Hi all. We heve in company many devices witch MSP430. But actually we have a big
problem. When EMI are big (electric industry) procesor is hand on. In program
of course is setting WDT in watchdog mode, but this circuit NOT restart a
procesor. Processor is DEAD and clients are Very Angry.
Anybody have some experience witch MSP430 in EMI ?
Thanks for advice !!!
--------------------------------
Doaduj swoje konto. Bez zdrapywania, bez kod, na dowoln kwot
http://link.interia.pl/f19f6 >>> - Szybkie i atwe doadowania GSM
EMI + WDT BIG Problem
Started by ●January 23, 2007
Reply by ●January 23, 20072007-01-23
I've found that the MSP "may" be more prone to EMI issues than say a PIC
operating at 5V. However, in many cases it was my own poor design that
didn't account for the EMI issues.
Flaky operation, random failures that happen in the field but work great in the lab, mysterious resets may be symptoms of EMI issues.
We use external WDT's on the MSP but even these are only as good as the code you write. If the code keeps patting the dog even if things are not okay then you don't get a reset.
It sounds like you may have a big problem that cannot be answered simply by an email. I would recommend finding someone that knows how to solve EMI issues directly. Probably someone with years in that field. Your solution may well involve a complete new design with shielding, filters, multilayer boards, small monkeys, and careful consideration and understanding of the true electrical environment. Okay, I just threw the small monkeys in there. I can recommend several EMI trouble shooters if you are interested.
Dan M
-----Original Message-----
From: m... [mailto:m...] On Behalf Of lukasz_krysiewicz
Sent: Tuesday, January 23, 2007 4:48 PM
To: m...
Subject: [msp430] EMI + WDT BIG Problem
Hi all. We heve in company many devices witch MSP430. But actually we have a big problem. When EMI are big (electric industry) procesor is hand on. In program of course is setting WDT in watchdog mode, but this circuit NOT restart a procesor. Processor is DEAD and clients are Very Angry.
Anybody have some experience witch MSP430 in EMI ?
Thanks for advice !!!
--------------------------------
Doaduj swoje konto. Bez zdrapywania, bez kod, na dowoln kwot
http://link.interia.pl/f19f6 >>> - Szybkie i atwe doadowania GSM
Flaky operation, random failures that happen in the field but work great in the lab, mysterious resets may be symptoms of EMI issues.
We use external WDT's on the MSP but even these are only as good as the code you write. If the code keeps patting the dog even if things are not okay then you don't get a reset.
It sounds like you may have a big problem that cannot be answered simply by an email. I would recommend finding someone that knows how to solve EMI issues directly. Probably someone with years in that field. Your solution may well involve a complete new design with shielding, filters, multilayer boards, small monkeys, and careful consideration and understanding of the true electrical environment. Okay, I just threw the small monkeys in there. I can recommend several EMI trouble shooters if you are interested.
Dan M
-----Original Message-----
From: m... [mailto:m...] On Behalf Of lukasz_krysiewicz
Sent: Tuesday, January 23, 2007 4:48 PM
To: m...
Subject: [msp430] EMI + WDT BIG Problem
Hi all. We heve in company many devices witch MSP430. But actually we have a big problem. When EMI are big (electric industry) procesor is hand on. In program of course is setting WDT in watchdog mode, but this circuit NOT restart a procesor. Processor is DEAD and clients are Very Angry.
Anybody have some experience witch MSP430 in EMI ?
Thanks for advice !!!
--------------------------------
Doaduj swoje konto. Bez zdrapywania, bez kod, na dowoln kwot
http://link.interia.pl/f19f6 >>> - Szybkie i atwe doadowania GSM
Reply by ●January 23, 20072007-01-23
f you mention "small mo" that is a reference to some of the original
animal tracking project going back over 20 years. Literally having
something running code while on a monkey's back is quite a static
electricity (i.e. EMI) problem where you need to have a lot taken care
of before you strap it on. Oddly, MSP430 is a great microcontroller for
the "augmented biological entity tracking" market.
Steve
an Muzzey wrote:
> I've found that the MSP "may" be more prone to EMI issues than say a PIC operating at 5V. However, in many cases it was my own poor design that didn't account for the EMI issues.
>
> Flaky operation, random failures that happen in the field but work great in the lab, mysterious resets may be symptoms of EMI issues.
>
> We use external WDT's on the MSP but even these are only as good as the code you write. If the code keeps patting the dog even if things are not okay then you don't get a reset.
>
> It sounds like you may have a big problem that cannot be answered simply by an email. I would recommend finding someone that knows how to solve EMI issues directly. Probably someone with years in that field. Your solution may well involve a complete new design with shielding, filters, multilayer boards, small monkeys, and careful consideration and understanding of the true electrical environment. Okay, I just threw the small monkeys in there. I can recommend several EMI trouble shooters if you are interested.
>
> Dan M
>
> -----Original Message-----
> From: m... [mailto:m...] On Behalf Of lukasz_krysiewicz
> Sent: Tuesday, January 23, 2007 4:48 PM
> To: m...
> Subject: [msp430] EMI + WDT BIG Problem
> Hi all. We heve in company many devices witch MSP430. But actually we have a big problem. When EMI are big (electric industry) procesor is hand on. In program of course is setting WDT in watchdog mode, but this circuit NOT restart a procesor. Processor is DEAD and clients are Very Angry.
>
> Anybody have some experience witch MSP430 in EMI ?
>
> Thanks for advice !!!
> --------------------------------
> Doaduj swoje konto. Bez zdrapywania, bez kod, na dowoln kwot
> http://link.interia.pl/f19f6 >>> - Szybkie i atwe doadowania GSM
>
>
animal tracking project going back over 20 years. Literally having
something running code while on a monkey's back is quite a static
electricity (i.e. EMI) problem where you need to have a lot taken care
of before you strap it on. Oddly, MSP430 is a great microcontroller for
the "augmented biological entity tracking" market.
Steve
an Muzzey wrote:
> I've found that the MSP "may" be more prone to EMI issues than say a PIC operating at 5V. However, in many cases it was my own poor design that didn't account for the EMI issues.
>
> Flaky operation, random failures that happen in the field but work great in the lab, mysterious resets may be symptoms of EMI issues.
>
> We use external WDT's on the MSP but even these are only as good as the code you write. If the code keeps patting the dog even if things are not okay then you don't get a reset.
>
> It sounds like you may have a big problem that cannot be answered simply by an email. I would recommend finding someone that knows how to solve EMI issues directly. Probably someone with years in that field. Your solution may well involve a complete new design with shielding, filters, multilayer boards, small monkeys, and careful consideration and understanding of the true electrical environment. Okay, I just threw the small monkeys in there. I can recommend several EMI trouble shooters if you are interested.
>
> Dan M
>
> -----Original Message-----
> From: m... [mailto:m...] On Behalf Of lukasz_krysiewicz
> Sent: Tuesday, January 23, 2007 4:48 PM
> To: m...
> Subject: [msp430] EMI + WDT BIG Problem
> Hi all. We heve in company many devices witch MSP430. But actually we have a big problem. When EMI are big (electric industry) procesor is hand on. In program of course is setting WDT in watchdog mode, but this circuit NOT restart a procesor. Processor is DEAD and clients are Very Angry.
>
> Anybody have some experience witch MSP430 in EMI ?
>
> Thanks for advice !!!
> --------------------------------
> Doaduj swoje konto. Bez zdrapywania, bez kod, na dowoln kwot
> http://link.interia.pl/f19f6 >>> - Szybkie i atwe doadowania GSM
>
>
Reply by ●January 23, 20072007-01-23
The MSP430 from what I have read needs a pretty stable power supply.
If you have a lot of electrical noise in your power it can screw up the
processor operation. The watchdog will try to reset to get it working
again.
If you are blowing MSP430 then you have that noise with an amplitude
over 3.6V or less than 0V on the power or any line.
My first guess is your power supply. It is not filtering the noise out.
My second guess is that one of the signals is going too high or
negative.
You need to get out the oscilloscope to see what is happening.
Kip
On Tue, 2007-01-23 at 23:47 +0100, lukasz_krysiewicz wrote:
> Hi all. We heve in company many devices witch MSP430. But actually we have a big problem. When EMI are big (electric industry) procesor is hand on. In program of course is setting WDT in watchdog mode, but this circuit NOT restart a procesor. Processor is DEAD and clients are Very Angry.
>
> Anybody have some experience witch MSP430 in EMI ?
>
> Thanks for advice !!!
> --------------------------------
> Doładuj swoje konto. Bez zdrapywania, bez kodów, na dowolną kwotę.
> http://link.interia.pl/f19f6 >>> - Szybkie i łatwe doładowania GSM
>
> Yahoo! Groups Links
--
Kipton Moravec
If you have a lot of electrical noise in your power it can screw up the
processor operation. The watchdog will try to reset to get it working
again.
If you are blowing MSP430 then you have that noise with an amplitude
over 3.6V or less than 0V on the power or any line.
My first guess is your power supply. It is not filtering the noise out.
My second guess is that one of the signals is going too high or
negative.
You need to get out the oscilloscope to see what is happening.
Kip
On Tue, 2007-01-23 at 23:47 +0100, lukasz_krysiewicz wrote:
> Hi all. We heve in company many devices witch MSP430. But actually we have a big problem. When EMI are big (electric industry) procesor is hand on. In program of course is setting WDT in watchdog mode, but this circuit NOT restart a procesor. Processor is DEAD and clients are Very Angry.
>
> Anybody have some experience witch MSP430 in EMI ?
>
> Thanks for advice !!!
> --------------------------------
> Doładuj swoje konto. Bez zdrapywania, bez kodów, na dowolną kwotę.
> http://link.interia.pl/f19f6 >>> - Szybkie i łatwe doładowania GSM
>
> Yahoo! Groups Links
--
Kipton Moravec
Reply by ●January 23, 20072007-01-23
> Processor is DEAD and clients are Very Angry.
Can I talk to your mummy ?
Best Regards,
Kris
-----Original Message-----
From: m... [mailto:m...] On Behalf Of lukasz_krysiewicz
Sent: Wednesday, 24 January 2007 9:48 AM
To: m...
Subject: [msp430] EMI + WDT BIG Problem
Hi all. We heve in company many devices witch MSP430. But actually we have a big problem. When EMI
are big (electric industry) procesor is hand on. In program of course is setting WDT in
watchdog mode, but this circuit NOT restart a procesor. Processor is DEAD and clients are Very
Angry.
Anybody have some experience witch MSP430 in EMI ?
Thanks for advice !!!
Can I talk to your mummy ?
Best Regards,
Kris
-----Original Message-----
From: m... [mailto:m...] On Behalf Of lukasz_krysiewicz
Sent: Wednesday, 24 January 2007 9:48 AM
To: m...
Subject: [msp430] EMI + WDT BIG Problem
Hi all. We heve in company many devices witch MSP430. But actually we have a big problem. When EMI
are big (electric industry) procesor is hand on. In program of course is setting WDT in
watchdog mode, but this circuit NOT restart a procesor. Processor is DEAD and clients are Very
Angry.
Anybody have some experience witch MSP430 in EMI ?
Thanks for advice !!!
Reply by ●January 24, 20072007-01-24
> Processor is DEAD and clients are Very Angry.
Can I talk to your mummy ?
Best Regards,
Kris
Can I talk to you ?
:)
Thansk for any advice. Today I will be testing devices, and trying solved a problem.
--------------------------------
Doaduj swoje konto. Bez zdrapywania, bez kod, na dowoln kwot
http://link.interia.pl/f19f6 >>> - Szybkie i atwe doadowania GSM
Can I talk to your mummy ?
Best Regards,
Kris
Can I talk to you ?
:)
Thansk for any advice. Today I will be testing devices, and trying solved a problem.
--------------------------------
Doaduj swoje konto. Bez zdrapywania, bez kod, na dowoln kwot
http://link.interia.pl/f19f6 >>> - Szybkie i atwe doadowania GSM
Reply by ●January 24, 20072007-01-24
I had found the same problem in an industrial application.
In my case, the application have a frequency converter and a
lot of 220Vac relays (~25). The problem was solved adding
snubbers to the contactors (the contactor's manufacturer have
a snubber specific to each contactor type).
-Paulo.
> -----Mensagem original-----
> De: m... [mailto:m...] Em nome de
> lukasz_krysiewicz
> Enviada em: quarta-feira, 24 de janeiro de 2007 04:09
> Para: m...
> Assunto: Re: [msp430] EMI + WDT BIG Problem
>
>
> > Processor is DEAD and clients are Very Angry.
>
> Can I talk to your mummy ?
>
> Best Regards,
> Kris
>
> Can I talk to you ?
>
> :)
>
> Thansk for any advice. Today I will be testing devices, and trying solved
> a problem.
>
> --------------------------------
> Doaduj swoje konto. Bez zdrapywania, bez kod, na dowoln kwot
> http://link.interia.pl/f19f6 >>> - Szybkie i atwe doadowania GSM
>
>
>
>
>
>
>
>
In my case, the application have a frequency converter and a
lot of 220Vac relays (~25). The problem was solved adding
snubbers to the contactors (the contactor's manufacturer have
a snubber specific to each contactor type).
-Paulo.
> -----Mensagem original-----
> De: m... [mailto:m...] Em nome de
> lukasz_krysiewicz
> Enviada em: quarta-feira, 24 de janeiro de 2007 04:09
> Para: m...
> Assunto: Re: [msp430] EMI + WDT BIG Problem
>
>
> > Processor is DEAD and clients are Very Angry.
>
> Can I talk to your mummy ?
>
> Best Regards,
> Kris
>
> Can I talk to you ?
>
> :)
>
> Thansk for any advice. Today I will be testing devices, and trying solved
> a problem.
>
> --------------------------------
> Doaduj swoje konto. Bez zdrapywania, bez kod, na dowoln kwot
> http://link.interia.pl/f19f6 >>> - Szybkie i atwe doadowania GSM
>
>
>
>
>
>
>
>
Reply by ●January 24, 20072007-01-24
>I've found that the MSP "may" be more prone to
EMI issues than say a PIC
>operating at 5V. However, in many cases it was my own poor design that
>didn't account for the EMI issues.
>
>Flaky operation, random failures that happen in the field but work great in
>the lab, mysterious resets may be symptoms of EMI issues.
>
>We use external WDT's on the MSP but even these are only as good as the
>code you write. If the code keeps patting the dog even if things are not
>okay then you don't get a reset.
>
In addition to what Dan mentions (he and I work together), I wanted to state
that our processor also was dead in that even the internal watchdog would
not reset the part. We had to turn power off then back on to get it to
restart again. So, we went with an external WDT circuit which finally
worked like a WDT should work. I was as if when the noise came in, it would
turn off all clocks, including the clock to the internal WDT circuit and it
would totally freeze. I don't believe the internal WDT on the MSP430 uses
any analog timing components, but rather relies on a digital oscillator to
work.
But, he is right, we did have a lot of noise situations we needed to take
care of. So the 200% solution to the problem was to clean up the noise and
put in a good WDT circuit. We got an education on EMI/EMS and found several
areas where we could make improvements in our design.
Lou
_________________________________________________________________
Turn searches into helpful donations. Make your search count.
http://click4thecause.live.com/search/charity/default.aspx?source=hmemtagline_donation&FORM=WLMTAG
>operating at 5V. However, in many cases it was my own poor design that
>didn't account for the EMI issues.
>
>Flaky operation, random failures that happen in the field but work great in
>the lab, mysterious resets may be symptoms of EMI issues.
>
>We use external WDT's on the MSP but even these are only as good as the
>code you write. If the code keeps patting the dog even if things are not
>okay then you don't get a reset.
>
In addition to what Dan mentions (he and I work together), I wanted to state
that our processor also was dead in that even the internal watchdog would
not reset the part. We had to turn power off then back on to get it to
restart again. So, we went with an external WDT circuit which finally
worked like a WDT should work. I was as if when the noise came in, it would
turn off all clocks, including the clock to the internal WDT circuit and it
would totally freeze. I don't believe the internal WDT on the MSP430 uses
any analog timing components, but rather relies on a digital oscillator to
work.
But, he is right, we did have a lot of noise situations we needed to take
care of. So the 200% solution to the problem was to clean up the noise and
put in a good WDT circuit. We got an education on EMI/EMS and found several
areas where we could make improvements in our design.
Lou
_________________________________________________________________
Turn searches into helpful donations. Make your search count.
http://click4thecause.live.com/search/charity/default.aspx?source=hmemtagline_donation&FORM=WLMTAG
Reply by ●January 24, 20072007-01-24
I've had issues with ESD corrupting RAM on an older design right around
4kV contact. I'm not the only one - Jack Ganssle wrote an article that
mentioned one of his friends dealing with the msp430 and ESD.
http://www.ganssle.com/tem/tem71.pdf . It is an interesting read.
At any rate, I'm in agreement with Lou - an external WDT, stop the noise
with hardware, and make provisions for error detection and correction in
your software.
Take every test that you need to pass for certification in terms of
susceptibility, reproduce the conditions as best you can but at elevated
levels ( 200% ;) ) and try to kill your board. Once satisfied, send it
off for certification.
A painful, but worthwhile lesson.
Michael
________________________________
From: m... [mailto:m...] On Behalf
Of Lou C
Sent: Wednesday, January 24, 2007 6:17 AM
To: m...
Subject: RE: [msp430] EMI + WDT BIG Problem
>I've found that the MSP "may" be more prone to EMI issues than say a
PIC
>operating at 5V. However, in many cases it was my own poor design that
>didn't account for the EMI issues.
>
>Flaky operation, random failures that happen in the field but work
great in
>the lab, mysterious resets may be symptoms of EMI issues.
>
>We use external WDT's on the MSP but even these are only as good as the
>code you write. If the code keeps patting the dog even if things are
not
>okay then you don't get a reset.
>
In addition to what Dan mentions (he and I work together), I wanted to
state
that our processor also was dead in that even the internal watchdog
would
not reset the part. We had to turn power off then back on to get it to
restart again. So, we went with an external WDT circuit which finally
worked like a WDT should work. I was as if when the noise came in, it
would
turn off all clocks, including the clock to the internal WDT circuit and
it
would totally freeze. I don't believe the internal WDT on the MSP430
uses
any analog timing components, but rather relies on a digital oscillator
to
work.
But, he is right, we did have a lot of noise situations we needed to
take
care of. So the 200% solution to the problem was to clean up the noise
and
put in a good WDT circuit. We got an education on EMI/EMS and found
several
areas where we could make improvements in our design.
Lou
__________________________________________________________
Turn searches into helpful donations. Make your search count.
http://click4thecause.live.com/search/charity/default.aspx?source=hmemta
gline_donation&FORM=WLMTAG
agline_donation&FORM=WLMTAG>
4kV contact. I'm not the only one - Jack Ganssle wrote an article that
mentioned one of his friends dealing with the msp430 and ESD.
http://www.ganssle.com/tem/tem71.pdf . It is an interesting read.
At any rate, I'm in agreement with Lou - an external WDT, stop the noise
with hardware, and make provisions for error detection and correction in
your software.
Take every test that you need to pass for certification in terms of
susceptibility, reproduce the conditions as best you can but at elevated
levels ( 200% ;) ) and try to kill your board. Once satisfied, send it
off for certification.
A painful, but worthwhile lesson.
Michael
________________________________
From: m... [mailto:m...] On Behalf
Of Lou C
Sent: Wednesday, January 24, 2007 6:17 AM
To: m...
Subject: RE: [msp430] EMI + WDT BIG Problem
>I've found that the MSP "may" be more prone to EMI issues than say a
PIC
>operating at 5V. However, in many cases it was my own poor design that
>didn't account for the EMI issues.
>
>Flaky operation, random failures that happen in the field but work
great in
>the lab, mysterious resets may be symptoms of EMI issues.
>
>We use external WDT's on the MSP but even these are only as good as the
>code you write. If the code keeps patting the dog even if things are
not
>okay then you don't get a reset.
>
In addition to what Dan mentions (he and I work together), I wanted to
state
that our processor also was dead in that even the internal watchdog
would
not reset the part. We had to turn power off then back on to get it to
restart again. So, we went with an external WDT circuit which finally
worked like a WDT should work. I was as if when the noise came in, it
would
turn off all clocks, including the clock to the internal WDT circuit and
it
would totally freeze. I don't believe the internal WDT on the MSP430
uses
any analog timing components, but rather relies on a digital oscillator
to
work.
But, he is right, we did have a lot of noise situations we needed to
take
care of. So the 200% solution to the problem was to clean up the noise
and
put in a good WDT circuit. We got an education on EMI/EMS and found
several
areas where we could make improvements in our design.
Lou
__________________________________________________________
Turn searches into helpful donations. Make your search count.
http://click4thecause.live.com/search/charity/default.aspx?source=hmemta
gline_donation&FORM=WLMTAG
agline_donation&FORM=WLMTAG>
