Automotive temp MCU

Ralph Malph wrote:
<snip>
> I am a bit surprised that no one has been able to point me toward
> another maker or two of automotive MCUs.  I guess the pickin's are
> pretty slim out there.

  This will likely be because the 125'C is mainly under-bonnet - and 
that's high volumes / large NRE / narrow customer focus.
  It's also not where a $3 uC will be found....
  For Dash/Boot/Cabin applications, 85'C is fine, and that's where
a much broader cutomer range applies - mostly it's the lower
temp range that is important.

  Do you have numbers on the amb your uC will ACTUALLY be asked to
operate in ?

-jg

Microchip's sampling is very generous! I've gotten at least 5
different mcu samples. They ship it to you in less than a week, direct
from Thailand or something like that.




rwyoung@ieee.org (Rob Young) wrote in message news:<69e5135f.0401261444.39151264@posting.google.com>...
> Ralph Malph <noone@yahoo.com> wrote in message news:<4014C024.EF30F1D@yahoo.com>...
> > Ralph Malph wrote:
> > > 
> > > I am looking at replacing several hardware chips with an MCU.  I
> > > considered this once before but was looking for a lot more in the way of
> > > IO than I am now.  Also, this time the MCU has to operate over the full
> > > automotive temp range of -40 to +125C.  I have looked at several
> > > manufacturers web sites, but they typically don't include temperature
> > > range in the guides.  So I have to pick the dozen or so that are a
> > > likely fit otherwise and then download every data sheet.
> > > 
> > > I thought I might ask here for chips that anyone has used before that
> > > would fit my needs.
> > > 
> > > -40C to 125C range (of course)
> > > ISP
> > > 18 or more IOs (after accounting for ISP)
> > > at least two outputs must drive LEDs at 20 mA
> > > very little RAM and Flash size requirements, they just have to be there
> > > the flash has to be writable by the MCU (or data EEPROM will do)
> > > temperature measurement or ADC inputs
> > > small package; TSSOP or QFN or very small QFP (like 48 TQFP)
> > > low price, <$3 in 1k qty
> > > easy to buy in 100's or 1000's
> > 
> > 
> > I think I found a good one at Microchip.  The PIC16F872 comes in
> > extended temperature rather than "automotive" which explains why I
> > couldn't find parts in a search.  It also turns out that Microchip will
> > make extended temp versions of chips and not put it in the data sheet or
> > in any selection guides.  They really do make it hard to use their
> > chips.  The price on this one is a bit higher that my target, but not by
> > much.  It meets all the above criteria except for the package, but they
> > have a SSOP28 which is not overly large at about 50% larger than a 28
> > pin QFN.  I can live with that.  I can buy them at Digikey although not
> > in onesies-twosies.  There minimum on most extended temp parts is what
> > the maker requires, 260 or so in this case.  
> > 
> > I would still like to have a second choice and I wouldn't mind cutting
> > the price a bit more (sometimes I can be a real penny pincher).  I
> > looked at Motorola (nothing below $4 that meets the specs), Atmel
> > (nothing automotive that I could find), National (nothing to the spec),
> > Fairchild (only very tiny parts), Analog Devices (MicroConverts are a
> > bit steep), Dallas (too pricey) and I even looked at Intersil (the old
> > RCA 1802 which has no memory on chip it seems).  
> > 
> > Did I miss any good candiates?  I am shy of the Asian companies because
> > they can be very hard to buy from, much less get support.
> 
> 
> You can probably get better pricing on the part if you check with
> Future Electronics  http://www.futureelectronics.com/
> 
> PIC16F872-I/SS is listed at $2.90 each if you buy them 47 at a time (1
> tube).  Somewhat cheaper than Digikey.
> 
> And if you just want a few to "play" with, Microchip has a pretty good
> sampling policy.  I think you can get up to 3 or 5 parts per 6 month
> period free.
> 
> Rob Young
> rwyoung@ieee.org

>Did I miss any good candiates?  I am shy of the Asian companies because
>they can be very hard to buy from, much less get support.


Have a look at the ST6 series from ST http://www.st.com/ 

They may do what you want but the temp range is only -40
to +85.  Something like the ST6265 has 21 i/o, 3884 byte
user ROM, 128 bytes RAM, 128 bytes EEPROM. 8 i/o can drive
20mA, 8 can be ADC input, any digital i/o can be used for 
interrupt.  Has other "nice" features as well - but it is
slow nowadays - 8MHz crystal divided by 12/13 for internal 
clocks.

Alan

++++++++++++++++++++++++++++++++++++++++++
Jenal Communications
Manufacturers and Suppliers of HF Selcall
P O Box 1108, Morley, WA, 6943
Tel: +61 8 9370 5533 Fax +61 8 9370 5106
Web Site: http://www.jenal.com 
e-mail: http://www.jenal.com/contact.php
++++++++++++++++++++++++++++++++++++++++++

CBarn24050 wrote:
> 
> hi, 125deg used to be the military grade, have things changed while I wasn't
> looking? Can't you get away with 85deg?

Mil temp is *-55C* to +125C while automotive (or Microchip calls it
"extended") is *-40C* to +125C.  

I am building a board that will run at up to 85C, but it will shut
itself off at 85C and survive up to 125C.  The part that controls the
on/off (which also includes a low power mode with multiple wake up
options) is simplest if I use a small MCU.  If I use other,
non-programmable parts, I end up with several devices to replace the
MCU.  By adding the MCU I also get several other features that otherwise
need even more chips.

Jim Granville wrote:
> 
> Ralph Malph wrote:
> <snip>
> > I am a bit surprised that no one has been able to point me toward
> > another maker or two of automotive MCUs.  I guess the pickin's are
> > pretty slim out there.
> 
>   This will likely be because the 125'C is mainly under-bonnet - and
> that's high volumes / large NRE / narrow customer focus.
>   It's also not where a $3 uC will be found....
>   For Dash/Boot/Cabin applications, 85'C is fine, and that's where
> a much broader cutomer range applies - mostly it's the lower
> temp range that is important.
> 
>   Do you have numbers on the amb your uC will ACTUALLY be asked to
> operate in ?

125C or higher if I could do it...

This will be used (among other places) on a locomotive as part of some
test gear.  When in a tunnel the temps can get very high from what I am
told.

Alan wrote:
> 
> >Did I miss any good candiates?  I am shy of the Asian companies because
> >they can be very hard to buy from, much less get support.
> 
> Have a look at the ST6 series from ST http://www.st.com/
> 
> They may do what you want but the temp range is only -40
> to +85.  Something like the ST6265 has 21 i/o, 3884 byte
> user ROM, 128 bytes RAM, 128 bytes EEPROM. 8 i/o can drive
> 20mA, 8 can be ADC input, any digital i/o can be used for
> interrupt.  Has other "nice" features as well - but it is
> slow nowadays - 8MHz crystal divided by 12/13 for internal
> clocks.

Thanks for the reply, but the temperature is what this question is all
about.  I can find lots of chips that meet all the other requirements
including the tiny QFN package and the low price.  

I can also find parts that are automotive temp, but they fail in one of
the other requirements.  I found one that is only slightly out of my
price range and is about twice as large as a 28QFN.  So I can use it if
I have to.  I am just trying to cover all the ground to make sure I have
not missed anything.  

Too bad Atmel does not have an automotive temp AVRs.  I would love to
use one of them.

In message <4016B281.7BB2760A@yahoo.com>, Ralph Malph <noone@yahoo.com> 
writes

[snip]

>I am building a board that will run at up to 85C, but it will shut
>itself off at 85C and survive up to 125C.  The part that controls the
>on/off (which also includes a low power mode with multiple wake up
>options) is simplest if I use a small MCU.

An 85C PIC will easily run at well over 125C - just don't use the 
internal RC oscillator or watchdog.  Are 125C rated parts any different? 
Don't think so.  Try it.

-- 
Syd

On Tue, 27 Jan 2004 13:50:24 -0500, Ralph Malph <noone@yahoo.com>
wrote:

>
>125C or higher if I could do it...
>
>This will be used (among other places) on a locomotive as part of some
>test gear.  When in a tunnel the temps can get very high from what I am
>told.

How long does it spend in the tunnel ? If it is only momentary, some
insulation and/or some massive metal parts may limit the rate of
temperature change.

Also look at the MCU maximum junction temperature and thermal
resistance specifications. You might be able to run the MCU at a clock
frequency well below the nominal clock frequency and since the power
dissipation is usually directly proportional to the clock frequency,
you might be able to run it at a sufficiently low junction
temperature.

Also adding some large heat sink may help. Assuming maximum Tj of 150
C and a thermal resistance from junction to ambient of 25 C/W, you
could dissipate 1 W, however, if the total thermal resistance can be
reduced to 10 C/W, 2.5 W could be dissipated or alternatively the
junction temperature could be 135 C at 1 W.  A heatsink could be
useful even with many low power devices that you would never consider
using at normal temperatures.

However, a large heatsink may be a problem, if there are large
accelerations or vibrations.

If everything else fails, you might consider using a Peltier cooler,
but remember, that you must have a place for dumping not only the heat
generated by the electronics, but also the heat generated by the
Peltier element (U x I). The efficiency of the Peltier is not very
spectacular, so be prepared to dissipate quite a lot of power.

Paul
       

Syd Rumpo wrote:
> 
> In message <4016B281.7BB2760A@yahoo.com>, Ralph Malph <noone@yahoo.com>
> writes
> 
> [snip]
> 
> >I am building a board that will run at up to 85C, but it will shut
> >itself off at 85C and survive up to 125C.  The part that controls the
> >on/off (which also includes a low power mode with multiple wake up
> >options) is simplest if I use a small MCU.
> 
> An 85C PIC will easily run at well over 125C - just don't use the
> internal RC oscillator or watchdog.  Are 125C rated parts any different?
> Don't think so.  Try it.

Yes, they are different, they are guaranteed to meet spec at 125C if
nothing else.  I can't come close to testing the parts the way the maker
does.  Some things like leakage current are important and not worth
testing on chips before I make the boards.  I have considered building
the boards with industrial chips and testing each one, but this can be
very expensive.  

Thanks for the suggestion anyway.

Paul Keinanen wrote:
> 
> On Tue, 27 Jan 2004 13:50:24 -0500, Ralph Malph <noone@yahoo.com>
> wrote:
> 
> >
> >125C or higher if I could do it...
> >
> >This will be used (among other places) on a locomotive as part of some
> >test gear.  When in a tunnel the temps can get very high from what I am
> >told.
> 
> How long does it spend in the tunnel ? If it is only momentary, some
> insulation and/or some massive metal parts may limit the rate of
> temperature change.

They can be parked/stuck in a tunnel for hours.  


> Also look at the MCU maximum junction temperature and thermal
> resistance specifications. You might be able to run the MCU at a clock
> frequency well below the nominal clock frequency and since the power
> dissipation is usually directly proportional to the clock frequency,
> you might be able to run it at a sufficiently low junction
> temperature.

Yes, I am doing that with the LDO.  But I prefer to use fully spec'd
automotive parts.  I have found a number of PICs that can do
automotive.  Now I am just trying to optimize my choice.  


> Also adding some large heat sink may help. Assuming maximum Tj of 150
> C and a thermal resistance from junction to ambient of 25 C/W, you
> could dissipate 1 W, however, if the total thermal resistance can be
> reduced to 10 C/W, 2.5 W could be dissipated or alternatively the
> junction temperature could be 135 C at 1 W.  A heatsink could be
> useful even with many low power devices that you would never consider
> using at normal temperatures.

This MCU will be dissipating nearly no power (uWs) when over 85C.  So I
don't think this is needed.  

 
> However, a large heatsink may be a problem, if there are large
> accelerations or vibrations.

This part is on the bottom of the board with 0.1" clearance, not much
room for a heatsink  :) 


> If everything else fails, you might consider using a Peltier cooler,
> but remember, that you must have a place for dumping not only the heat
> generated by the electronics, but also the heat generated by the
> Peltier element (U x I). The efficiency of the Peltier is not very
> spectacular, so be prepared to dissipate quite a lot of power.

Far too exotic, but thanks for the suggestions.