| DETECTOR READ TIME | PIXEL READOUT RATE |
| Detector | Format | Pixel size (um) | Charge shift of 1 row (us) | NON
ASTRO (sec) |
FAST
(sec) |
NORMAL
(sec) |
SLOW
(sec) |
XTRA
SLOW (sec) |
NON
ASTRO (us/pix) |
FAST
(us/pix) |
NORMAL
(us/pix) |
SLOW
(us/pix) |
XTRA
SLOW (us/pix) |
| Thompson | 1K x 1K | 18 | 12.5 | 28 | 38 | 60 | 100 | 160 | 28 | 38 | 60 | 100 | 160 |
| Tek | 1K x 1K | 24 | 500 | 33 | 52 | 75 | 120 | 394 | 33 | 52 | 75 | 120 | 394 |
| MITLL 2 | 2K x 4K | 15 | 46.5 | 60 | 95 | 160 | 300 | 940 | 6.5 | 10.5 | 16 | 32 | 104 |
| MITLL 3 | 2K x 4K | 15 | 46.5 | 58 | 108 | 143 | 285 | 924 | 6.5 | 10.5 | 18 | 34 | 106 |
You can now see that the rate of charge shifts row-to-row is typically 3 orders of mag faster than the time to read out through the amplifiers.
Note that the Thompson and the Tek timings are with the old version of the AAO-1 controller. (A small qualification is that the Thompson did not really have an XTRASLOW mode, more a CTE-enhanced mode.)
In 1998, the AAO-1 controller was upgraded with new hardware: these are the timings reflected for the MITLL chips.