Hello all, Is there a recommended temperature rang...
# analog-design
j
Hello all, Is there a recommended temperature range for the SKY130 technology? What if I try to test a SKY130 chip at e.g. 200°C, what could be the problem? Is there a destructive maximum temperature? Also, what's the range of validity for the models?
s
According to chracterization data on standard cells:
Copy code
sky130_fd_sc_hd__ff_100C_1v65.lib           sky130_fd_sc_hd__ss_n40C_1v28.lib
sky130_fd_sc_hd__ff_100C_1v95.lib           sky130_fd_sc_hd__ss_n40C_1v35.lib
sky130_fd_sc_hd__ff_n40C_1v56.lib           sky130_fd_sc_hd__ss_n40C_1v40.lib
sky130_fd_sc_hd__ff_n40C_1v65.lib           sky130_fd_sc_hd__ss_n40C_1v44.lib
sky130_fd_sc_hd__ff_n40C_1v76.lib           sky130_fd_sc_hd__ss_n40C_1v60_ccsnoise.lib
sky130_fd_sc_hd__ff_n40C_1v95_ccsnoise.lib  sky130_fd_sc_hd__ss_n40C_1v60.lib
sky130_fd_sc_hd__ff_n40C_1v95.lib           sky130_fd_sc_hd__ss_n40C_1v76.lib
sky130_fd_sc_hd__ss_100C_1v40.lib           sky130_fd_sc_hd__tt_025C_1v80.lib
sky130_fd_sc_hd__ss_100C_1v60.lib           sky130_fd_sc_hd__tt_100C_1v80.lib
i would say the models are characterized from -40C to 100C. A common consumer range is -40C : 125C, automotive applications require some wider range.
a
@Jorge Marin
j
cool, thanks both for the tips now, the next question would be: what if we need to go out of that range? I've tested XFAB180 up to 175°C (with flying melted soldering paste inside the thermostream capsule 😬 for the first sample which - of course - failed). Do you think SKY130 can reach those temperatures or a bit higher, too?
and also, do you know why 125°C is used as upper limit? Is there a fundamental problem for higher temperatures, or it's just the definition because normal applications shouldn't go higher?
a
@Jorge Marin That's a very hard question and you could only answer this by testing. There are many issues targeting such temperature like packaging, the devices would behave totally different. Most likely you will need custom devices even at the doping profile not just the layout level. You need a process PhD designer to work with you on this.
I don't know why 125
Also, you will need to consider the derating and the change in lattice structure in silicon at 200 degrees.
I honestly don't know what will happen.
h
Usually silicon technologies are qualified for 125degC (that is the industry standard). At this temperature, lifetime is calculated, electro-migration, device wear, etc. and guaranteed by the foundry (for e.g. 10years lifetime). Automotive silicon technologies are often qualified for 150degC (it gets hot near the combustion engine). Silicon is melting at 1415degC, so there is room for your experiments 😁. But beyond the qualified temperature, don’t expect that ICs last as long as promised, and the device models will not be valid, so expect all kind of issues in analog and digital circuits.
j
thanks a lot for the feedback!