'saw this thread a bit late (...)
Just for info: if one plans using an industrial motherboard and in particular the LVDS video connector (assuming you can get your hands on the corresponding LVDS cable...), pay attention to have a board where the BIOS eventually supports inverting the signal that controls the backlight's intensity... with a Kontron board, I had to 'invert' the signal with a little home-made circuit... (not a drama, but if you can avoid it with a BIOS settings...).
Another point is the compatibility of the front panel controller (using a Cypress MCU) with the USB ports on the motherboard, my personal experience shows that anything more recent that Intel's series 6 chipset will NOT work...
And yet another feedback: with a WR6K, I had mixed success with W7 and a quad-core CPU (I7 3770T), I got periodical freezes, requiring hard reset... a Core 2 Duo brought much less problems, but I couldn't isolate the problem to be CPU-bound...
Have fun modding your Lecroy beasts !!
I've spent some time modding my WR6100A, testing various approaches and platforms and my conclusions are as such:
* bought Asrock IMB-181-L ITX industrial motherboard (
http://www.ebay.co.uk/itm/382070632688 - the price is good) and Core i7-4770. The LVDS worked OK, but I didn't check the backlight control pin, just wired 12V and jumpered the control on the CCFL inverted provided by LeCroy (8mA23002 by ERG).
* I've also upgraded the LCD to 8.4" XGA (1024x768), Mitsubishi model AA084XA03, with some rewiring and swapping the connectors the CCFL inverter can power this panel without a trouble. I also bought some random 8.4" resistive 4-wire touch digitizer. This is a big improvement, but unfortunately despite the same display sizes, the mechanical details are different, therefore a lot of butchery of the plastic and metal, both on the panel and LeCroy side was required - not for the faint-hearted (had to disassemble the panel into layers for example or dremel LeCroy's plastics) and doesn't look very pretty inside.
* crappy Iwatsu front panel controller will work with Intel up to Q45 chipset, later the introduction of rate matching hubs messes something up and even plugging old USB hubs inline won't help. There is a way out of this by installing a separate USB controller. In my case it was mini-PCIe card (only left PCIe slot on an ITX board) with µPD720202 USB 3.0 controller and the front panel connected to the USB 2.0 lines. Now the front panel works.
* on the subject of front panel devices, I hate the touch screen chip there (UR7HCTS2-USB), some long obsolete IC for which there is no driver available. It is not provided with X-Stream, LeCroy only provides tbupdd.dll file (no inf files or calib. app) with XStream, which I guess is used by the X-Stream app to launch the calibration app. The drivers are made by a company called Touch-Base (
https://touch-base.com/) that develops drivers for various touch screen chips and one can request a trial-driver for a particular chip from them. It works, but every 100 clicks it shows an annoying message. That's why I'm thinking about replacing this chip with Microchip AR1100, it also uses Touch-Base drivers (therefore compatible with X-Stream) but Microchip provides them free of charge.
Now the scope PC is super fast, has high resolution display over LVDS, both front panel and touch screen work, but there is a big elephant (or two) in the room - the PCI interface.
I bought 3 different PCIe to PCI adapters, based on various chipsets, of different mechanical and connectivity approaches. In theory all of them work, but in practice none of them. Some are super crappy mechanically, don't fit reliably. The PCI card works in all of them, gets detected, driver is loaded and beeps and the scope works more or less. Usually with hiccups and error messages, sometimes more, sometimes just SWTriggerThreshold search failed (some errors are only visible in the developer logs). Usually, after the initial errors and after few internal calibration cycles things sort themselves out and the scope works.
My guess it has something to do with shared IRQs and the different way the PCIe interrupts work. Disabling on-board devices that wanted to share the IRQ improved things a bit, but not entirely because well the PCIe-to-PCI adapter chipset shares the same IRQ and there was some other thing (mobo chipset) that I couldn't disable in the BIOS. If the LecS65AcqDrv.sys driver could be rewritten (unlikely) to accommodate for modern PCIe/IRQ approaches (maybe some extra critical sections etc), then I guess the X-Stream app would work without these hiccups.
The second elephant in the room is, well, x32. Not much use of i7-4770 and 8GB of RAM on Windows 7 x32. There seems to be no way around this, even if one, by some miracle, rewrote and recompiled the LecS65AcqDrv.sys driver for x64 then I doubt the 64-bit version of the X-Stream app has the necessary DCOM modules (dlls) for old scopes.
Therefore, I'm leaning towards going back to my intermediate platform (the original LeCroy mobo was leaky and the CPU was a joke) - Intel DQ45CB motherboard with Q9400 and 4GB RAM. Recently, I scored an SDVO-LVDS card so I'll be trying to get the LVDS out of Intel graphics this way. If not I'll have to build some DVI to LVDS contraption and use a short DVI cable loop at the back to feed the front panel LCD from one of the rear DVI ports. Don't want to use monitor scaler boards as this is even messier and I don't know if it would support my particular panel.
If I find some time I want to do a benchmark - measure the trigger rate with some intensive functions turned on, e.g. 24Mpt FFT + math + measurements and compare how much faster i7-4770 (despite running in x32 mode) is over Q9400. If it is much faster then it may be worth to stay on i7-4770 and suffer the hiccups at the X-Stream app startup.