I think your changes are really nice. I haven't used the LDD before but it has pretty good specs and is cheap for a commercial driver. The plug-in capability is really cool, making for a nicely integrated unit vs. having a sloppy mess of wiring. The Typhon was a good achievement in some ways but it really fails in other ways - ease of integration being one of them. I definitely see this as an improvement in that arena.
Can you show some photos of the board with the components removed or post the EAGLE files? It looks like the design is all SMT?
What would be really nice would be to take this a step further. Integrate a voltage reg that could step down the LDD's source DC to something the Typhon hardware could use, to eliminate the separate power supply.
How are you finding the LDD's? It looks like they only go up to 1A which kinda sucks if you want to push XM-L or even XP-G.
I'm working on the design a bit and making a few changes with component placement, just to correct a few "bugs" that I found in the first board run. Being that this was my first attempt to marry the "Typhon" to a sub PcB, I want to correct any flaws before I post the build files. The Atmega chip, Ds1307, UA7805 regulator, capacitors, and resistors are all SMT. The Screw headers, watch crystal, battery holder, and LCD contrast trim pot are all "through hole" I plan to convert more of the design to SMT as I source more parts for the next batch. I now prefer to work with SMT components, as all the soldering can be completed in one step.
I thought about integrating the Typhon's power source with the LDD's, but didn't for a couple of reasons-
1.) Cost of the additional components.
2.) Keeping the "Typhon" powered on it's own supply allows me to use it to "turn off" the 48V supply at night.
I really like the LDD's! They're hard to get right now, but at $5 a piece they can't be beat. I can barely build a CAT4101 based driver for that cost! and the LDD will drive 2X the number of Led's!
As for their 1000ma output? we're still working on that. I've managed to parallel drive a couple and make 2,000ma output (Hello XML's), but the jury's still out if my wiring methods will result in long term performance or not.