I managed to get another one of these cameras working, this time a 7290A, and in my experiments and image tweaks have developed a basic starting procedure as well as a couple of tips/things to check if you're having image trouble on your unit. I haven't worked with any other vidicon devices, and the silkscreen labeling can be somewhat inconsistent between models, so it took more trial and error than I had expected, but I've at least got a reasonable procedure.
When trying to get image out of a camera that you know is working, my first port of call is the 300V adjustment, just to verify that the 300V cap (and the various other voltages devised from the same boost converter) is biased at the right amount. This also effects the target voltage, so it's worth messing with first. Then, I measure the target voltage on the board and set the corresponding pot to 10-15V. In my limited experience these tubes will produce an image at lower voltages still, especially in the visible band, but with reduced sensitivity and it's useful to have a lower starting point to adjust later. Some cameras seem to have their tube's target voltage marked on a sticker, but this isn't always present and I've found it's not really required.
From there, we want to verify basic imaging, which is basically making sure the tube is actually reading out usable information. When powered up, the camera should spike slightly, then gradually decrease current usage after the high voltage biasing is charged up. The screen will go white after generating the sync signal, then dark (or potentially reverse) and in a few seconds the screen will sort of fade into an image after that point, usually in a sort of spreading puddle shape and sometimes including some wavy horizontal lines (more on that later). The 7290 (non A) I have takes longer to start the image fading in, but this could very well be due to aging components, as it actually takes my 7290 a couple of minutes to stabilize at full image size/correct rotation/no flickering - I just haven't bothered to track down the responsible components. The 7290A I just worked on will get an image in less than 10s from power on, but it goes through the full screen cycles then fades into the image, even with a well illuminated scene.
From the starting target voltage set, I adjust the black level pot (if the lens is facing towards you, it is the second to top potentiometer facing upwards on the left hand board) so that it is an in-between setting, then I adjust the potentiometer on the vidicon base's board (the one on the left side of the board when both pots are facing up and the lens is pointing towards you) and test to see if I can see motion in the image. Once you see motion, it's all about trying to maximize that signal, this usually involves some turning down the black level to be darker, but also turning up the target voltage slowly. More target voltage means more sensitivity and less ghosting, but too much will make the image go fully white and then, if pushed farther, fully black, and running too high can damage the tube. I like to start low and gradually increase the level until I can notice the image ghosting decrease - there seems to be a region where a little increase in voltage makes a notable change, and at a point it stops being noticeable. It is important to mention that light will leak in from the sides and you want to be setting this level as close as you can to usage conditions, so inserting the filter holder (to block some stray light) or just turning off bench lights to work with a minimum to see the adjustment pots can be quite helpful. If you're looking for SWIR or NIR only bands, it's also worth putting in that filter, so that if your target is brightly lit in visible, you don't end up setting the brightness too low in the band of interest. It is best to do these adjustments with the lens mounted and with a target that is high contrast.
Once you have good contrast on your image, you want to focus it, using the focus pot on the right hand board and the pot on the right hand side of the tube base board (again with the lens pointing towards you). I try to focus the lens as sharp as I can get, then adjust the focus pot to as sharp as I can get, then treat the right pot on the tube base board as a fine adjustment for it, since it has some similar effect on the image. This usually has to be repeated, and don't worry if the image isn't plumb. At this point I move to the horizontal and vertical adjustments on the right hand board to both try and center the image in the lens aperture (you can see the circular edges when the deflections are maxed), fill as much of the aperture as possible, and stay away from any bright or dark spots that track with the movement of the position adjustments. Remember that with the image persistence it takes a few seconds for the image to settle, so it's best to adjust, wait a second, then check, rather than try to operate in realtime.
At this point I turn to the four pots that are outward facing on the left hand board and adjust them for as even a field illumination as possible. Proper adjustment of these should cancel almost all vignetting at edges or bright/dark regions of the screen, and you can usually adjust them by watching the bright band or dark spot move around on the image and just try to null it out. At this point if your image is off the vertical axis, loosen the four small set screws that are holding the metal sleeve around the tube and gently rotate the tube in the required direction. This is best done in small increments, while power is off, and then repower and wait for the image to check your work. No need to be sticking your hands on the tube when it's biased at a few hundred voltages and the whole assembly is pretty sensitive to vibration and EM noise (coupled in by your hands, metallic tools, whatever). That basic process takes a bit of time, but has yielded good results for my image quality and can be started basically with unknown settings.
Now a couple of quirks of the design that I've realized and which could be helpful:
First off, if you have your basic voltages set, you think the electronics and tube work, but are seeing no image (and not the normal dark and white transitions that then fade into an image on power up), your tube could be the wrong distance from the lens mount. Basically, the glass tube itself can move inside the metal sleeve with the deflection coils that is anchored in place, and if the tube face is too close to the lens, it will be far enough outside of the coils for no image to appear. You can gently press on the face of the tube with a clean cloth to slowly press it farther in - in my case a millimeter or two behind the metal bracket that retains the filter mount - to make the tube set back farther in the coils, and this should get your image. Distance to the lens also effects where it can focus so it's worth testing the full range of the lens to find a good place for it, but if you need to reverse the direction and move the tube towards the lens, gently press and rotate the board on the back of the tube. The direction of rotation is not important because the image will be aligned from the coils around the tube, but turning it too far could cause access issues for the cables or to the pots.
Another thing to be on the lookout for: if you see wavy horizontal lines that look vaguely like scalloped cracks... these are not cracks. I haven't actually figured out what they are, but I've now seen two tubes that with more adjustment (and some of this is tube position in the coil dependent), even when these lines are visible at the first power on image, they go away and the image can look normal. An artifact of the scanning, I think, maybe someone can shed some light on the phenomenon.
Finally, the image is read off the tube through the light blue wire at the front left of the tube, into a SMD JFET on the inside face of that left board. Getting your hand or tools near that cable will couple in noise that is visible on screen, and leaving the screw out of the nearest mounting hole can cause grounding issues that give you a lot of garbage on screen.
I've got another 7290A to work on that's just not able to run its boost converter for more than about 2 seconds for some reason, but I hope to test out the adjustment method again on it, which has basically the same parts and layout, but earlier board revisions with several alterations not seen in the unit I've now got working. Expect to see a comparison with the 7290 (non A) at some point too, the 7290A definitely has less of a vignetting effect, but it also could be the age/condition of the electronics.