> 140c ARM M0 based board

[revengeofpodus]

Hi All I am new.
I am also new to my job and company.

MY job is running the daily operations of a Fabless semiconductor start up. We deisgn chips that can function while really , really hot   - 250c for thousands of hours  +2000hrs so far. While I have plenty of experience running a business, I have little in the IC world.

So without wanting to look dumb I am trying to find a way to educate myself. On A) why its even a big deal that we can do this B) why anyone would ever need it?

MY company is testing  18mb SRAM at 250c and it has been running continuously while at 250C and passed 2000 hours at temp. We also built an ARM M0 and plan to test it at the same temp, there are high temp boards in development where the entire board will be on a ceramic sub straight and every device is 250c capable (its seems they can redesign any circuit to be high temp)

 IT all sounds really exciting... but why would anyone want a board running at that temperature? we keep talking about down hole oil drilling- and some of those holes they drill do get to 250c.. but I still don't get the business model...?

I keep wanting to ask what is wrong with 150C? I know cars use chips and boards, and while not everything in the car gets 250C a lot of it gets 150c.. but more so why not just add cooling to the board (like the video game computers have). Anyone here really know why being able to operate at up to 250c would be considered revolutionary.. ?

any insight into computer chips in hot places would be greatly appreciated ( I would like to be able to have more conversation than just employee reimbursments or equipment purchases..

ROP   

[AndyC_772]

Cooling a board is tricky if the entire evironment is hot and there's no supply of cooling air. You can cool parts locally to some extent using a Peltier heat pump, but they generate heat of their own which has to be managed somehow.

Most electronics can be located somewhere more benign - so, for example, if you're monitoring and controlling an internal combustion engine, the ECU doesn't have to be in the engine bay. The sensors themselves do, though, so if you've got something producing a small signal that needs to be locally measured, digitised and sent up a cable, then the hardware to do that needs to tolerate whatever the temperature happens to be in the place where the signal is available.

Down hole electronics do certainly fall into this area too, and there's certainly no ready supply of cooling air half a mile underground. It's a bit of a niche industry, though. Not too many customers... though those which do exist can afford expensive kit.

Are you licensing an ARM core and making a high temperature chip yourselves, or building modules with other peoples' chips?

[revengeofpodus]

thank you for the reply.. The company licenses the ARM M0 ( I do the utilization reports) and we build the devices ourselves at foundry fab. The goal is to build everything on the board with our High temp technology. I am happy to PM you the website with the current high temp data..(i do not know the forum rules on that kind of stuff)

I am curious about a sensor needing "local measurement" why not, as you said, put the board in a remote place and run a wire? I don't understand the concept of local measurement..

any response would be greatly appreciated.

rop

[SaabFAN]

Every wire is more or less vulneral to interference, and also causes signal degradation with its parasitic resistance, inductance and capacitance.
Once the signal is digitized, this doesn't matter anymore, but if, for example, the ECU in your car gets a distorted signal from the Lambda-Probe because of too long wires, the engine will probably not meet the newest emission-standards, barring you from selling it in all the countries that demand those standards to be met.

High temperature-ICs sound interesting btw. Can you tell us the name of the company you're working at? Might come in handy to have that information available should the need for those components arise

[chickenHeadKnob]

The way you have worded your initial post sounds fishy to me. You are the manager of chip IP company and you don't know what end markets your products are good for?!!  It sounds more like you are a salesman who wants to prepare for a job interview at such a place.

The one end market that I know of that needs high temp operation is "down hole" drilling instrumentation. I wrote some software for such a thing a long time ago and was glad I didn't have to worry about the hardware much, but the temperature requirements were pretty extreme.

[revengeofpodus]

The way you have worded your initial post sounds fishy to me. You are the manager of chip IP company and you don't know what end markets your products are good for?!!  It sounds more like you are a salesman who wants to prepare for a job interview at such a place.

The one end market that I know of that needs high temp operation is "down hole" drilling instrumentation. I wrote some software for such a thing a long time ago and was glad I didn't have to worry about the hardware much, but the temperature requirements were pretty extreme.

I am not the manager of a chip IP company.. I am doing the daily business operations management .. there is more than one side to a IC company. I think if you trucked down to the operations , AP, payroll, , even HR of that down hole instrumentation company you said you worked for, you would find may people who are not intimately knowledgeable of the technology process the company does.

but thank you for the response

[revengeofpodus]

Every wire is more or less vulneral to interference, and also causes signal degradation with its parasitic resistance, inductance and capacitance.
Once the signal is digitized, this doesn't matter anymore, but if, for example, the ECU in your car gets a distorted signal from the Lambda-Probe because of too long wires, the engine will probably not meet the newest emission-standards, barring you from selling it in all the countries that demand those standards to be met.

High temperature-ICs sound interesting btw. Can you tell us the name of the company you're working at? Might come in handy to have that information available should the need for those components arise

PM sent.. I have to be wary of the AUP

[electr_peter]

I think high temperature technology can be employed in areas such as device/equipment black-boxes (data retention in fire), sensor/robotics applications for fire fighters (robots/cameras in fire) and in chemical industry (sensors in high temperature chemical environment).

I wonder about soldering and connections to be rated to at least 250-300C. I do not know common solders to be suited for such temps, I recall reading an article about >350-400C rated solder composition.

[Howardlong]

I think high temperature technology can be employed in areas such as device/equipment black-boxes (data retention in fire), sensor/robotics applications for fire fighters (robots/cameras in fire) and in chemical industry (sensors in high temperature chemical environment).

I wonder about soldering and connections to be rated to at least to 250-300C. I do not know common solders to be suited for such temps, I recall reading an article about >350-400C rated solder composition.

Indeed, not to mention all the other parts also required to function, none of which will be COTS. Sounds like a very limited market, i.e. low volume, very high value.

[revengeofpodus]

I think high temperature technology can be employed in areas such as device/equipment black-boxes (data retention in fire), sensor/robotics applications for fire fighters (robots/cameras in fire) and in chemical industry (sensors in high temperature chemical environment).

I wonder about soldering and connections to be rated to at least to 250-300C. I do not know common solders to be suited for such temps, I recall reading an article about >350-400C rated solder composition.

Indeed, not to mention all the other parts also required to function, none of which will be COTS. Sounds like a very limited market, i.e. low volume, very high value.

cots = commercial off the shelf correct??

 That term is used frequently. I understand the value proposition of replacement of the commercial products, but like you mentioned the application seems constrained since the entire system would have to be high temperature, not just the "juicy high dollar" chips. currently we only produce SRAM and an ARM M0 but my question (that I have kept to myself - and reason im am here) is wont those chips need the support chips to also have the same capability.

OR

like mentioned above is there basic function that a single chip (or perhaps 2) could perform without having to build the entire system.. IE digitizing sensors... does that seem more practical ...?

I have don't quite a bit of research on Semi;'s and thay all seem to specialize on a device or device category, but the "board" consists of many devices. I have yet to discover a complete system Semi company.. with our costs per device I just don't see the biz model where we could afford to develop (much less produce) the entire boards of chips..

SO I am trying to learn about basic digitizing (as referenced above) a minimal board of a single (perhaps 2 chips) that does ....? I keep getting drawn to the internet of things concept...