Jim, thanks for the feedback. However, I've been doing some testing on my approximately 3-year-old ASI120MM using a new Acer Aspire Core i3 notebook and the situation seems to be more nuanced that I could have expected.
For several days I tried to get the camera to work well on that notebook running a clean copy of Windows 10 Home (Anniversary Edition) and I could never get it to work reliably for guiding using PhD v2.6.2dev7. Although I saw some connection problems and disconnects during the first 30 minutes of testing those problem eventually went away (completely) and I proceeded to do some guiding tests indoors using an artificial star. Well, over two days I did about 10 hours of guiding and I had consistent problems with the horizontal frame sync on the camera's output (frame splits about 1/3 of the way from the right edge of the image). When that happens PhD loses the guide star and then you might have to disconnect from the camera and reconnect to get the camera's output to return to normal. I did a lot of testing to try and fix that problem, rebooting the camera and Windows, trying the "compatible" firmware that ZWO offers (didn't fix the frame split problem and even introduced new and undesired behaviors, so I returned to the standard firmware), different USB cables, then trying to use a USB2 hub instead of a direct connection to the notebook's USB2 port, switching from the ST4 guiding port ("on camera") to a direct serial connection to the mount, and other things. Nothing worked and I had these frame sync errors on average about every 15 minutes (sometime more often, sometimes more like every 20 minutes).
I should also mention that I'm using the latest version of the "ZWO ASI Camera" driver from within PhD.
Well, since I had seen this exact same problem on two other computers (an Asus notebook Windows 8.1 and an Intel Compute Stick running Windows 10 Home) I was just about ready to give up and then I tried something that was suggested by jsmoraes in post #8 (above). So, I uninstalled the ZWO driver software and then installed the software using that software's Windows 7 compatibility mode (through the preference pane for each installer app that you can access through the app's desktop icon) . Then, in kind of a "shot gun" approach I decided to try a third USB cable. Amazingly, for the next day and after at least another 6 hours of testing I had zero errors while guiding using PhD.
At first, I said to myself that it can't be the installation using the compatibility mode, it must be the USB cable. So, I switched back to the cable I had been using previously and everything worked fine. Then I switched back to using the ST4 guider port, then I started stress testing the connection, rebooting, stopping and the restarting the guiding again and again, disconnecting the camera and then reconnecting (both through software and by disconnecting the USB cable), installing a USB hub between the camera and the computer (usually not recommended), then installing a second hub between the first hub and the computer (daisy chaining through two hubs), then trying a 25' unpowered USB repeater while plugged in to the hub. Well, everything continued to work without a single hiccup.
So, you might think problem solved? Well, not so "fast" because then I went back to the Intel Compute Stick which I also could never get to work and amazingly that system was also now working (without ever having changed anything on that particular installation).
Actually, the situation is more complicated that this because when I first got the camera about three years ago I was using a small Acer netbook running Windows 7 and I used that configuration for probably over a year and never had any problems with guiding (at least not with anything that appeared to be related to either the camera or the computer). In fact, when I couldn't get my Asus Core i3 notebook with Windows 8.1 working with that camera (because of the split frames) I would often go back to the netbook running Windows 7 and the camera always worked correctly under that system (as it still does today, since I tested that configuration just in the last week).
So, what does all of this mean? As I see it there are at least three possibilities.
First, the camera itself may be changing its characteristics, meaning the hardware is just unstable and you may have periods when it will appear to work and then for no apparent reason it just stops "behaving" and nothing you can do will make it work again. That seems somewhat unlikely although I have had a few nights when guiding did work even on my Windows v8.1 Asus notebook (but very rarely, and it aways did return to the split framing errors on the next night or perhaps after a week of inactivity). The problem with this theory is that the camera always seems to work on the netbook running Windows 7 (even during periods when it wouldn't work on the other computers).
Possibility two, it really is the installation under the Windows 7 compatibility mode that fixed the problem on my new Acer Core i3 notebook (thanks jsmoraes), and the fact that it is running fine under the Intel Compute Stick is just a random fluke. I guess that is a possibility but it seems highly unlikely. Unfortunately, I can't go back and test on my other notebook computer (the one running Windows 8.1) because that computer now has a hardware fault in the motherboard (it crashes very frequently, basically that 18-month-old computer is now pretty much nothing more than a door stop).
Possibility three, the problem is in the ZWO drivers and it just depends upon where those drivers are loaded in memory (the drivers may be doing something like corrupting their own memory from a buffer overrun or some type of other fault in the handling of their resources). Actually, this scenario matches every behavior that I've seen but I have no real way of knowing for sure whether this is true. So, the reason why the Windows 7 compatibility mode seems to work is because the drivers are being loaded into a different area of memory, or the order of the memory allocations has been changed, etc. Given this, it's certainly possible that on any given computer the driver might work or it might not and this behavior could change with any modification to that same computer's runtime. I'm going to continue to investigate this to see if I can confirm whether this may be true.
Of course, if case three is true then their will be no way to fix this problem with certainty until ZWO looks at (or fixes) their driver code and that seems a very unlikely outcome because I think ZWO has basically stopped supporting that line of their software. What I mean is that all of ZWO's other products are using a new set of driver code and the USB2-based ASI120MM and ASI120MC are still using old, camera-specific drivers. I guess I should add one other note here, I also have the ASI120MC color camera and that also stopped working once I switched from using my old Windows 7 netbook, so it seems unlikely that I have a simple, one-off hardware problem with the mono ASI120MM.
Lastly, it's interesting to note that right now my ASI120MM seems rock solid and it continues to work no matter what I do, even things that ZWO recommends against, like using the camera through a USB hub, or using a USB repeater, or using long runs of USB cables between the camera and the host computer. Thus, at least in my case, I think those kinds of setups can actually work as long as the camera and drivers are functioning correctly.
Edited by james7ca, 17 December 2016 - 08:19 AM.