+-----------------------------------------+ | MIPS Operating Parameters During IOC/SV | +-----------------------------------------+ KG Oct. 27, 2003 Campaign G-K bias/stims JS Oct. 29, 2003 Notes re. sides, modules, bias/reset; posted on web Bias/stims predicted for O, P JS March 1, 2004 Note re. bias change on 70um array Operating parameters for the MIPS Ge:Ga arrays were changed numerous times during IOC/SV in response to the offset and elevated noise on side B of the 70um array, and the unexpectedly high responsivity of the 160um array. The tables below summarize those changes and others in the areas of: Bias and Stims Vreset (affects bias) Voffset Anneals See the NOTES towards the bottom for more details. Current version of this is available at: http://mips.as.arizona.edu/Rincon/Released/IOC/TestTrack/OpParam.txt Bias and Stim Settings ---------------------- Bias given vs. side on 70um, and for modules 1, and 2-4 on the 160um, for campaigns G-K. After that, it is broken down in more detail (see notes). 70 micron 160 micron IOC/SV V_bias mV stim V_bias mV stim Campaign sA sB DAC m1 m2-4 DAC ----------------------------------------------------- past: G 40 40 0x33 25 30 0x42 H 40 40 0x33 25 39 0x42 I 40 40 0x33 25 39 0x42 J_1* 55 55 0x2D 10 15 0x43 J_2* 55 55 0x2F 10 15 0x37 K 30 22 0x3C 10 15 0x4D O_1 62 50 0x2D 10 15 0x4D O_2 62 50 0x2F 10 15 0x37 P 62 50 0x2D 10 15 0x4D ------ Real Ge Anneals started here ----- Q* 62 50/62 0x2D 10/15/15/18 0x4D -- 70/160 Frames set to good side for Phot modes -- R 62 50/62 0x2D 10/15/15/18 0x4D future: V W 5MC* 43/35 23/25 ------------------------------------------------------ *J: stim DAC was mistakenly reset after mips-666 ran in campaign J. *O_1 / O_2: Got the stim settings we wanted at the beginnin of the campaign, but a mips_ier180 reset them to the wrong values when it completed. Those were used for the rest of the campaign. *Q: 70um B: mods 5-7 @ 50mV, mod 8 @ 62mV; Side A 62mV 160um: module biases of 10/15/15/18 (mods 1-4) *5MC: Array bias reduced from 62 to 35mV to aleviate problems with Stim latents. No change to stim DAC, resulting in about 6k DN/s in the stim flashes with the new bias. 70um A: module 1 bias increased by 8mV by tweaking its Vreset value. Mods 2-4 left with nominal, 'optimized' Vreset (i.e gives zero output for zero bias). 70um B: modules 5-7 @ 23mV due to vreset tweak, module 8 at 35mV. Vreset ------ Voffset ------- Anneals ------- +-------+ | NOTES | +-------+ Array Structure: Sides, Modules, and Flips ========================================== Prior to launch: ---------------- 70um Flight: Side A = right 1/2 Side B = left 1/2 Modules 1-8 decreasing from left to right Modules 4x32 blocks, long axis vertical Stim hot-spot in lower-left quadrant Cadien pixel in lower-left quadrant Shorted/low-response readout [5:8,0:8] Offset readout [16:20, 0:8] 160um Flight: Side A = left 1/2 Side B = right 1/2 Modules 1-4 increasing from left to right Modules 5x2 blocks, long axis horizontal Stim brighter at left by ~ x4 Offset readout bottom row, pixels 6-10 from left After launch: ------------- Arrays were flipped left-right starting at around Campaign J. All data reprocessed after that (~ Oct 20, 2003) have this flip applied. Bias and Vreset =============== Bias is set at the 1/2 array (side) level by V_Bias. The 70 and 160um arrays are divided in 1/2 vertically as described above. Bias is also determined on a module level by Vreset. Bias and Vreset were tuned to try and optimize response to both sky and stims, to improve noise performance on the 70um array, and to lower responsivity on the 160um array. The 70um noise that became apparent at the time the offset pattern appeared (campaign F) was determined to be additive in nature, at least for short (1-10 second) timescales. Increasing Vbias helped overcome the additive noise component on the data ramps. Voffset ======= Voffset is an additive offset applied to the DN. Voffset was increased on side B of the 70um array after the offset pattern appeared there in campaign F.