Forums

Why Hasn't This Been Done with Silicon Carbide

Started by John Savard December 12, 2014
On Fri, 12 Dec 2014 16:06:15 -0700, Jim Thompson
<To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:

>On Fri, 12 Dec 2014 12:27:13 -0800, John Larkin ><jlarkin@highlandtechnology.com> wrote: > >>On Fri, 12 Dec 2014 13:56:40 -0600, "Tim Williams" >><tiwill@seventransistorlabs.com> wrote: >> >>>"John Larkin" <jlarkin@highlandtechnology.com> wrote in message >>>news:i9am8ah27bp9bf5ifch26p4gso3ip8a744@4ax.com... >>>> Compound semiconductors, in addition to defects, tend to make only >>>> N-type devices. So, no equivalent of CMOS. >>> >>>Specifically, the hole mobility is balls. >>> >>>So for example, the Cray-1 or whatever it was that was constructed >>>entirely from GaAs and ran at a blazing 60MHz -- was made in old fashioned >>>NMOS. No CMOS, no low voltage, no ECL. Almost as hot as ECL though. >>> >>>At least, that's what I remember reading. >> >> >>The Gigabit Logic parts were all n-fets, with depletion fets for >>loads, pretty much RTL type logic; absurd power hogs. >> >>> >>>> Silicon is unique in that SiO2 makes a great implantation mask, so the >>>> photolithography process works really well. >>> >>>Well, SiC too I suppose. Unless they need enough implantation energy that >>>it goes on through regardless. >>> >>>Not sure about Ge, but no one uses that stuff pure because temp >>>performance is nonexistent. >> >>GeO isn't good like SiO. >> >>There are very few germanium parts made any more. One germanium device >>made with actual lithography is a back diode, a fairly obscure >>microwave detector diode. >> >>There are germanium photodiodes. >> >>> >>>> SiC fets on the market now are slow, due to high gate contact >>>> resistances. They need a lot of gate voltage swing, too. GaN seems to >>>> be better. >>> >>>Too bad no one's making GaN these days. Not much anyway, not for power. >>>There's the lower voltage EPC dies, and this thing is brand new, >>>http://www.mouser.com/ProductDetail/GaN-Systems/GS66508P-E03-TY/?qs=sGAEpiMZZMshyDBzk1%2fWi6IaY8dKw19I0Ueh0isV9Zg5enZmT3gDjg%3d%3d >>>and with suspiciously familiar ratings if I do say so myself... >> >> >>GaN is big in RF, especially military stuff. Lots of people are >>investing big time, and shipping parts. Macom, Cree, Triquint, >>Nitronics, Amcom, IR, probably more. >> >>I haven't seen a SiC or GaN fet with avalanche ratings. > >SiGe makes for screamingly fast bipolar devices... > ><http://en.wikipedia.org/wiki/Silicon-germanium> > > ...Jim Thompson
The EclipsPlus ECL gates are SiGe. And some of the Analog Devices fast comparators, I think. -- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
On 12 Dec 2014 21:53:22 GMT, Glen Walpert <nospam@null.void> wrote:

>On Fri, 12 Dec 2014 12:27:13 -0800, John Larkin wrote: > ><clip> > >> I haven't seen a SiC or GaN fet with avalanche ratings. > >http://www.cree.com/Power/Products/MOSFETs/TO247/C2M0040120D > >data sheet claims it is avalanche rugged and includes an avalanche SOA >graph.
That's nice. They have got the series gate resistance down some, too. -- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
On Fri, 12 Dec 2014 15:25:10 -0800, John Larkin
<jlarkin@highlandtechnology.com> wrote:

>On Fri, 12 Dec 2014 16:06:15 -0700, Jim Thompson ><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: > >>On Fri, 12 Dec 2014 12:27:13 -0800, John Larkin >><jlarkin@highlandtechnology.com> wrote: >> >>>On Fri, 12 Dec 2014 13:56:40 -0600, "Tim Williams" >>><tiwill@seventransistorlabs.com> wrote: >>> >>>>"John Larkin" <jlarkin@highlandtechnology.com> wrote in message >>>>news:i9am8ah27bp9bf5ifch26p4gso3ip8a744@4ax.com... >>>>> Compound semiconductors, in addition to defects, tend to make only >>>>> N-type devices. So, no equivalent of CMOS. >>>> >>>>Specifically, the hole mobility is balls. >>>> >>>>So for example, the Cray-1 or whatever it was that was constructed >>>>entirely from GaAs and ran at a blazing 60MHz -- was made in old fashioned >>>>NMOS. No CMOS, no low voltage, no ECL. Almost as hot as ECL though. >>>> >>>>At least, that's what I remember reading. >>> >>> >>>The Gigabit Logic parts were all n-fets, with depletion fets for >>>loads, pretty much RTL type logic; absurd power hogs. >>> >>>> >>>>> Silicon is unique in that SiO2 makes a great implantation mask, so the >>>>> photolithography process works really well. >>>> >>>>Well, SiC too I suppose. Unless they need enough implantation energy that >>>>it goes on through regardless. >>>> >>>>Not sure about Ge, but no one uses that stuff pure because temp >>>>performance is nonexistent. >>> >>>GeO isn't good like SiO. >>> >>>There are very few germanium parts made any more. One germanium device >>>made with actual lithography is a back diode, a fairly obscure >>>microwave detector diode. >>> >>>There are germanium photodiodes. >>> >>>> >>>>> SiC fets on the market now are slow, due to high gate contact >>>>> resistances. They need a lot of gate voltage swing, too. GaN seems to >>>>> be better. >>>> >>>>Too bad no one's making GaN these days. Not much anyway, not for power. >>>>There's the lower voltage EPC dies, and this thing is brand new, >>>>http://www.mouser.com/ProductDetail/GaN-Systems/GS66508P-E03-TY/?qs=sGAEpiMZZMshyDBzk1%2fWi6IaY8dKw19I0Ueh0isV9Zg5enZmT3gDjg%3d%3d >>>>and with suspiciously familiar ratings if I do say so myself... >>> >>> >>>GaN is big in RF, especially military stuff. Lots of people are >>>investing big time, and shipping parts. Macom, Cree, Triquint, >>>Nitronics, Amcom, IR, probably more. >>> >>>I haven't seen a SiC or GaN fet with avalanche ratings. >> >>SiGe makes for screamingly fast bipolar devices... >> >><http://en.wikipedia.org/wiki/Silicon-germanium> >> >> ...Jim Thompson > >The EclipsPlus ECL gates are SiGe. And some of the Analog Devices fast >comparators, I think.
The first Garmin GPS chip (my design ;-) was on IBM (Burlington, VT) SiGe... all bipolar. ...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
On Saturday, 13 December 2014 04:53:49 UTC+11, John Larkin  wrote:
> On Fri, 12 Dec 2014 02:35:43 -0800 (PST), John Savard > <jsavard@ecn.ab.ca> wrote: >=20 > >Today's microprocessors are all made using the CMOS logic family. > > > >It has the advantage of using a minimum amount of electricity, since - e=
xcept for leakage currents, which are becoming more important as transistor= s and wires shrink - electrical power is only used during changes of state.
> > > >However, the performance of CMOS circuits is limited by the slower P-typ=
e FET branch of them - this can be helped by going to Germanium, which has = high hole mobility, or by using stretched silicon, or by using domino logic= ... the IBM CELL processor used an alternative approach instead of domino l= ogic which also worked.
> > > >And the fastest logic family used to be ECL, because the transistors did=
n't saturate. But it was an energy hog.
> > > >Silicon carbide can tolerate high temperatures, and it's a semiconductor=
. The trouble is that silicon carbide crystals are riddled with defects, so= it's a challenge to even make decent single transistors out of it, let alo= ne microprocessors with millions of transistors!
> > > >However, the thought occurred to me that surely there must be some mater=
ial that forms nice single crystals, free of defects, with the same interat= omic lattice spacing as silicon carbide. One could use wafers of _that_ - p= resuming it's also an insulator - and using chemical vapor deposition, prod= uce good silicon carbide transistors in large numbers per die...
> > > >No doubt there are good reasons why that is harder than it seems. > > > >John Savard >=20 > Compound semiconductors, in addition to defects, tend to make only > N-type devices. So, no equivalent of CMOS. >=20 > Silicon is unique in that SiO2 makes a great implantation mask, so the > photolithography process works really well. >=20 > I don't think anyone has sold even an opamp from SiC. There are > microwave ICs made from InP and GaAs and such, but tend to be > relatively simple, with distributed amps being the high end of > commercial products. >=20 > http://www.semiconductor-today.com/news_items/2014/AUG/KTH_010814.shtml >=20 > http://ieeexplore.ieee.org/xpl/login.jsp?tp=3D&arnumber=3D6019027&url=3Dh=
ttp%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D6019027
>=20 > That is NOT very impressive so far. >=20 >=20 > Gigabit Logic once made GaAs logic chips, long gone now.
I used them - quite successfully. ECLinPS is now almost as fast http://www.onsemi.com/pub/Collateral/MC100EP016A-D.PDF looks as if it can be cascaded to count at 600MHz, rather than the 800MHz w= e got out of the GaAs equivalent. The catch seems to have been that the production yield never got high enoug= h for GigaBit to make money out of them, even through they were ten times m= ore expensive than regular 100k ECL at the time. --=20 Bill Sloman, Sydney
On Friday, December 12, 2014 5:33:51 PM UTC, Tim Wescott wrote:
> On Fri, 12 Dec 2014 08:25:56 -0800, meow2222 wrote: > > On Friday, December 12, 2014 10:35:48 AM UTC, John Savard wrote:
> >> Silicon carbide can tolerate high temperatures, and it's a > >> semiconductor. The trouble is that silicon carbide crystals are riddled > >> with defects, so it's a challenge to even make decent single > >> transistors out of it, let alone microprocessors with millions of > >> transistors!
> > What exactly is the result of a transistor built on a defective piece of > > crystal?
> Generally, a useless piece of crystal with an interesting doping profile. > > I'm not sure down to the atomic level what's going on, but any crystal > defect that conducts electricity and that slashes across what should be an > insulating layer will kill the thing, as would any crystal defect that > carries the wrong dopant into the wrong part of the device during > manufacture.
I guess I was idly wondering if one could put devices in series or parallel to get round the problem upto a point. Then in theory one could create parallel data paths with voting on inputs... gets very wasteful though.
> I try not to dwell too much on "if only", if only because you end up > wasting a lot of time on useless wishing.
Most of it is useless, but 0.01% or so is surprisingly valuable NT
Am 13.12.2014 um 01:36 schrieb Bill Sloman:

>> Gigabit Logic once made GaAs logic chips, long gone now. > > I used them - quite successfully. ECLinPS is now almost as fast > > http://www.onsemi.com/pub/Collateral/MC100EP016A-D.PDF > > looks as if it can be cascaded to count at 600MHz, rather than the 800MHz we got out of the GaAs equivalent. >
Try these: < http://www.onsemi.com/pub_link/Collateral/MC100EP016A-D.PDF > regards, Gerhard
On Sat, 13 Dec 2014 02:01:41 +0100, Gerhard Hoffmann
<ghf@hoffmann-hochfrequenz.de> wrote:

>Am 13.12.2014 um 01:36 schrieb Bill Sloman: > >>> Gigabit Logic once made GaAs logic chips, long gone now. >> >> I used them - quite successfully. ECLinPS is now almost as fast >> >> http://www.onsemi.com/pub/Collateral/MC100EP016A-D.PDF >> >> looks as if it can be cascaded to count at 600MHz, rather than the 800MHz we got out of the GaAs equivalent. >> > >Try these: >< http://www.onsemi.com/pub_link/Collateral/MC100EP016A-D.PDF > > >regards, Gerhard
This is nice, http://www.onsemi.com/pub/Collateral/NB7V52M-D.PDF under $10 for a 10 GHz flop. Internal terminations. -- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
On Saturday, 13 December 2014 12:01:46 UTC+11, Gerhard Hoffmann  wrote:
> Am 13.12.2014 um 01:36 schrieb Bill Sloman: >=20 > >> Gigabit Logic once made GaAs logic chips, long gone now. > > > > I used them - quite successfully. ECLinPS is now almost as fast > > > > http://www.onsemi.com/pub/Collateral/MC100EP016A-D.PDF > > > > looks as if it can be cascaded to count at 600MHz, rather than the 800M=
Hz we got out of the GaAs equivalent.
> > >=20 > Try these: > < http://www.onsemi.com/pub_link/Collateral/MC100EP016A-D.PDF >
Isn't that exactly the URL that I posted? My problem - back in 1988 - was that I needed a 32-bit wide counter clocked= at 800MHz. The plan was to do most of the counting in 100k ECL which was g= oing to be tricky, and - in the event - never worked, because the guy who d= id the detailed design of the relevant board didn't bother reading the bit = of my specification that discussed how to get the 100k ECL to work as fast = as was needed, and an unrealistic schedule meant that the design reviews th= at would have caught this had had to be by-passed to meet the schedule. Comically, a completely unrelated screw up meant that that board never work= ed - some idiot in the drafting office okayed a redistribution of the inner= layers of the board which meant that none of the critical printed circuit = tracers had the characteristic impedance (or the immunity from cross-talk) = that they needed. The board was persuaded to sort of work by rerouting all the critical paths= into subminature coax (1.1mm OD Filotex - it's still available from Farnel= l, but the name has changed). By the third go-around, the board worked adeq= uately, but half as fast as had been originally intended - nobody who matte= red was prepared to risk changing the logic to make it go that fast, reflec= ting the fact that the original demand for 800MHz had been management-impos= ed to allow the boss to claim a 10psec granularity in time resolution which= we'd never actually needed, nor would have been able to prove that we had. --=20 Bill Sloman, Sydney
On Saturday, December 13, 2014 2:32:59 AM UTC, Bill Sloman wrote:
> On Saturday, 13 December 2014 12:01:46 UTC+11, Gerhard Hoffmann wrote: > > Am 13.12.2014 um 01:36 schrieb Bill Sloman: > >=20 > > >> Gigabit Logic once made GaAs logic chips, long gone now. > > > > > > I used them - quite successfully. ECLinPS is now almost as fast > > > > > > http://www.onsemi.com/pub/Collateral/MC100EP016A-D.PDF > > > > > > looks as if it can be cascaded to count at 600MHz, rather than the 80=
0MHz we got out of the GaAs equivalent.
> > > > >=20 > > Try these: > > < http://www.onsemi.com/pub_link/Collateral/MC100EP016A-D.PDF > >=20 > Isn't that exactly the URL that I posted? >=20 > My problem - back in 1988 - was that I needed a 32-bit wide counter clock=
ed at 800MHz. The plan was to do most of the counting in 100k ECL which was= going to be tricky, and - in the event - never worked, because the guy who= did the detailed design of the relevant board didn't bother reading the bi= t of my specification that discussed how to get the 100k ECL to work as fas= t as was needed, and an unrealistic schedule meant that the design reviews = that would have caught this had had to be by-passed to meet the schedule.
>=20 > Comically, a completely unrelated screw up meant that that board never wo=
rked - some idiot in the drafting office okayed a redistribution of the inn= er layers of the board which meant that none of the critical printed circui= t tracers had the characteristic impedance (or the immunity from cross-talk= ) that they needed.
>=20 > The board was persuaded to sort of work by rerouting all the critical pat=
hs into subminature coax (1.1mm OD Filotex - it's still available from Farn= ell, but the name has changed). By the third go-around, the board worked ad= equately, but half as fast as had been originally intended - nobody who mat= tered was prepared to risk changing the logic to make it go that fast, refl= ecting the fact that the original demand for 800MHz had been management-imp= osed to allow the boss to claim a 10psec granularity in time resolution whi= ch we'd never actually needed, nor would have been able to prove that we ha= d. I so dont miss all that NT
John Larkin wrote:
> On Fri, 12 Dec 2014 02:35:43 -0800 (PST), John Savard > <jsavard@ecn.ab.ca> wrote: > >> Today's microprocessors are all made using the CMOS logic family. >> >> It has the advantage of using a minimum amount of electricity, since - except for leakage currents, which are becoming more important as transistors and wires shrink - electrical power is only used during changes of state. >> >> However, the performance of CMOS circuits is limited by the slower P-type FET branch of them - this can be helped by going to Germanium, which has high hole mobility, or by using stretched silicon, or by using domino logic... the IBM CELL processor used an alternative approach instead of domino logic which also worked. >> >> And the fastest logic family used to be ECL, because the transistors didn't saturate. But it was an energy hog. >> >> Silicon carbide can tolerate high temperatures, and it's a semiconductor. The trouble is that silicon carbide crystals are riddled with defects, so it's a challenge to even make decent single transistors out of it, let alone microprocessors with millions of transistors! >> >> However, the thought occurred to me that surely there must be some material that forms nice single crystals, free of defects, with the same interatomic lattice spacing as silicon carbide. One could use wafers of _that_ - presuming it's also an insulator - and using chemical vapor deposition, produce good silicon carbide transistors in large numbers per die... >> >> No doubt there are good reasons why that is harder than it seems. >> >> John Savard > > Compound semiconductors, in addition to defects, tend to make only > N-type devices. So, no equivalent of CMOS. > > Silicon is unique in that SiO2 makes a great implantation mask, so the > photolithography process works really well. > > I don't think anyone has sold even an opamp from SiC. There are > microwave ICs made from InP and GaAs and such, but tend to be > relatively simple, with distributed amps being the high end of > commercial products. > > http://www.semiconductor-today.com/news_items/2014/AUG/KTH_010814.shtml > > http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=6019027&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D6019027 > > That is NOT very impressive so far. > > > Gigabit Logic once made GaAs logic chips, long gone now. > > SiC fets on the market now are slow, due to high gate contact > resistances. They need a lot of gate voltage swing, too. GaN seems to > be better. > >
...idiot mumblings here... Sapphire substrate for isolation, silicon Nfets and Germanium Pfets sputtered / CVD on for CMOS-like circuitry?