Hi,
I'm having some trouble with an LPC2103 design related to higher
temperatures; the design becomes unstable and therefore unusuable at
higher temperatures (somewhere between 60 and 70 degrees celsius).
According to the datasheet it should do +85, and all the components i'm
using around it should make this aswell.
The strange thing is that some hand assembled boards do make this
temperature range, but the initial serie of professionally produced
boards don't. I've been swapping most components but the MCU, which
now
seems to be the cultprint, which atm doesn't make sense to me.
Anyone having some experience related to higher temperatures and the
LPC2103 or related, and wanting to share some?
Regards,
-Bastiaan
LPC2103 temperature range
Started by ●July 17, 2012
Reply by ●July 17, 20122012-07-17
Have you tried freezing it as well? This could be related to soldering
problems that are showing up with temperature cycling, and it would be
more likely with the finer pitch parts.
Gently flexing the board may also show soldering issues.
Mike
-----Original Message-----
From: l... [mailto:l...]On Behalf
Of Bastiaan van Kesteren
Sent: Tuesday, July 17, 2012 7:01 AM
To: l...
Subject: [lpc2000] LPC2103 temperature range
Hi,
I'm having some trouble with an LPC2103 design related to higher
temperatures; the design becomes unstable and therefore unusuable at
higher temperatures (somewhere between 60 and 70 degrees celsius).
According to the datasheet it should do +85, and all the components i'm
using around it should make this aswell.
The strange thing is that some hand assembled boards do make this
temperature range, but the initial serie of professionally produced
boards don't. I've been swapping most components but the MCU, which now
seems to be the cultprint, which atm doesn't make sense to me.
Anyone having some experience related to higher temperatures and the
LPC2103 or related, and wanting to share some?
Regards,
-Bastiaan
problems that are showing up with temperature cycling, and it would be
more likely with the finer pitch parts.
Gently flexing the board may also show soldering issues.
Mike
-----Original Message-----
From: l... [mailto:l...]On Behalf
Of Bastiaan van Kesteren
Sent: Tuesday, July 17, 2012 7:01 AM
To: l...
Subject: [lpc2000] LPC2103 temperature range
Hi,
I'm having some trouble with an LPC2103 design related to higher
temperatures; the design becomes unstable and therefore unusuable at
higher temperatures (somewhere between 60 and 70 degrees celsius).
According to the datasheet it should do +85, and all the components i'm
using around it should make this aswell.
The strange thing is that some hand assembled boards do make this
temperature range, but the initial serie of professionally produced
boards don't. I've been swapping most components but the MCU, which now
seems to be the cultprint, which atm doesn't make sense to me.
Anyone having some experience related to higher temperatures and the
LPC2103 or related, and wanting to share some?
Regards,
-Bastiaan
Reply by ●July 18, 20122012-07-18
I tested LPC2106 used in 70 or so boards with lots of other components... OK up
to about 60C. I would not try for 85C with commercial parts.
--- In l..., Bastiaan van Kesteren wrote:
>
> Hi,
>
> I'm having some trouble with an LPC2103 design related to higher
> temperatures; the design becomes unstable and therefore unusuable at
> higher temperatures (somewhere between 60 and 70 degrees celsius).
>
> According to the datasheet it should do +85, and all the components i'm
> using around it should make this aswell.
>
> The strange thing is that some hand assembled boards do make this
> temperature range, but the initial serie of professionally produced
> boards don't. I've been swapping most components but the MCU, which now
> seems to be the cultprint, which atm doesn't make sense to me.
> Anyone having some experience related to higher temperatures and the
> LPC2103 or related, and wanting to share some?
>
> Regards,
> -Bastiaan
>
--- In l..., Bastiaan van Kesteren wrote:
>
> Hi,
>
> I'm having some trouble with an LPC2103 design related to higher
> temperatures; the design becomes unstable and therefore unusuable at
> higher temperatures (somewhere between 60 and 70 degrees celsius).
>
> According to the datasheet it should do +85, and all the components i'm
> using around it should make this aswell.
>
> The strange thing is that some hand assembled boards do make this
> temperature range, but the initial serie of professionally produced
> boards don't. I've been swapping most components but the MCU, which now
> seems to be the cultprint, which atm doesn't make sense to me.
> Anyone having some experience related to higher temperatures and the
> LPC2103 or related, and wanting to share some?
>
> Regards,
> -Bastiaan
>
Reply by ●July 24, 20122012-07-24
Hi,
Thanks for the replies!
I've been doing some more testing here, and am starting to suspect the
soldering-process.
I now have two boards of the professionally produced series which have
the crash-problem i'm having close after powerup (between 10 seconds and
one minute), even at room temperature (which is a tad high here right
now, but... :) )
To clarify what I mean with 'crash-problem'; the product has serial
output on an uart; when the board crashes, the baudrate of the
uart-output goes way off (haven't measured the baudrate, but output is
garbage). Other perhiperals stop working aswell, I assume speeds are
changing there aswell.
Boards running in hot locations seem to be more crash-prone, hence my
initial temperature range suspicion.
I attempted a reflow of the MCU on the 10-second-crash board, which made
matters worse; it did not run properly at all, or only for a few seconds.
So, i removed the MCU and hand-soldered a new one. And behold, the board
runs... I only have it running for about 30 minutes now, but anyways the
MCU seems to be the problem.
So, I suspect the MCU has got too hot during the production-process,
therby damaging it in a half/half way, causing these random crashes I'm
seeing now; some boards crash after less than a minute, some a few days,
some run fine for months.
Is this making sense? Anyone seen something like this?
I'll contact my PCB assembler in the morning, see what they can say
about this...
Regards,
-Bastiaan
Op 18-7-2012 23:01, stevec schreef:
> I tested LPC2106 used in 70 or so boards with lots of other
> components... OK up to about 60C. I would not try for 85C with
> commercial parts.
>
> --- In l... ,
> Bastiaan van Kesteren wrote:
> >
> > Hi,
> >
> > I'm having some trouble with an LPC2103 design related to higher
> > temperatures; the design becomes unstable and therefore unusuable at
> > higher temperatures (somewhere between 60 and 70 degrees celsius).
> >
> > According to the datasheet it should do +85, and all the components i'm
> > using around it should make this aswell.
> >
> > The strange thing is that some hand assembled boards do make this
> > temperature range, but the initial serie of professionally produced
> > boards don't. I've been swapping most components but the MCU, which now
> > seems to be the cultprint, which atm doesn't make sense to me.
> >
> >
> > Anyone having some experience related to higher temperatures and the
> > LPC2103 or related, and wanting to share some?
> >
> > Regards,
> > -Bastiaan
> >
Thanks for the replies!
I've been doing some more testing here, and am starting to suspect the
soldering-process.
I now have two boards of the professionally produced series which have
the crash-problem i'm having close after powerup (between 10 seconds and
one minute), even at room temperature (which is a tad high here right
now, but... :) )
To clarify what I mean with 'crash-problem'; the product has serial
output on an uart; when the board crashes, the baudrate of the
uart-output goes way off (haven't measured the baudrate, but output is
garbage). Other perhiperals stop working aswell, I assume speeds are
changing there aswell.
Boards running in hot locations seem to be more crash-prone, hence my
initial temperature range suspicion.
I attempted a reflow of the MCU on the 10-second-crash board, which made
matters worse; it did not run properly at all, or only for a few seconds.
So, i removed the MCU and hand-soldered a new one. And behold, the board
runs... I only have it running for about 30 minutes now, but anyways the
MCU seems to be the problem.
So, I suspect the MCU has got too hot during the production-process,
therby damaging it in a half/half way, causing these random crashes I'm
seeing now; some boards crash after less than a minute, some a few days,
some run fine for months.
Is this making sense? Anyone seen something like this?
I'll contact my PCB assembler in the morning, see what they can say
about this...
Regards,
-Bastiaan
Op 18-7-2012 23:01, stevec schreef:
> I tested LPC2106 used in 70 or so boards with lots of other
> components... OK up to about 60C. I would not try for 85C with
> commercial parts.
>
> --- In l... ,
> Bastiaan van Kesteren wrote:
> >
> > Hi,
> >
> > I'm having some trouble with an LPC2103 design related to higher
> > temperatures; the design becomes unstable and therefore unusuable at
> > higher temperatures (somewhere between 60 and 70 degrees celsius).
> >
> > According to the datasheet it should do +85, and all the components i'm
> > using around it should make this aswell.
> >
> > The strange thing is that some hand assembled boards do make this
> > temperature range, but the initial serie of professionally produced
> > boards don't. I've been swapping most components but the MCU, which now
> > seems to be the cultprint, which atm doesn't make sense to me.
> >
> >
> > Anyone having some experience related to higher temperatures and the
> > LPC2103 or related, and wanting to share some?
> >
> > Regards,
> > -Bastiaan
> >
Reply by ●July 24, 20122012-07-24
It has been a long time since I have used the LPC2103, but some of the other LPC
parts I have found have gone awry for no apparent reason and replacing them was
the only thing that fixed them.
Things like, IAP flash programming stopped working, JTAG would not work even after a complete flash erase via RS232, etc. And some very strange operational symptoms. But usually, if they make it to final test and out the door, they rarely have a problem after this. There have been a few though, but we've shipped many thousands of products with the LPC chips now.
You might also check the CM's anti-static practices. My guess would be that it is not the soldering temperature, but sure doesn't hurt to check. If the CM has been in business long time, it's probably something else.
I do feel your pain though, as about 8 years ago, I had about 1 year of nightmares while the company I was at was using the AVR Atmega 32 processor and Atmel had a silicon problem they did not own up to and it drove me crazy figuring out why processors were dying in very mysterious ways until I got them to own up to their problem. Almost killed our company. I don't believe that is your problem here though.
boB
--- In l..., Bastiaan van Kesteren wrote:
>
> Hi,
>
> Thanks for the replies!
>
> I've been doing some more testing here, and am starting to suspect the
> soldering-process.
>
> I now have two boards of the professionally produced series which have
> the crash-problem i'm having close after powerup (between 10 seconds and
> one minute), even at room temperature (which is a tad high here right
> now, but... :) )
>
> To clarify what I mean with 'crash-problem'; the product has serial
> output on an uart; when the board crashes, the baudrate of the
> uart-output goes way off (haven't measured the baudrate, but output is
> garbage). Other perhiperals stop working aswell, I assume speeds are
> changing there aswell.
>
> Boards running in hot locations seem to be more crash-prone, hence my
> initial temperature range suspicion.
>
> I attempted a reflow of the MCU on the 10-second-crash board, which made
> matters worse; it did not run properly at all, or only for a few seconds.
>
> So, i removed the MCU and hand-soldered a new one. And behold, the board
> runs... I only have it running for about 30 minutes now, but anyways the
> MCU seems to be the problem.
>
> So, I suspect the MCU has got too hot during the production-process,
> therby damaging it in a half/half way, causing these random crashes I'm
> seeing now; some boards crash after less than a minute, some a few days,
> some run fine for months.
>
> Is this making sense? Anyone seen something like this?
>
> I'll contact my PCB assembler in the morning, see what they can say
> about this...
>
> Regards,
> -Bastiaan
>
> Op 18-7-2012 23:01, stevec schreef:
> > I tested LPC2106 used in 70 or so boards with lots of other
> > components... OK up to about 60C. I would not try for 85C with
> > commercial parts.
> >
> > --- In l... ,
> > Bastiaan van Kesteren wrote:
> > >
> > > Hi,
> > >
> > > I'm having some trouble with an LPC2103 design related to higher
> > > temperatures; the design becomes unstable and therefore unusuable at
> > > higher temperatures (somewhere between 60 and 70 degrees celsius).
> > >
> > > According to the datasheet it should do +85, and all the components i'm
> > > using around it should make this aswell.
> > >
> > > The strange thing is that some hand assembled boards do make this
> > > temperature range, but the initial serie of professionally produced
> > > boards don't. I've been swapping most components but the MCU, which now
> > > seems to be the cultprint, which atm doesn't make sense to me.
> > >
> > >
> > > Anyone having some experience related to higher temperatures and the
> > > LPC2103 or related, and wanting to share some?
> > >
> > > Regards,
> > > -Bastiaan
> > >
> >
>
Things like, IAP flash programming stopped working, JTAG would not work even after a complete flash erase via RS232, etc. And some very strange operational symptoms. But usually, if they make it to final test and out the door, they rarely have a problem after this. There have been a few though, but we've shipped many thousands of products with the LPC chips now.
You might also check the CM's anti-static practices. My guess would be that it is not the soldering temperature, but sure doesn't hurt to check. If the CM has been in business long time, it's probably something else.
I do feel your pain though, as about 8 years ago, I had about 1 year of nightmares while the company I was at was using the AVR Atmega 32 processor and Atmel had a silicon problem they did not own up to and it drove me crazy figuring out why processors were dying in very mysterious ways until I got them to own up to their problem. Almost killed our company. I don't believe that is your problem here though.
boB
--- In l..., Bastiaan van Kesteren wrote:
>
> Hi,
>
> Thanks for the replies!
>
> I've been doing some more testing here, and am starting to suspect the
> soldering-process.
>
> I now have two boards of the professionally produced series which have
> the crash-problem i'm having close after powerup (between 10 seconds and
> one minute), even at room temperature (which is a tad high here right
> now, but... :) )
>
> To clarify what I mean with 'crash-problem'; the product has serial
> output on an uart; when the board crashes, the baudrate of the
> uart-output goes way off (haven't measured the baudrate, but output is
> garbage). Other perhiperals stop working aswell, I assume speeds are
> changing there aswell.
>
> Boards running in hot locations seem to be more crash-prone, hence my
> initial temperature range suspicion.
>
> I attempted a reflow of the MCU on the 10-second-crash board, which made
> matters worse; it did not run properly at all, or only for a few seconds.
>
> So, i removed the MCU and hand-soldered a new one. And behold, the board
> runs... I only have it running for about 30 minutes now, but anyways the
> MCU seems to be the problem.
>
> So, I suspect the MCU has got too hot during the production-process,
> therby damaging it in a half/half way, causing these random crashes I'm
> seeing now; some boards crash after less than a minute, some a few days,
> some run fine for months.
>
> Is this making sense? Anyone seen something like this?
>
> I'll contact my PCB assembler in the morning, see what they can say
> about this...
>
> Regards,
> -Bastiaan
>
> Op 18-7-2012 23:01, stevec schreef:
> > I tested LPC2106 used in 70 or so boards with lots of other
> > components... OK up to about 60C. I would not try for 85C with
> > commercial parts.
> >
> > --- In l... ,
> > Bastiaan van Kesteren wrote:
> > >
> > > Hi,
> > >
> > > I'm having some trouble with an LPC2103 design related to higher
> > > temperatures; the design becomes unstable and therefore unusuable at
> > > higher temperatures (somewhere between 60 and 70 degrees celsius).
> > >
> > > According to the datasheet it should do +85, and all the components i'm
> > > using around it should make this aswell.
> > >
> > > The strange thing is that some hand assembled boards do make this
> > > temperature range, but the initial serie of professionally produced
> > > boards don't. I've been swapping most components but the MCU, which now
> > > seems to be the cultprint, which atm doesn't make sense to me.
> > >
> > >
> > > Anyone having some experience related to higher temperatures and the
> > > LPC2103 or related, and wanting to share some?
> > >
> > > Regards,
> > > -Bastiaan
> > >
> >
>
