Log in

No account? Create an account
More progress on the computer - Alex Belits
More progress on the computer
The power supply filter bracket arrived on Monday, so I was able to assemble the power supply and turn it on. Without fans because the fans connectors board wasn't ready yet, and existing fan connectors on the power supply board aren't reachable in this design.

Another bracket, one that holds the case walls together, wasn't as good, its geometry either was made with an assumption that bending forces can be passed through a half a millimeter of air, or some welding was missing.

Then I have spent time working on the aluminum brackets that hold rj-45 connectors and the LCD board. After one failed attempt I have made two brackets, one for the console and LCD, another for four ethernet ports. The good news are, both brackets, doubling as EM shields, took exactly as much space as was allocated for them, hold all the connectors securely, cover all boards, and require two screws for LCD/console, and one screw for ethernet. The bad news are, the replacement of power supply, or HD cables (IDE or SCSI), can't be done without taking out that single screw and lifting the ethernet bracket, pulling and folding over four cables that are attached to it, then removing the power supply that is held by three more screws. In my opinion, if anyone is doing such a major replacement, one more bracket and a single screw shouldn't be a problem. After all, all four screws are strictly vertical, and the only nontrivial operation is lifting the bracket and folding the cables over, so they open the access to the power supply and the last screw that holds it. The last part, a screw under the bracket, is an unintentional but nice foolproofing measure -- it's impossible to lift the power supply while the ethernet cables are still hanging over it, something that would damage the jacks.

Cables routing didn't give me any surprises, though I think, folding ribbon cables to pass through all tunnels, and place all connectors against the drives and boards, should be considered a primitive form of origami.

Now, to make this thing run in its full configuration, I should finish (what at this point means drill, etch and solder) the fan power connectors board, the simpliest of three custom PCBs that I have in this design. It's nothing but a bunch of connectors attached to 12V, and with tachometers outputs connected to the wires going to the motherboard, so it will be able to monitor them. This version has no fan speed control, however it can be added if necessary. If anyone is curious what software I use for a simple PCB design, it's an old "PCB". I probably would use Eagle if it was more complex, but for this purpose "PCB" is sufficient, and I am not afraid of Xaw-based GUI.
2 comments or Leave a comment
j_b From: j_b Date: May 7th, 2004 12:31 am (UTC) (Link)
Is the target mobo for the first revision SATA or regular ATA? Rounded cables potential? (I've heard reviews against them)
abelits From: abelits Date: May 7th, 2004 01:59 am (UTC) (Link)
First model will be regular ATA, however both should be possible -- cable "tunnels" under the power supply can pass either. Cables have to be flat ribbons, the whole thing relies on the wide cables being folded, and fitting under the drives.

The only place where a rounded cable can make things easier in this box is a short piece between drives, however the current version has everything flat.
2 comments or Leave a comment