I was looking to see if there were any HIGH temperature Silicon Carbide FETs out there and I see some R&D and some future products from a couple of years ago talking about 500 degrees C but nothing available now that has a higher die temperature spec than about 200 C.
Was this just a dream that some were trying for or did somebody actually make a high temperature part that you could buy ??
Here is one link I found but the PDF spec sheet does not come up but the picture they show looks something like what I would expect for an expensive high temp FET.
I seem to recall from reading about the regular devices that, although the SiC itself can be made to operate at very high temperatures, it's something about making the schottky contacts to the stuff, or the metallization, or bonding, or something like that, which actually has a substantially lower temp rating. So they aren't losing much by putting them in regular 150-175C plastic epoxy cases.
You're right, the stuff *should* be able to handle high temps like they say, and there must be people somewhere making them... I don't have any experience with down-hole or other extreme enviroment stuff unfortunately.
Deep Friar: a very philosophical monk.
Note the link to the datasheet is broke - translation: if you have to ask "how much" then you cannot afford it. I think Halliburton makes high temp FETs, SOS if i remember right. $400 a pop last i heard..
I see some NASA related R&D where they've run die at 500 C for 2000+ hours for aerospace and in engine applications. This was like, more than 5 years ago.
One product that showed up from a search were these guys.
A bullet point at the top of the sheet say 225 degree C max operation temperature but everything else I see in this data sheet says
175 C at the top end. How can they say 225 C max operating rempterature but 175 C is the maximum storage temperature ??
I would think that if a converter was made with parts that ~could~ run up to 500 C, it would be very ineffiecient. I can see using a high temp version in an extreme environment application but not in a consumer or industrial/commercial application.
You can get high temperatures if you isolate enough - even with a modest power over time. E.g. lift the SiC-microchip from the heat-sink.
The SiC-transistors seem to be very fast and has reasonbly low leakage current, when operated at high temperature.
You can compare Si-transistors with the old Ge-transistors - and see the SiC-transistor as Si-transistors was better than Ge-....
Oct 28, 2011, powerelectronics.com: SiC ?Super? Junction Transistors Deliver High Temp Performance:
Citat: "... GeneSiC?s SiC-based 1200 V/220 m? Super Junction Transistors (SJTs) feature high temperature (> 300 °C) operation capability, ultra-fast switching transitions (< 15 ns) ... The leakage current in the SJT at VDS = 1200 V is below 5 µA up to temperatures as high as 225 °C. Leakage currents of < 100 µA were measured even at 325 °C ..."
No Si-transistor can match the above, at that temperature!
2001, purdue.edu: Lateral power MOSFETs in silicon carbide:
Citat: "... silicon carbide is considered to be the material of choice for power switching electronics in the future ... we present the first lateral power devices on a semi-insulating vanadium doped substrate of silicon carbide. The first generation of lateral DMOSFETs in 4H-SiC yielded a blocking voltage of 2.6 kV ..."
25 August, 2004, BBC News: Door open for silicon replacement:
Citat: "... Previous research has already shown that even at red-hot temperatures as high as 650C (1,202F), silicon carbide devices can function unperturbed and without the need for cooling ... One exciting application for silicon carbide could be in deep-space missions, where nuclear power would be needed for the craft. Radiation-hardened silicon carbide devices would reduce the shielding needed to protect reactor control electronics ..."
Downhole devices have short and difficult lives, so storage conditions can last a lot longer than operating conditions.
I once spent a fair amount of effort trying to persuade a large oil services company that the right approach to downhole instruments was to build a thermally excited thermoelectric refrigerator (a bit like a pulse-tube cryostat but in a different operating regime). These are solid metal tubes with zero moving parts--just a heater, three heat exchangers, and a bunch of small stainless-steel tubes. They're used for making LNG from well gas, so making a 25 C volume in a 250C ambient is way within the state of the art.
If that was done, just once, they could use normal COTS components forever, instead of laboriously qualifying all the parts for each development. Didn't make a dent, unfortunately, but it was fun learning about it. It's still the right way to go.
The only parts I've seen for sale rely on insulated substrates to overcome some of the limitations of higher temperature operation.
No mention of SiC to do this.
Although it might be comforting to have a higher temperature capability in semiconductor switches, they don't function in complete independance, and any heat they generate or tolerate also has to be tolerated by local ancilliary hardware.
Too often, I see inquiries about high temperature hardware where there's no real justification to place parts in locations where they're subjected to this extreme.
I see what you mean from switching loss graphs for various higest temperature comparisons in your first powerelectronics.com link.
I still haven't found anything available yet that mentions higher than maybe 225 C operation that can be bought.
I have seen some of the university R&D links (and some in cahoots with NASA), and they have evidently found some good results... But I don't usually hold my breath too long for university research becoming reality in short time. We will see though. The reason I have my doubts is because I have seen R&D showing promise in papers published since at least 2005 and some earlier stuff I believe in your links here.
Some of the stuff they put down there is very clever indeed, but it's a complete waste of resources to keep doing it that way. It would cost probably $250k to develop a 38 mm diameter thermoacoustic fridge for
2-inch cased holes. That isn't much for a big oil services outfit, but there's a lot of "engineering judgement" against downhole refrigeration in general.
You can see the point--imagine a roughneck taking a free-piston Stirling cooler out of the hole, chucking it into the back of the truck with all the casings and gizmos, and bouncing along the washboard road to the next hole. It wouldn't survive long. Of course a thermoacoustic fridge is made of solid copper and stainless steel, and its only moving part is helium-argon gas fill. It wouldn't be much daintier than a length of steel casing.
The real reason seems to be that nobody wants to take the risk of having all the good ole boys rib them for years if it failed. "Puttin' a _fridge_ down a *drill*hole? 'S a long way to go for a cold beer, Bob. 'Specially since it weren't so cold, wuz it? uh huh huh. That's ol' Bob, boys, right down to the ground. Fridges down drillholes to boyl his beer for him."
see how you could cool that length with the type of cooler you're suggesting.
My customer was trying to put laser interferometers down there. You can get hundreds of watts worth of cooling at 25C, so a heat pipe plus a bit of insulation will cool quite a way down. And what in the world do you need 30 feet for? Two feet was more than enough for our stuff.