Manually testing the shutters and FWs.
We disconnected the CCD camera so we could look through the telescope along the optical axis, at a lamp. We ran the camrea so that the internal shutter opened and closed regularly, and then manually changed the color and ND filters.
Strange behaviour was observed: Setting the color of the filter was not a simple affair - the colour would be a bit random, and the ND filter would reset - apparently by itself - after a few seconds.
We will try to run scripts next, to see if the command of the devices from the scripts are more consequential and whether the ND resets in the middle of a sequence. Our experience from masses of observations from MLO (before the lightning struck) is that stacks of images are all OK as long as the first image is all right.