• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Custom PSU Cabling Q

SoulkeepHL

SoulkeepHL

2[H]4U
Joined
Jun 16, 2001
Messages
2,571
I have a big sever thats still a tangle of wires on the insides despite my best attempts, so I was thinking that doing some custom PSU wires for the drives would clean things up a good bit, my only worry is over voltage drops. I've got 13 drives, one on top of another, that would clean up the tangle of splitters that it is right now if I reduced it to a single run and a lot of right angle molex connectors, but I'm worried about the draw. Any advice grounded in ohms law is more than welcome.
 
SoulkeepHL said:
I have a big sever thats still a tangle of wires on the insides despite my best attempts, so I was thinking that doing some custom PSU wires for the drives would clean things up a good bit, my only worry is over voltage drops. I've got 13 drives, one on top of another, that would clean up the tangle of splitters that it is right now if I reduced it to a single run and a lot of right angle molex connectors, but I'm worried about the draw. Any advice grounded in ohms law is more than welcome.
Click to expand...

If you do a parallel run of connectors from drive to drive right to the PSU I don’t think wire length will be an issue (resistance wise), in fact you will probably have lower resistance by dropping all those jumpers. 16-gauge wire would be a nice touch as well, again probably better then what the jumpers use. Don’t think OHM’s law comes into play here, anything you do will improve things.;)

Since your modding, you could clean up the fan wires as well. Should look nice when you’re done.

That many drives makes one heck of a start up load, what PSU are you using?

Luck
 
BillR said:
If you do a parallel run of connectors from drive to drive right to the PSU I don’t think wire length will be an issue (resistance wise), in fact you will probably have lower resistance by dropping all those jumpers. 16-gauge wire would be a nice touch as well, again probably better then what the jumpers use. Don’t think OHM’s law comes into play here, anything you do will improve things.;)

Since your modding, you could clean up the fan wires as well. Should look nice when you’re done.

That many drives makes one heck of a start up load, what PSU are you using?

Luck
Click to expand...

When you say parallel, one set of wires per drive? one set for the entire thing?

And the PSU is a 3Y Power Technology AP3450, which is a 2+1 redundant psu, each module capable of 450w, and a combined active 900w. It's a little weak on the 12v though, only 30a, I'm pushing it towards its limit with the 14hdds and 1 optical drive. Plenty of overhead on the 3.3v and 5v lines though (60a and 100a respectively).

Its giant.

DCP02782.jpg


Thats a standard atx psu sitting on top of it.
 
Heh, helps to have all the info ;)

Ok, the question now becomes, for that entire PSU is there only one complete wiring harness or more then one?

That will make a difference how to give you a good answer to your question.

I’m assuming (always dangerous I know) that it’s just one harness set. If that’s the case we have to assume that there is one “total” 12v rail available. If you can verify that then you would build a harness from the usual 4 wires simply jumping from drive to drive.

Since there are probably more then one full sized 12 volt set of wires you could combine two (match the colors, solder each to a longer wire the next gauge up use heat shrink for safety and neatness) and use those just for the drive array. Use the rest as you normally would.

My explanation may need clarifying, so please ask. Sometimes it’s hard to type what you can see so clearly in your head

That is a lot of drives, but apparently it boots, so that part is good. Boot time is the max draw current wise, then things settle down and the load drops.

The only thing I would suggest, and I can’t stress this enough, would be a huge UPS if you don’t already have one. That is a lot of drives and data to have take a dump for lack of proper protection. Nobody’s AC is that good :D
 
It has a single harness ;)
have a few N+1 supplies myself, its a shared hotswap bus

I agree with BillR, actually removing the wye splitters will offset any added resistance a slightly longer wire would have, but your likely shortening that as well

redoing it provides you with the opportunity to upgrade it quite alot, ensuring better pins, connectors (soldered and criped) and even shielding if you choose

most HDDs will drop to a quarter of their spinnup draw
unfortunately since that occurs before any software is in place to monitor it
observing rail flutuation at spinup requires a multimeter

that is the reason its so common to see a delayed spinup option on many SCSI controllers and even a few top shelf ATA RAID controllers
 
Yeah, the 3ware 7506-12 supports staggered spin up, and I have a big excel spreadsheet detailing the power usage of the system (2.8 a @12vdc max at spinup for the drives, .97a @ 12vdc nom. for seeking). I'll have to make this my christmas project then. And the psu does indeed have a shared backplane, complete with 3x atx connectors (two of which I use :D). Thanks for all the input.
 
You must log in or register to reply here.
Back
Top