Cruise ship stuck in ice
in [Design]
|
Prev: Basic Stamp
Next: Energy harvesting
From: JosephKK on 8 Mar 2010 00:07 On Sat, 6 Mar 2010 23:00:26 +0200 (EET), Okkim Atnarivik <Okkim.Atnarivik(a)twentyfour.fi.invalid> wrote: >Paul Keinanen <keinanen(a)sci.fi> wrote: >: In order to keep this discussion on topic, why is it so hard to get >: component specifications from semiconductor manufacturers for >: temperatures below 0 C ? > > Actually, more and more modern CMOS semiconductors seem to work even >at 4.2K . More so than, say, 15 years ago. My wild guess is that >linewidth reduction combined with higher switching speed requirements >have driven the manufacturers to use heavier doping levels, attempting >to get higher mobility. A side effect is that the carriers no longer >freeze out. > > Regards, > Mikko Hmmm. Been studying up a bit recently. Logic fets may be doubly degenerately doped.
From: JosephKK on 8 Mar 2010 00:10 On Sat, 06 Mar 2010 16:28:07 -0500, Spehro Pefhany <speffSNIP(a)interlogDOTyou.knowwhat> wrote: >On Sat, 6 Mar 2010 23:00:26 +0200 (EET), the renowned Okkim Atnarivik ><Okkim.Atnarivik(a)twentyfour.fi.invalid> wrote: > >>Paul Keinanen <keinanen(a)sci.fi> wrote: >>: In order to keep this discussion on topic, why is it so hard to get >>: component specifications from semiconductor manufacturers for >>: temperatures below 0 C ? >> >> Actually, more and more modern CMOS semiconductors seem to work even >>at 4.2K . More so than, say, 15 years ago. My wild guess is that >>linewidth reduction combined with higher switching speed requirements >>have driven the manufacturers to use heavier doping levels, attempting >>to get higher mobility. A side effect is that the carriers no longer >>freeze out. >> >> Regards, >> Mikko > > >I thought it was primarily bipolar that had problems with 4K >(including such weirdities as Si diodes oscillating. Do you have/would >you be willing to share a list of parts that have been found to work >at liquid He temperatures? > > >Best regards, >Spehro Pefhany I have seen many reports of PCs being super-hotrodded in LN2.
From: Spehro Pefhany on 8 Mar 2010 02:07 On Sun, 07 Mar 2010 21:10:58 -0800, the renowned "JosephKK"<quiettechblue(a)yahoo.com> wrote: >On Sat, 06 Mar 2010 16:28:07 -0500, Spehro Pefhany <speffSNIP(a)interlogDOTyou.knowwhat> wrote: > >>On Sat, 6 Mar 2010 23:00:26 +0200 (EET), the renowned Okkim Atnarivik >><Okkim.Atnarivik(a)twentyfour.fi.invalid> wrote: >> >>>Paul Keinanen <keinanen(a)sci.fi> wrote: >>>: In order to keep this discussion on topic, why is it so hard to get >>>: component specifications from semiconductor manufacturers for >>>: temperatures below 0 C ? >>> >>> Actually, more and more modern CMOS semiconductors seem to work even >>>at 4.2K . More so than, say, 15 years ago. My wild guess is that >>>linewidth reduction combined with higher switching speed requirements >>>have driven the manufacturers to use heavier doping levels, attempting >>>to get higher mobility. A side effect is that the carriers no longer >>>freeze out. >>> >>> Regards, >>> Mikko >> >> >>I thought it was primarily bipolar that had problems with 4K >>(including such weirdities as Si diodes oscillating. Do you have/would >>you be willing to share a list of parts that have been found to work >>at liquid He temperatures? >> >> >>Best regards, >>Spehro Pefhany > >I have seen many reports of PCs being super-hotrodded in LN2. Sure, but LN2 is a balmy 77K .. unfortunately a lot more tends to go wrong in the next 73K down.. Best regards, Spehro Pefhany -- "it's the network..." "The Journey is the reward" speff(a)interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com
From: Paul Keinanen on 8 Mar 2010 16:11 On Sat, 06 Mar 2010 11:44:04 -0800, Joerg <invalid(a)invalid.invalid> wrote: > >> In practice, most problems at low temperature systems are associated >> with getting oscillators to oscillate at low temperatures. >> > >Look for categories such as "space and harsh environments", like here: >http://www.linear.com/pc/viewCategory.jsp?navId=H0,C1,C1778,C1503 > >They also have resistor-set oscillators if you can use those: >http://cds.linear.com/docs/Information%20Card/LTCMP.pdf > >What kind of problems do you have to get oscillators started? Russian >truckers sometimes light a little wood fire under the engine to warm it >a bit but I guess that's not an option in your case :-) An oscillator is simply an amplifier with a frequency selective feedback path that satisfies the Barkhausen criterion (gain larger than the feedback path loss). When an amplifier is powered up with a frequency selective feedback network, first the wide band thermal noise (about -174 dBm/Hz at room temperatures) is amplified and the noise after frequency shaping in the LC network is feed back to the input of the amplifier. After a few times though the amplifier, the broadband noise has been reduced to a narrow band single frequency signal, when the amplifier is saturated by noise. For a proper startup, the oscillator must start in class A or AB, while later on, it can drop to class C. At very low temperatures, if the Barkhausen requirement is not satisfied, if the gain drops too much at low temperatures, preventing the oscillation.
From: Joerg on 8 Mar 2010 17:31
Paul Keinanen wrote: > On Sat, 06 Mar 2010 11:44:04 -0800, Joerg <invalid(a)invalid.invalid> > wrote: > >>> In practice, most problems at low temperature systems are associated >>> with getting oscillators to oscillate at low temperatures. >>> >> Look for categories such as "space and harsh environments", like here: >> http://www.linear.com/pc/viewCategory.jsp?navId=H0,C1,C1778,C1503 >> >> They also have resistor-set oscillators if you can use those: >> http://cds.linear.com/docs/Information%20Card/LTCMP.pdf >> >> What kind of problems do you have to get oscillators started? Russian >> truckers sometimes light a little wood fire under the engine to warm it >> a bit but I guess that's not an option in your case :-) > > An oscillator is simply an amplifier with a frequency selective > feedback path that satisfies the Barkhausen criterion (gain larger > than the feedback path loss). > > When an amplifier is powered up with a frequency selective feedback > network, first the wide band thermal noise (about -174 dBm/Hz at room > temperatures) is amplified and the noise after frequency shaping in > the LC network is feed back to the input of the amplifier. After a few > times though the amplifier, the broadband noise has been reduced to a > narrow band single frequency signal, when the amplifier is saturated > by noise. > > For a proper startup, the oscillator must start in class A or AB, > while later on, it can drop to class C. > > At very low temperatures, if the Barkhausen requirement is not > satisfied, if the gain drops too much at low temperatures, preventing > the oscillation. > This is not the only way to start an oscillator. Another method when there isn't enough initial gain and/or noise is a kicker circuit. Can be a bit tricky because often there must be additional circuitry that detects a successful start and if necessary can initiate restart attempts. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM. |