnit wrote:
> >>
> >> Hi,
> >>
> >> I am facing a problem with BOR on MSP430F2121. I am using a solar

cell
> >> as power supply. The power supply on my MCU rise very slowly (above
> >> 30mV/s). When the power supply goes up to 1.9V, the MCU does not
start
> >> propery. I assume that the BOR signal is not generate. I expected
that

> >> a BOR signal reset the MCU in order to start the application when
> >> power voltage is sufficient.
> >>
> >> Any idea to fix that?
> >>
> >> Thanks
> >>
> >>
> >>
> >
> > Hi 2121 user!
> > I have experienced the same problem, both in 2121 and 2101 devices

and the
> > problem seams to vary between batches of devices. I have recently
done
> > some measurments to really understand when and why the BOR does not
fully

> > resets the MSP.
> > My setup:
> > - MSP2121 with pullup resistor (22k) on RST pin.
> > - Vcc is generated by solar cell, via voltage limiter (3V).
> > - Software inits. external resonator (3,4MHz) on startup.
> > - When resonator is running, the MCLK is devided by 2 and shortly
> > thereafter by 4.
> >
> > My findings:
> > Normally, the CPU start execute code at ~1,5V. (reservation for

measuring
> > errors) and stops executing code at ~1.4V. (hysteresis in BOR).
> >
> > 1. Once the MSP is woken up properly, the Vcc may vary in any dV/ds
as
> > long as Vcc is over ~100mV. The BOR seams to work in this scenario
> > (ofcource, I have to do further tests on this).
> >
> > 2. If Vcc starts from 0V and increases ~1V/s, sometimes the BOR does
not
> > seam to work (See uploaded screendump of oscilloscope were the
yellow
> > track shows CPU code execution and the red track shows Vcc).
> >
http://www.nabble.com/file/p14618026/LCRY0301J11835_20080104_113849.jpg
> > Because the CPU never starts even if the Vcc reaches 3V. In this
state,
> > the MSP cannot be reseted with low signal on RST pin. The only way
to get
> > the MSP out of this state is to set Vcc < 100mV (or 0V).
> >
> > 3. If the Vcc starts from 0V and increases 2V/s or faster, the MSP
starts
> > up fine and acts according to the normal scenario described above.
> >
> > My conclusion so far:
> > The MSP may not power up/reset itself properly when Vcc increases
from 0V.
> >
> > Is it not the BOR task to ensure proper reset regardless of
> > positive/negative Vcc ramps (dV/Ds)?
> > So, what is the solution...?
> > What happends in the MSP if the RST pin is held low (with external
voltage
> > detector) until Vcc is 1.6-1.8V? Can it solve all strange Vcc
scenarios
> > like the one described above in point 2?
> >
> > Kind Regards
> > /DeeAitch
> >
> >
>
> As a reply to my own post recenly, I would like to add an answer to my
own
> questions:
> Q. What happends in the MSP if the RST pin is held low (with external
> voltage detector) until Vcc is 1.6-1.8V? Can it solve all strange Vcc
> scenarios like the one described above in point 2?
> A. No. I just discovered that holding the RST low until Vcc gets over
1.7V
> does not help. :-/
>
> /DeeAitch
>
> --
> View this message in context:
http://www.nabble.com/MSP430x2xx-family---Power-up---Problem-with-Browno\
ut-Reset-tp14174168p14618466.html
> Sent from the MSP430 - Discuss mailing list archive at Nabble.com.
>
Try switching power witha MOSFET when it is over 2.7V.  You can use a
power level detector to take the MCU out of reset at that time as well
as have a fast power rise time to the MCU.

[Onestone-2]

 on page 19 of the data sheet you will find the specification for the
Vcc rise time for correct operation of the POR/BOR circuit.

Al

DeeAitch wrote:

>
>DeeAitch wrote:
>  
>
>>dinaebnit wrote:
>>    
>>
>>>Hi,
>>>
>>>I am facing a problem with BOR on MSP430F2121. I am using a solar cell
>>>as power supply. The power supply on my MCU rise very slowly (above
>>>30mV/s). When the power supply goes up to 1.9V, the MCU does not start
>>>propery. I assume that the BOR signal is not generate. I expected that
>>>a  BOR signal reset the MCU in order to start the application when
>>>power voltage is sufficient.
>>>
>>>Any idea to fix that?
>>>
>>>Thanks
>>>
>>>
>>>
>>>      
>>>
>>Hi 2121 user!
>>I have experienced the same problem, both in 2121 and 2101 devices and the
>>problem seams to vary between batches of devices. I have recently done
>>some measurments to really understand when and why the BOR does not fully
>>resets the MSP.
>>My setup:
>>- MSP2121 with pullup resistor (22k) on RST pin.
>>- Vcc is generated by solar cell, via voltage limiter (3V).
>>- Software inits. external resonator (3,4MHz) on startup.
>>- When resonator is running, the MCLK is devided by 2 and shortly
>>thereafter by 4.
>>
>>My findings:
>>Normally, the CPU start execute code at ~1,5V. (reservation for measuring
>>errors) and stops executing code at ~1.4V. (hysteresis in BOR).
>>
>>1. Once the MSP is woken up properly, the Vcc may vary in any dV/ds as
>>long as Vcc is over ~100mV. The BOR seams to work in this scenario
>>(ofcource, I have to do further tests on this).
>>
>>2. If Vcc starts from 0V and increases ~1V/s, sometimes the BOR does not
>>seam to work (See uploaded screendump of oscilloscope were the yellow
>>track shows CPU code execution and the red track shows Vcc).
>>http://www.nabble.com/file/p14618026/LCRY0301J11835_20080104_113849.jpg 
>>Because the CPU never starts even if the Vcc reaches 3V. In this state,
>>the MSP cannot be reseted with low signal on RST pin. The only way to get
>>the MSP out of this state is to set Vcc < 100mV (or 0V).
>>
>>3. If the Vcc starts from 0V and increases 2V/s or faster, the MSP starts
>>up fine and acts according to the normal scenario described above.
>>
>>My conclusion so far:
>>The MSP may not power up/reset itself properly when Vcc increases from 0V.
>>
>>Is it not the BOR task to ensure proper reset regardless of
>>positive/negative Vcc ramps (dV/Ds)?
>>So, what is the solution...?
>>What happends in the MSP if the RST pin is held low (with external voltage
>>detector) until Vcc is 1.6-1.8V? Can it solve all strange Vcc scenarios
>>like the one described above in point 2?
>>
>>Kind Regards
>>/DeeAitch
>>
>>
>>    
>>
>
>As a reply to my own post recenly, I would like to add an answer to my own
>questions:
>Q. What happends in the MSP if the RST pin is held low (with external
>voltage detector) until Vcc is 1.6-1.8V? Can it solve all strange Vcc
>scenarios like the one described above in point 2?
>A. No. I just discovered that holding the RST low until Vcc gets over 1.7V
>does not help. :-/
>
>/DeeAitch
>
>  
>

[Non-text portions of this message have been removed]

[old_cow_yellow]

--- In msp430@..., DeeAitch <sm5wpx@...> wrote:

>
> As a reply to my own post recenly, I would like to add an
> answer to my own questions:
> Q. What happends in the MSP if the RST pin is held low
> (with external voltage detector) until Vcc is 1.6-1.8V?
> Can it solve all strange Vcc scenarios like the one
> described above in point 2?
> A. No. I just discovered that holding the RST low until
> Vcc gets over 1.7V does not help. :-/
>
> /DeeAitch

Somewhat puzzled by your latest finding.

Just to make sure I did not misunderstand you. If you hold nRST to
ground **WHEN Vcc=0** until Vcc slowly rises to say 1V before you
releases nRST, the CPU still does not start right?

I knew that when the CPU is already stuck, nRST may not be able to
un-stuck it. What surprised me was that the CPU may get stuck
**WHILE** nRST is held at ground.

[DeeAitch]

old_cow_yellow wrote:

Somewhat puzzled by your latest finding.

Just to make sure I did not misunderstand you. If you hold nRST to
ground **WHEN Vcc=0** until Vcc slowly rises to say 1V before you
releases nRST, the CPU still does not start right?

I knew that when the CPU is already stuck, nRST may not be able to
un-stuck it. What surprised me was that the CPU may get stuck
**WHILE** nRST is held at ground.

Yes you understand correct. Allthough i let Vcc rice to maybee 2V or more before I release nRST.
So the CPU may get stuck in slow voltage rising allthough the nRST is held low during the time.

/DeeAitch

[DeeAitch]

redbeard4545 wrote:

Try switching power witha MOSFET when it is over 2.7V.  You can use a
power level detector to take the MCU out of reset at that time as well
as have a fast power rise time to the MCU.

Hi redbeard.

Yes this may be the final solution to it. To let a XC61C CMOS voltage detector keep Vcc = 0 until a certain level. The XC61C can drive ~20mA so I think I can use it in series with Vcc to the MSP in my circuit. I am qurious if the Vd (detection level) of the XC61C can be perhaps 1.6V or 1.7V to make it work.....? Do you think that level is enough to make a stable start of the CPU?

note:
The MCLK is initiated with an external resonator (3,64MHz) and then devided to ~1,5MHz after startup. The datasheet does not tell how the MSP acts below 1.8V. But I can only test with Vd 1.6V to see...

What I do not understand is why the BOR flipps out at the low voltage ~100mV... Nothing in the datasheet tells me what happends at that low voltage. According to my measurments, there will not be a reset if Vcc has a slow rise from 0V. But the reset seams to work if Vcc varies in the region inbetween ~100mV - 1.6V with the condition that the MSP has been reset properly earlier.

/DeeAitch

[DeeAitch]

dinaebnit wrote:

Well, the DCO frequency is not modified, I think it is about 1MHz on
startup.
It seens that the MCU does not start even if the power supply is
suffisant.
I understand that the MCU can loop on startup routine when heavy
peripherals are switched on. I do not think this is the point.

External BOR will probably not work out. In my experience, a hardware
reset cannot solve the problem.

Here is a new point. When power supply is low (below 1.0V), I apply a
reset pulse on RST pin. The I rise up the power supply slowly. When
the supply voltage is above 1.4V, the MCU starts up; the program I
wrote togges an output at 1Hz.

Does anybody know if such external BOR chip is available for a 1.5V
threshold?

Both ST microelectronics and Torex has such devices, among others. Check their sites under Voltage detectors or alike.

/DeeAitch

[old_cow_yellow]

I tried to reproduce the power up problem you observed. But I found no
problem in my case.

a) I only have an old (Rev.D) F2131 chip. It is in PW package with the
markings on the chip:
430F2131
59K G4
AR65 D

b) I programmed an old F169 to generate a voltage ramp. The DAC12 can
source a few mA. The internal 2.5V reference is used, thus the max
voltage is 2.5V and the resolution is ~0.6mV (2.5V/4095). I start with
0 and increment by ~0.6mV every ~0.25Sec. It took ~10min to ramp up to
~1.5V and the F2131 starts up normally every time!

--- In msp430@..., DeeAitch <sm5wpx@...> wrote:

>
> old_cow_yellow wrote:
> >  
> > Somewhat puzzled by your latest finding.
> >
> > Just to make sure I did not misunderstand you. If you hold nRST to
> > ground **WHEN Vcc=0** until Vcc slowly rises to say 1V before you
> > releases nRST, the CPU still does not start right?
> >
> > I knew that when the CPU is already stuck, nRST may not be able to
> > un-stuck it. What surprised me was that the CPU may get stuck
> > **WHILE** nRST is held at ground.
>
> Yes you understand correct. Allthough i let Vcc rice to maybee 2V or

more
> before I release nRST.
> So the CPU may get stuck in slow voltage rising allthough the nRST
is held
> low during the time.
>
> /DeeAitch
> --
> View this message in context:
http://www.nabble.com/MSP430x2xx-family---Power-up---Problem-with-Brownout-Reset-tp14174168p14631271.html
> Sent from the MSP430 - Discuss mailing list archive at Nabble.com.
>

[DeeAitch]

old_cow_yellow wrote:

I tried to reproduce the power up problem you observed. But I found no
problem in my case.

a) I only have an old (Rev.D) F2131 chip. It is in PW package with the
markings on the chip:
430F2131
59K G4
AR65 D

b) I programmed an old F169 to generate a voltage ramp. The DAC12 can
source a few mA. The internal 2.5V reference is used, thus the max
voltage is 2.5V and the resolution is ~0.6mV (2.5V/4095). I start with
0 and increment by ~0.6mV every ~0.25Sec. It took ~10min to ramp up to
~1.5V and the F2131 starts up normally every time!

This is very strange, allthough other sais that it might not happen on all devices, just a fex. Perhaps batch related. Thank you for your revision notes.... I am using 2121 chips with the following prints:

430F2121
63K G4
HVCX D

and

430F2121
74k G4
D9JC E

Tha latter devices shown more problems than the first mentioned.
I am now awaiting the Torex voltage detector and I shall try to control the Vcc line to the MSP and see if it solves the problem.

Regards

[DeeAitch]

old_cow_yellow wrote:

Somewhat puzzled by your latest finding.

Just to make sure I did not misunderstand you. If you hold nRST to
ground **WHEN Vcc=0** until Vcc slowly rises to say 1V before you
releases nRST, the CPU still does not start right?

I knew that when the CPU is already stuck, nRST may not be able to
un-stuck it. What surprised me was that the CPU may get stuck
**WHILE** nRST is held at ground.

I have now tested several voltage detectors from Torex. They are placed at MSP Vcc so that the voltage switches ON to the MSP when the main voltage source reaches a certain level.

When I use the 1,6V voltage detector, the msp seams to work fine. It wakes up normally and starts executing code correctly.

But...

When I use any lower detection voltage (as 1,2V or 1,3V), the MSP starts to oscillate at ~1,5kHz when the voltage reaches ~1,5V and a big amount of current is drawn by the msp.
I have no idea why, and the thing is that I cannot use 1,6V of special reasons....
This problem is not there when I do not use any voltage detector and letting the MSP get its voltage supply directly from the solar cells.

So what causes the MSP to behave like this when there is a voltage detector controlling the Vcc? Even if the detection voltage is as low as 1,2V.

Very confused.....

Regards

[tintronic]

I am wondering what the final supply voltage is. Does it reach the
minimum operating voltage of 1.8V after the voltage detection and
before your code starts running?
What is the special reason why you cannot use the 1.6V detector,
knowing that then your circuit does work?


Michael K.


--- In msp430@..., DeeAitch <sm5wpx@...> wrote:

>
>
>
>
> old_cow_yellow wrote:
> >
> >
> > Somewhat puzzled by your latest finding.
> >
> > Just to make sure I did not misunderstand you. If you hold nRST to
> > ground **WHEN Vcc=0** until Vcc slowly rises to say 1V before you
> > releases nRST, the CPU still does not start right?
> >
> > I knew that when the CPU is already stuck, nRST may not be able to
> > un-stuck it. What surprised me was that the CPU may get stuck
> > **WHILE** nRST is held at ground.
> >
> >
> >
> >
>
> I have now tested several voltage detectors from Torex. They are

placed at
> MSP Vcc so that the voltage switches ON to the MSP when the main voltage
> source reaches a certain level.
>
> When I use the 1,6V voltage detector, the msp seams to work fine. It
wakes
> up normally and starts executing code correctly.
>
> But...
>
> When I use any lower detection voltage (as 1,2V or 1,3V), the MSP
starts to
> oscillate at ~1,5kHz when the voltage reaches ~1,5V and a big amount of
> current is drawn by the msp.
> I have no idea why, and the thing is that I cannot use 1,6V of special
> reasons....
> This problem is not there when I do not use any voltage detector and
letting
> the MSP get its voltage supply directly from the solar cells.
>
> So what causes the MSP to behave like this when there is a voltage
detector
> controlling the Vcc? Even if the detection voltage is as low as 1,2V.
>
> Very confused.....
>
> Regards
>
> --
> View this message in context:
http://www.nabble.com/MSP430x2xx-family---Power-up---Problem-with-Brownout-Reset-tp14174168p15043874.html
> Sent from the MSP430 - Discuss mailing list archive at Nabble.com.
>