EmbeddedRelated.com
Forums

High temperature opperation of the MSP430

Started by chrismholt123 October 3, 2007
I am curious as to the highest temperature that the MSP 430 can be
successfully operated at. Does anyone have any experience testing the
micro in extreme environments?

I searched your forum and found a test that was done by a forum member
of the flash memory at temperatures up to 100C.

Has anyone tested temperatures higher than this?

I do not need to write to flash but just need the micro to read an
external ADC via spi and record the readings to external memory.

I need a micro that will operate at 180C. I know that the PIC 18C
will reliably (over 1000 hours) operate at this temperature as well as
a the Motorola MC68332.

I like the development tools and emulation options provided by the MSP
430 as well as the incredible current consumption of the part.

Does anyone have experience using this micro at high temp?

I am a University Student at the University of Alberta and any help
could save me a few $$ on my project by not needing to test the MSP430
myself.

Thank you very much for your help!

Beginning Microcontrollers with the MSP430

High temperature also causes Alzheimer. What is already burned into to
the Flash memory will disappear a lot sooner.

--- In m..., Adriano Caye wrote:
>
> MSP430 shouldn't be adequate for your application. The MSP430F147, for
> example, supports up to 85 C. It can work on 180 C, but its operation
> will not be reliable.
>
> Regards,
> Adriano.
>
> chrismholt123 wrote:
>
> > I am curious as to the highest temperature that the MSP 430 can be
> > successfully operated at. Does anyone have any experience testing the
> > micro in extreme environments?
> >
> > I searched your forum and found a test that was done by a forum member
> > of the flash memory at temperatures up to 100C.
> >
> > Has anyone tested temperatures higher than this?
> >
> > I do not need to write to flash but just need the micro to read an
> > external ADC via spi and record the readings to external memory.
> >
> > I need a micro that will operate at 180C. I know that the PIC 18C
> > will reliably (over 1000 hours) operate at this temperature as well as
> > a the Motorola MC68332.
> >
> > I like the development tools and emulation options provided by the MSP
> > 430 as well as the incredible current consumption of the part.
> >
> > Does anyone have experience using this micro at high temp?
> >
> > I am a University Student at the University of Alberta and any help
> > could save me a few $$ on my project by not needing to test the MSP430
> > myself.
> >
> > Thank you very much for your help!
> >
> >
> > --
> > Esta mensagem foi verificada pelo sistema de antivus e
> > acredita-se estar livre de perigo.
>
>
> --
> Esta mensagem foi verificada pelo sistema de antivus e
> acredita-se estar livre de perigo.
>



Thank you for your help!

old cow yellow,
I was not aware of the "Alzheimer" effect on flash at higher
temperatures. I knew that programming and erasing flash at high
temperatures did not work but I did not know that the flash memory
would randomly flip bits at high temp. Do you know of a study or
paper that elaborates on this effect?

Adriano,
Reliability is a very general term. Many applications of micro
controllers cannot allow any failures after several years of
continuous operation and are expected to last well over 50,000 hours.
My application only needs to work for 100 hours at high temperature.

Has anyone actually taken a MSP 430 above 100C for any length of time?

Thank you again for your help.
--- In m..., "old_cow_yellow"
wrote:
>
> High temperature also causes Alzheimer. What is already burned into to
> the Flash memory will disappear a lot sooner.
>
> --- In m..., Adriano Caye wrote:
> >
> > MSP430 shouldn't be adequate for your application. The MSP430F147,
for
> > example, supports up to 85 C. It can work on 180 C, but its
operation
> > will not be reliable.
> >
> > Regards,
> > Adriano.
> >
> > chrismholt123 wrote:
> >
> > > I am curious as to the highest temperature that the MSP 430 can be
> > > successfully operated at. Does anyone have any experience
testing the
> > > micro in extreme environments?
> > >
> > > I searched your forum and found a test that was done by a forum
member
> > > of the flash memory at temperatures up to 100C.
> > >
> > > Has anyone tested temperatures higher than this?
> > >
> > > I do not need to write to flash but just need the micro to read an
> > > external ADC via spi and record the readings to external memory.
> > >
> > > I need a micro that will operate at 180C. I know that the PIC 18C
> > > will reliably (over 1000 hours) operate at this temperature as
well as
> > > a the Motorola MC68332.
> > >
> > > I like the development tools and emulation options provided by
the MSP
> > > 430 as well as the incredible current consumption of the part.
> > >
> > > Does anyone have experience using this micro at high temp?
> > >
> > > I am a University Student at the University of Alberta and any help
> > > could save me a few $$ on my project by not needing to test the
MSP430
> > > myself.
> > >
> > > Thank you very much for your help!
> > >
> > >
> > > --
> > > Esta mensagem foi verificada pelo sistema de antivus e
> > > acredita-se estar livre de perigo.
> >
> >
> > --
> > Esta mensagem foi verificada pelo sistema de antivus e
> > acredita-se estar livre de perigo.
>



You are scratching the surface of a very detailed subject, with many factors which
interact. But to answer your basic question I believe you want to look at Black's equation.
Some information can be found on Wikipedia:- http://en.wikipedia.org/wiki/Black's_equation

At a basic level, the IAR compiler has the ability to generate a ROM integrity checksum at
Link-time, which you could call at regular intervals to detect such failures.

To go one stage further, Error Correction Code, can be held in RAM (through the vulnerable
high temperature time), which would allow you patch a limited number of bit errors as and
when you need the data (or relocatable-code segments you might also need - luckily IAR can
also generate Position Independent Code segments you can overlay), but which can more permanently
be subsequently fixed in Flash, at a later time when the temperature comes back into spec.

Regards,

Colin,

-----Original Message-----
From: m... [mailto:m...] On Behalf Of chrismholt123
Sent: 05 October 2007 16:22
To: m...
Subject: [msp430] Re: High temperature opperation of the MSP430

Thank you for your help!

old cow yellow,
I was not aware of the "Alzheimer" effect on flash at higher temperatures. I knew that programming and erasing flash at high temperatures did not work but I did not know that the flash memory would randomly flip bits at high temp. Do you know of a study or paper that elaborates on this effect?

Adriano,
Reliability is a very general term. Many applications of micro controllers cannot allow any failures after several years of continuous operation and are expected to last well over 50,000 hours.
My application only needs to work for 100 hours at high temperature.

Has anyone actually taken a MSP 430 above 100C for any length of time?

Thank you again for your help.
--- In m..., "old_cow_yellow"
wrote:
>
> High temperature also causes Alzheimer. What is already burned into to
> the Flash memory will disappear a lot sooner.
>
> --- In m..., Adriano Caye wrote:
> >
> > MSP430 shouldn't be adequate for your application. The MSP430F147,
for
> > example, supports up to 85 C. It can work on 180 C, but its
operation
> > will not be reliable.
> >
> > Regards,
> > Adriano.
> >
> > chrismholt123 wrote:
> >
> > > I am curious as to the highest temperature that the MSP 430 can be
> > > successfully operated at. Does anyone have any experience
testing the
> > > micro in extreme environments?
> > >
> > > I searched your forum and found a test that was done by a forum
member
> > > of the flash memory at temperatures up to 100C.
> > >
> > > Has anyone tested temperatures higher than this?
> > >
> > > I do not need to write to flash but just need the micro to read an
> > > external ADC via spi and record the readings to external memory.
> > >
> > > I need a micro that will operate at 180C. I know that the PIC 18C
> > > will reliably (over 1000 hours) operate at this temperature as
well as
> > > a the Motorola MC68332.
> > >
> > > I like the development tools and emulation options provided by
the MSP
> > > 430 as well as the incredible current consumption of the part.
> > >
> > > Does anyone have experience using this micro at high temp?
> > >
> > > I am a University Student at the University of Alberta and any
> > > help could save me a few $$ on my project by not needing to test
> > > the
MSP430
> > > myself.
> > >
> > > Thank you very much for your help!
> > >
> > >
> > > --
> > > Esta mensagem foi verificada pelo sistema de antivus e
> > > acredita-se estar livre de perigo.
> >
> >
> > --
> > Esta mensagem foi verificada pelo sistema de antivus e
> > acredita-se estar livre de perigo.
>



I should have said "Flash memory data retention time" is shorter at
higher temperature.

TI document SLAA334.pdf has a paragraph describing this.

The MSP430F data retention time of 100 years is for 25C. According to
Black's equation (which TI calls Arrhenius equation), this translates
to 16.6 years at 50C.

Please double check, but I think at 180C the data retention is 260
hours. But beside Flash data retention, there are a lot of other
issues at such temperature.

--- In m..., "Garlick, Colin" wrote:
>
>
> You are scratching the surface of a very detailed subject, with many
factors which
> interact. But to answer your basic question I believe you want to
look at Black's equation.
> Some information can be found on Wikipedia:-
http://en.wikipedia.org/wiki/Black's_equation
>
> At a basic level, the IAR compiler has the ability to generate a ROM
integrity checksum at
> Link-time, which you could call at regular intervals to detect such
failures.
>
> To go one stage further, Error Correction Code, can be held in RAM
(through the vulnerable
> high temperature time), which would allow you patch a limited number
of bit errors as and
> when you need the data (or relocatable-code segments you might also
need - luckily IAR can
> also generate Position Independent Code segments you can overlay),
but which can more permanently
> be subsequently fixed in Flash, at a later time when the temperature
comes back into spec.
>
> Regards,
>
> Colin,
>
> -----Original Message-----
> From: m... [mailto:m...] On
Behalf Of chrismholt123
> Sent: 05 October 2007 16:22
> To: m...
> Subject: [msp430] Re: High temperature opperation of the MSP430
>
> Thank you for your help!
>
> old cow yellow,
> I was not aware of the "Alzheimer" effect on flash at higher
temperatures. I knew that programming and erasing flash at high
temperatures did not work but I did not know that the flash memory
would randomly flip bits at high temp. Do you know of a study or
paper that elaborates on this effect?
>
> Adriano,
> Reliability is a very general term. Many applications of micro
controllers cannot allow any failures after several years of
continuous operation and are expected to last well over 50,000 hours.
> My application only needs to work for 100 hours at high temperature.
>
> Has anyone actually taken a MSP 430 above 100C for any length of time?
>
> Thank you again for your help.
>
>
> --- In m..., "old_cow_yellow"
> wrote:
> >
> > High temperature also causes Alzheimer. What is already burned
into to
> > the Flash memory will disappear a lot sooner.
> >
> > --- In m..., Adriano Caye wrote:
> > >
> > > MSP430 shouldn't be adequate for your application. The MSP430F147,
> for
> > > example, supports up to 85 C. It can work on 180 C, but its
> operation
> > > will not be reliable.
> > >
> > > Regards,
> > > Adriano.
> > >
> > > chrismholt123 wrote:
> > >
> > > > I am curious as to the highest temperature that the MSP 430
can be
> > > > successfully operated at. Does anyone have any experience
> testing the
> > > > micro in extreme environments?
> > > >
> > > > I searched your forum and found a test that was done by a forum
> member
> > > > of the flash memory at temperatures up to 100C.
> > > >
> > > > Has anyone tested temperatures higher than this?
> > > >
> > > > I do not need to write to flash but just need the micro to
read an
> > > > external ADC via spi and record the readings to external memory.
> > > >
> > > > I need a micro that will operate at 180C. I know that the PIC 18C
> > > > will reliably (over 1000 hours) operate at this temperature as
> well as
> > > > a the Motorola MC68332.
> > > >
> > > > I like the development tools and emulation options provided by
> the MSP
> > > > 430 as well as the incredible current consumption of the part.
> > > >
> > > > Does anyone have experience using this micro at high temp?
> > > >
> > > > I am a University Student at the University of Alberta and any
> > > > help could save me a few $$ on my project by not needing to test
> > > > the
> MSP430
> > > > myself.
> > > >
> > > > Thank you very much for your help!
> > > >
> > > >
> > > > --
> > > > Esta mensagem foi verificada pelo sistema de antivus e
> > > > acredita-se estar livre de perigo.
> > >
> > >
> > > --
> > > Esta mensagem foi verificada pelo sistema de antivus e
> > > acredita-se estar livre de perigo.
> > >
> >
>
>
>
>
>
>
>
>
I just thought a somewhat less theoretical response to this thread might be called for. For the last few years, we have been using MSP430 chips within oil drilling bits. Our systems are rigorously tested at high vibration levels and at temperatures of 150 degrees C for extended periods (hundreds of hours per run) by the bit manufacturer. They've come through the testing and hundreds of field runs just fine.

Obviously, when designing boards for such applications, you need to make sure that all components are rated for high temperature operation. Also, for our high vibration requirements, we also have to be careful with layout and fabrication.

The simple answer is that if you get the right chip, MSP430 processors work quite well at high temperatures.

---------------------------------
Fussy? Opinionated? Impossible to please? Perfect. Join Yahoo!'s user panel and lay it on us.
Interesting. Does the MSP430 you use in this application support 150
degrees C? Or did you have to cool down the MSP to a more suitable
temperature by using a cooler and/or heat dissipator?

Adriano.

Joseph A. Vrba wrote:

> I just thought a somewhat less theoretical response to this thread
> might be called for. For the last few years, we have been using MSP430
> chips within oil drilling bits. Our systems are rigorously tested at
> high vibration levels and at temperatures of 150 degrees C for
> extended periods (hundreds of hours per run) by the bit manufacturer.
> They've come through the testing and hundreds of field runs just fine.
>
> Obviously, when designing boards for such applications, you need to
> make sure that all components are rated for high temperature
> operation. Also, for our high vibration requirements, we also have to
> be careful with layout and fabrication.
>
> The simple answer is that if you get the right chip, MSP430 processors
> work quite well at high temperatures.
>
> ---------------------------------
> Fussy? Opinionated? Impossible to please? Perfect. Join Yahoo!'s user
> panel and lay it on us.
>
>
> --
> Esta mensagem foi verificada pelo sistema de antivus e
> acredita-se estar livre de perigo.
--
Esta mensagem foi verificada pelo sistema de antivus e
acredita-se estar livre de perigo.
Wow, thank you Joseph! That was exactly the answer that I was looking
for. What packaging/model of MSP 430 did your company use for the
drilling application?

Also thank you for the awesome ideas about reprogramming the flash
memory after high temperature operation to ensure that bit flipping is
less likely. This is a great idea for improving the reliability of
the micro controller if you have some extra memory available or have a
copy of the program stored in more reliable external memory. Using a
flash micro has some big advantages!

Thank you again for the great help. This message board is really
incredible! It is awesome to get help from so many knowledgeable
members!

Thank you again,

Chris

--- In m..., "Joseph A. Vrba" wrote:
>
> I just thought a somewhat less theoretical response to this thread
might be called for. For the last few years, we have been using
MSP430 chips within oil drilling bits. Our systems are rigorously
tested at high vibration levels and at temperatures of 150 degrees C
for extended periods (hundreds of hours per run) by the bit
manufacturer. They've come through the testing and hundreds of field
runs just fine.
>
> Obviously, when designing boards for such applications, you need
to make sure that all components are rated for high temperature
operation. Also, for our high vibration requirements, we also have to
be careful with layout and fabrication.
>
> The simple answer is that if you get the right chip, MSP430
processors work quite well at high temperatures.
>
>
> ---------------------------------
> Fussy? Opinionated? Impossible to please? Perfect. Join Yahoo!'s
user panel and lay it on us.
>
>
>
I thought I'd continue this thread to answer everyone else...

In this application, we started with the MSP430F149 and used is successfully in around 100 units spanning a period of over two years. Recently we've switched to the MSP430FG4617 for increased memory capacity (and chip availability). That board has just completed all of its formal stress testing as is currently being used in field trials. (We added quite a few more sensors as part of the new design.)

There's no extra cooling on the chip. Although we don't typically see temperatures that high in normal downhole runs, the 150 degree C requirement comes from our customer. The formal testing for both high temp operation, temperature cycling and vibration is performed by an independent subsidiary of our customer.

---------------------------------
Catch up on fall's hot new shows on Yahoo! TV. Watch previews, get listings, and more!
Thanks for sharing the info, Joseph. I guess the operating temperature
in the field shouldn't exceed the 85 degress C of the datasheet, for
being working all this time. Or maybe your board doesn't work all the
time...

And what about the mechanical vibration? Are there layout rules to
diminish its effects? Are they the same recommendations to avoid EMI?
Just curious about it.

Regards,
Adriano.

Joseph A. Vrba wrote:

> I thought I'd continue this thread to answer everyone else...
>
> In this application, we started with the MSP430F149 and used is
> successfully in around 100 units spanning a period of over two years.
> Recently we've switched to the MSP430FG4617 for increased memory
> capacity (and chip availability). That board has just completed all of
> its formal stress testing as is currently being used in field trials.
> (We added quite a few more sensors as part of the new design.)
>
> There's no extra cooling on the chip. Although we don't typically see
> temperatures that high in normal downhole runs, the 150 degree C
> requirement comes from our customer. The formal testing for both high
> temp operation, temperature cycling and vibration is performed by an
> independent subsidiary of our customer.
>
> ---------------------------------
> Catch up on fall's hot new shows on Yahoo! TV. Watch previews, get
> listings, and more!
>
>
> --
> Esta mensagem foi verificada pelo sistema de antivus e
> acredita-se estar livre de perigo.
--
Esta mensagem foi verificada pelo sistema de antivus e
acredita-se estar livre de perigo.