Forums

Ultra low temperature microcontroller ?

Started by Unknown July 12, 2007
Morten M J�rgensen wrote:

> Since the application is a small footprint temperature sensor I don't > think this will do. I think it will be hard to control the influence of > the heater !?!?
Do you really need the microcontroller to be part of the temperature sensor assembly? Dealing as I do in cryogenic temperature measurements we go to a great deal of effort to ensure we thermally ground the wiring to our sensors (4 wire plan always). This way we can be sure that the temperature sensed is that of the point we are monitoring and not the heating effect of the current flowing in the wiring (even a few uA is enough heat to give several degrees Kelvin error). We also take care with the source energisation to minimise the battery like effects at wiring junctions (plugs and sockets etc). -- ******************************************************************** Paul E. Bennett ....................<email://peb@amleth.demon.co.uk> Forth based HIDECS Consultancy .....<http://www.amleth.demon.co.uk/> Mob: +44 (0)7811-639972 Tel: +44 (0)1235-811095 Going Forth Safely ..... EBA. www.electric-boat-association.org.uk.. ********************************************************************
Paul Burke <paul@scazon.com> writes:
> Morten M J&#2013266168;rgensen wrote: > > We are having a hard time finding a suitable microcontroller for our new > > automotive project. The product will have to operate in a very harsh > > enviroment including temperatures lower than -40 deg. C. > > Put the microcontoller between the driver's thighs. If it drops below > -40 there, nobody's going anywhere.
A little more seriously, can anyone give a simple explanation as to why low temperatures are a problem for solid-state electronics? Do the eletrons freeze?
"Everett M. Greene" <mojaveg@mojaveg.lsan.mdsg-pacwest.com> wrote in message 
news:20070712.79D3A60.ABA3@mojaveg.lsan.mdsg-pacwest.com...
> Paul Burke <paul@scazon.com> writes: >> Morten M J&#2013266168;rgensen wrote: >> > We are having a hard time finding a suitable microcontroller for our >> > new >> > automotive project. The product will have to operate in a very harsh >> > enviroment including temperatures lower than -40 deg. C. >> >> Put the microcontoller between the driver's thighs. If it drops below >> -40 there, nobody's going anywhere. > > A little more seriously, can anyone give a simple explanation as to > why low temperatures are a problem for solid-state electronics? Do > the eletrons freeze?
When I talked to a Microchip Rep about using their chips for operating down to -76 deg C they said that one of the problems is differential expantion/contraction between the die and the package. At that temperature the die often will crack. Don

Everett M. Greene wrote:

> A little more seriously, can anyone give a simple explanation as to > why low temperatures are a problem for solid-state electronics? Do > the eletrons freeze?
Exactly. I am trying to explain it without going into details; it is actually lot more complicated. The behavior of a semiconductor material depends on how many charge carriers can jump across the energy gap from valent to conduction zone. The average energy of the charge carriers depends on the temperature. So, at low temperatures the semiconductor material behaves like a dielectric, and at high temperatures it behaves like a metal. A semiconductor acts as a semiconductor only within a range of temperatures. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
On Thu, 12 Jul 2007 20:15:21 GMT, "Donald Harris"
<harrisdw1@verizon.net> wrote:

> >"Everett M. Greene" <mojaveg@mojaveg.lsan.mdsg-pacwest.com> wrote in message >news:20070712.79D3A60.ABA3@mojaveg.lsan.mdsg-pacwest.com... >> Paul Burke <paul@scazon.com> writes: >>> Morten M J&#2013266168;rgensen wrote: >>> > We are having a hard time finding a suitable microcontroller for our >>> > new >>> > automotive project. The product will have to operate in a very harsh >>> > enviroment including temperatures lower than -40 deg. C. >>> >>> Put the microcontoller between the driver's thighs. If it drops below >>> -40 there, nobody's going anywhere. >> >> A little more seriously, can anyone give a simple explanation as to >> why low temperatures are a problem for solid-state electronics? Do >> the eletrons freeze? > >When I talked to a Microchip Rep about using their chips for operating down >to -76 deg C they said that one of the problems is differential >expantion/contraction between the die and the package. At that temperature >the die often will crack.
That is more a storage temperature issue. For bipolar silicon devices. the junction forward voltage drop has a temperature coefficient about -2 mV/C so with a junction temperature from -55 C to +200 C, for instance the transistor Vbe can vary about 500 mV, so extra care is required in designing the biasing arrangement. This can be problematic in low voltage battery operated systems, especially as the battery voltage also drops with temperature. Also the bipolar current gain drops significantly with temperature and in oscillators the active device gain must be able to compensate for the resonator losses i.e. the total loop gain must be at least 1. At low temperatures the total loop gain may be below 1, in which case oscillation is not possible. It can take a few seconds or a few minutes, before the bias current warms the transistor sufficiently, to increase the current gain and thus making oscillation possible. There are a large number of ways things can go wrong and if a module manufacturer has a heat/cold test unit that goes only down to -40 C, the published temperature range is limited to -40 C, even if the unit would in practice function reliably at lower temperatures. Paul
On Jul 12, 3:33 am, "Morten M J=F8rgensen" <n...@fake.mail.com> wrote:
> Hello all, > > We are having a hard time finding a suitable microcontroller for our new > automotive project. The product will have to operate in a very harsh > enviroment including temperatures lower than -40 deg. C. > > Most microcontroller has -40 deg. C. as the lowest limit. Anyone here kno=
ws
> of microcontrollers that maybe goes to -50 deg. C. ? > > All architectures are welcome! > > Best Regards > > Morten M. J=F8rgensen
I've worked on military products where we have to be fully operation at -55 Deg C, previously you could buy these off the shelf, but most manufacterers don't offer them anymore. What we do now is buy -40 Deg parts and screen them (test them at -55 Deg C), there are companies that can do this for you.
steve wrote:
> I've worked on military products where we have to be fully operation > at -55 Deg C, previously you could buy these off the shelf, but most > manufacterers don't offer them anymore. What we do now is buy -40 Deg > parts and screen them (test them at -55 Deg C), there are companies > that can do this for you.
What sort of yield do you get doing this, and how does it vary across parts ? -jg
Morten M J&#2013266168;rgensen wrote:
>Hello all, > >We are having a hard time finding a suitable microcontroller for our new >automotive project. The product will have to operate in a very harsh >enviroment including temperatures lower than -40 deg. C. > >Most microcontroller has -40 deg. C. as the lowest limit. Anyone here knows >of microcontrollers that maybe goes to -50 deg. C. ? > >All architectures are welcome! > >Best Regards > >Morten M. J&#2013266168;rgensen >
The Freescale MPC555, MPC566 and MPC5554 are all available in -55 to 125 degrees. I believe they are the standard devices tested to extended temperature. They're not cheap though. They are about $120 each in quantity, compared with approx $50 for the -40 to 125 devices. Andy Sinclair