As it's a rainy day, I did some bench testing this morning.
ISIS, the spectroscopy software I use with my spectrograph provides facility to check the read out noise and dark current of your CCD using flats, bias and dark frames taken through the spectrograph.
I first tested my QSI 583. It reported RON of 13e and dark current of 0.0279 e/s. no real surprises here. I was imaging binned x2 and the results compare to the RON of 8e and dark current of 0.2 e/s at -10C (i ran my tests at -15C) reported by QSI.
Then I tested my ST-10XME. RON of only 0.61 e (!) and dark current of 0.0122 e/s. The dark current looks about right but the RON is pretty low. Recognize that the gain is quite low too (0.07 e/ADU compared to 1.031 for the QSI) so the read out noise should be lower. I am surprised by the result despite this. I ran it a few times with different framing and received similar results. I plan to take some more images to check this.
Something is out of whack there, Pete. According to the specs on the ST-10XME the gain should be around 1.3e-/ADU with full well depth of ~77,000 electrons. With a gain of 0.07e-/ADU you would run out of headroom with less than 5,000 electrons!