Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Why we dare to go without DARE (library) Innovation for life ESA AMICSA workshop TNO is active in five core areas • Quality of Life • Defence, Security & Safety • Science & Industry • Built Environment & Geosciences • Information & Communication Technology 2 ESA AMICSA workshop Noordwijk 2010-09-06 Activities in Opto-Mechanical Instrumentation for Space Applications •Scientific payload instrumentation, such as spectrometers •S59 UV stellar spectrometer •ISO Short wave Spectrometer (5-15 microns) •Herschel HIFI design and production support •Herschel HIFI cryogenic alignment camera system •Earth-observation payload instrumentation & calibration •GOME, SCIAMACHY, OMI design, production & calibration •Diffuser design, production and calibration •Optical Ground Support Equipment (GOME, SCIA, IASI) •Multi-spectral imaging spectrometer (Earthcare) •Precision Mechanics •Refocussing mechanisms for MSG •Optical Delay-Lines (Darwin technology development) •Achromatic Phase Shifters, Nulling technology (Darwin) •GAIA basic angle monitoring; Wave Front Sensor; Test eqpt • 3 Avionics Equipment and Other business • Sun sensors • Precision Optical Metrology sensors for Formation Flying • Fibre-Bragg grating for in-situ measurement of stress, deformation etc ESA AMICSA workshop Noordwijk 2010-09-06 TNO’s current portfolio • Existing Products • Current developments - SOTS • Specials and mission specific sensors 4 ESA AMICSA workshop Noordwijk 2010-09-06 Start of the miniaturisation within Microned • Autonomous micro-digital sunsensor •Autonomous power •Wireless link •MEMS based 5 ESA AMICSA workshop Noordwijk 2010-09-06 Main deliverable • • • • • • • 6 APS+ chip Single chip sunsensor Optimised for low power TSMC 0.18 micron process Last spin-out 5th May Chips received 23th of june Standard design library Several circuit design modifications to avoid SEL and SEU related issues. ESA AMICSA workshop Noordwijk 2010-09-06 Key personel Albert Theuwissen Ning Xie 7 ESA AMICSA workshop Noordwijk 2010-09-06 Key personel Murat Durkut Johan Leijtens Henk Hakkesteegt 8 ESA AMICSA workshop Henk Jansen Noordwijk 2010-09-06 Total dose radiation tolerance 0.18 micron CMOS TSMC Pixels tested up to 100 krad No significant impact observed Using DARE library would increase power consumption by factor of 2 1,0 0,9 0,8 Quantum Efficiency • • • • 0,7 2 Pinned Photodiode area/Pixel: 53 (m ) Pixel Number: 5500 2 Total Pinned Photodiode Area: 291500 (m ) Integration Time: 0,03 (sec) 0,6 Before Radiation 86krad 106krad 0,5 0,4 0,3 0,2 0,1 0,0 200 300 400 500 600 700 800 900 100011001200 Wavelength (nm) • Low power was the main requirement • 20krad is sufficient for majority of applications • For small systems extra shielding has less impact • No DARE library was used 9 ESA AMICSA workshop Noordwijk 2010-09-06 Radiation tolerance tested with gated pixel • total ionizing dose (TID) level of 31krad, 86krad, 106krad, 109krad and 137krad with an average energy of 46.2keV Number of Pixels • No significant decrease in quantum efficiency • No appreciable increase in dark current for 30krad 1000 Before Radiation 31krad 109krad 100 TG Transistor is off 10 1 0 Ref: 10 20 40 Dark Random Noise (DN) ESA AMICSA workshop Noordwijk 2010-09-06 Increased darkcurrent not significant. • Significantly lower then threshold • Noise on reading less then 0.004 degrees for ±47 degree system 11 ESA AMICSA workshop Noordwijk 2010-09-06 Single Event Latch-up in CMOS • Parasitic SCR (through substrate) • SEL if gain >1 • SEU if gain <1 • Latch up can destroy the circuit N.B. Significant current can only flow if both N+ and P+ terminals are connected to low impedance, and substrate resistance is high enough 12 ESA AMICSA workshop Noordwijk 2010-09-06 SEL hardened pixel design Winner takes all 3T pixel design • Parasitic SCR through NMOS reset transistor and PMOS source follower • Reduced QE due to N well for PMOS transistor 3.3 V n+ n+ np- p+ p+ p+ p+ N.B. Reduced QE not important for this application, because there is an abundance of light. 13 ESA AMICSA workshop Noordwijk 2010-09-06 SEL hardened winner takes all pixel design 3.3 V • All PMOS design • PMOS reset transistor is not significantly reducing QE any further • Extra ground connection per pixel to reduce substrate resistance.(decreases the effect of any SEU) n+ n+ np- p+ p+ p+ p+ 1.8 V 3.3 V p+ p+ No SCR no SEL 14 ESA AMICSA workshop n+ p- p+ p+ p+ p+ Noordwijk 2010-09-06 Ten pixel pin-hole • No false detection of the sunspot due to bad pixels or SEU • Replace affected pixels by the average of neighbouring pixels • Automatic compensation of SEE 15 ESA AMICSA workshop Noordwijk 2010-09-06 Single cycle acquisition through winner takes all hardware. • No forbidden states in the internal state machines. • Hard wired configuration straps • Automatic switchback to acquisition mode at SEU Loss of accuracy for a single cycle due to SEU is the worst consequence 16 ESA AMICSA workshop Noordwijk 2010-09-06 Conclusions • Power consumption is our main driver • Total dose radiation tolerance is sufficient for most sunsensor applications • SEL/SEU effects are deemed more critical and tackled through several design decisions (DARE) 17 ESA AMICSA workshop Noordwijk 2010-09-06 That’s why we dare to go without DARE libraries But we don’t dare to go without DARE Thank you for your attention. 18 ESA AMICSA workshop For further information Johan Leijtens +31 15 269 2191 [email protected] Noordwijk 2010-09-06