If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Rate Thread | Display Modes |
#1
|
|||
|
|||
Power line splitter
I've googled for this, and got mixed advice. So I turn to the
experienced people in this group who've probably done this many times. I want to add an internal HD. I have a bundle of wires coming from the PSU; one of them has a female Molex plug on it, and a line with male converter goes to the existing HD. Can I change this to a split line, and use the second line for the new HD? This splitter https://goo.gl/Gt58ri Ed |
Ads |
#2
|
|||
|
|||
Power line splitter
Ed Cryer wrote:
I want to add an internal HD. I have a bundle of wires coming from the PSU; one of them has a female Molex plug on it, and a line with male converter goes to the existing HD. Can I change this to a split line, and use the second line for the new HD? Your PSU is almost certain to have enough capacity to stand one extra drive, obviously if you took it to the nth degree and kept splitting, and splitting it would overload at some point. |
#3
|
|||
|
|||
Power line splitter
"Ed Cryer" wrote in message news I've googled for this, and got mixed advice. So I turn to the experienced people in this group who've probably done this many times. I want to add an internal HD. I have a bundle of wires coming from the PSU; one of them has a female Molex plug on it, and a line with male converter goes to the existing HD. Can I change this to a split line, and use the second line for the new HD? This splitter https://goo.gl/Gt58ri Ed Can't think of any reason why not, unless you have a bare-minimum power supply. I've used similar splitters for SATA and IDE drives (most recently on a Lenovo that didn't have enough connectors for 2 HDDs and a DVD drive). -- SC Tom |
#4
|
|||
|
|||
Power line splitter
On Thu, 7 Jun 2018 12:38:08 +0100, Ed Cryer
wrote: I've googled for this, and got mixed advice. So I turn to the experienced people in this group who've probably done this many times. I want to add an internal HD. I have a bundle of wires coming from the PSU; one of them has a female Molex plug on it, and a line with male converter goes to the existing HD. Can I change this to a split line, and use the second line for the new HD? Been there, done that. Worked fine. -- Someone who thinks logically provides a nice contrast to the real world. (Anonymous) |
#5
|
|||
|
|||
Power line splitter
Ed Cryer wrote:
I've googled for this, and got mixed advice. So I turn to the experienced people in this group who've probably done this many times. I want to add an internal HD. I have a bundle of wires coming from the PSU; one of them has a female Molex plug on it, and a line with male converter goes to the existing HD. Can I change this to a split line, and use the second line for the new HD? This splitter https://goo.gl/Gt58ri Ed If you need to build a "tree" off one cable, and need a large fanout, you should use Molex to Molex Y cables for the initial fan. The idea here, is this might resist burnt connectors a slight bit better than using SATA. Molex ------ Molex -------- Molex ---- SATA 15p \ \_______ Molex ---- SATA 15p \____ Molex ________ Molex ---- SATA 15p \ -------- Molex ---- SATA 15p Using Molex to SATA Y as you propose, is perfectly good for *direct* usage, because the SATA is on the end of the run when doing this. If the loom had three Molex on it, you could power six SATA this way, but running six HDD off one cable would likely be a mistake. The connectors won't burn in this picture, but the drive at the bottom could be getting close to the 11V voltage mark. At some point, the skimpy cabling on modern PSUs, has too much voltage drop. 12V | | ------- +- Molex ------ SATA 15p ^ | \ | v \____ SATA 15p | | 4 drives is probably "enough" +- Molex ------ SATA 15p | | \ | v \____ SATA 15p v | ------- +- Molex ------ SATA 15p ~11V \ --- Check for voltage drop if \____ SATA 15p adding too many... The thing to be avoided, is chaining a bunch of SATAs in a row. If you find a male SATA to two female SATA, then that would be an "ampacity fail" in terms of a design. SATA has a 3 amp limit, and some SATAs on the left could end up carrying more than that. The DC peak current on hard drives is actually a bit variable, and not all that easy to predict (short of finding the detailed drive spec sheet that has an actual oscilloscope trace). SATA ------ SATA -------- SATA ---- SATA 15p \ \_______ SATA ---- SATA 15p \____ SATA ________ SATA ---- SATA 15p \ -------- SATA ---- SATA 15p Molex has a 6 to 10 amp limit, partially influenced by the wire gauge crimped into the connector, partially influenced by how many pins are next to one another. I might split the difference and claim it can carry 8 amps or so - but I have had a Molex "burn" with only 5.5 amps or so (video card). If a Molex works itself loose, so the connection becomes bad, that's how it burns. Actual authentic Molex brand would never do that (they have extremely high insertion force), while all the knockoffs don't have nearly the same insertion/retention characteristic. The loom can use various gauges of wire. From little surprises I've received here, around four drives running off a single cable run from the PSU is about the limit. At some point, there is too much DC voltage drop along the cable. You might need to use three or four Y's in a row, to encourage "bad electrical hygiene" when dealing with HDD loads. Hard drives have two voltage sensors. The drive is not put in running order, until the 12V wire reaches a bit more than 11V. If the wire is beginning to "sag" under the load, the DC voltage gets uncomfortably close to 11V. The HDD senses this, and thinks the power has just come on. It resets the drive in the middle of what you were doing. This is... not good. I've had a drive here go into a "reset loop" where it keeps spinning up, over and over again, and sloppy Y cable use was the root cause. This doesn't seem to hurt the drive, so that's not an issue. If that happens, re-distribute the drive electrical loads on two cable runs from the PSU, rather than continuing to run all those drives off one originating cable run. The run I had trouble with, was powering both hard drives and fans. The fans don't draw too much current, but they encourage making the "chain" of molex Y cables too long. Then, sticking a hard drive on the very end one, gets you below 11V again. With a little common sense, you shouldn't have a problem. Paul |
#6
|
|||
|
|||
Power line splitter
Paul wrote:
Ed Cryer wrote: I've googled for this, and got mixed advice. So I turn to the experienced people in this group who've probably done this many times. I want to add an internal HD. I have a bundle of wires coming from the PSU; one of them has a female Molex plug on it, and a line with male converter goes to the existing HD. Can I change this to a split line, and use the second line for the new HD? This splitter https://goo.gl/Gt58ri Ed If you need to build a "tree" off one cable, and need a large fanout, you should use Molex to Molex Y cables for the initial fan. The idea here, is this might resist burnt connectors a slight bit better than using SATA. Â*Â*Â*Â*Â*Â*Â* Molex ------ Molex -------- Molex ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \_______ Molex ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \____ Molex ________ Molex ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \ Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â* -------- Molex ---- SATA 15p Using Molex to SATA Y as you propose, is perfectly good for *direct* usage, because the SATA is on the end of the run when doing this. If the loom had three Molex on it, you could power six SATA this way, but running six HDD off one cable would likely be a mistake. The connectors won't burn in this picture, but the drive at the bottom could be getting close to the 11V voltage mark. At some point, the skimpy cabling on modern PSUs, has too much voltage drop. Â*12V | Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* ------- Â*Â*Â*Â* +- Molex ------ SATA 15pÂ*Â*Â*Â* ^ Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* | Â*Â*Â*Â* vÂ*Â*Â*Â*Â*Â*Â*Â* \____ SATA 15pÂ*Â*Â*Â* | Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â * 4 drives is probably "enough" Â*Â*Â*Â* +- Molex ------ SATA 15pÂ*Â*Â*Â* | Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* | Â*Â*Â*Â* vÂ*Â*Â*Â*Â*Â*Â*Â* \____ SATA 15pÂ*Â*Â*Â* v Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* ------- Â*Â*Â*Â* +- Molex ------ SATA 15p Â*~11VÂ*Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* --- Check for voltage drop if Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \____ SATA 15pÂ*Â*Â*Â*Â*Â* adding too many... The thing to be avoided, is chaining a bunch of SATAs in a row. If you find a male SATA to two female SATA, then that would be an "ampacity fail" in terms of a design. SATA has a 3 amp limit, and some SATAs on the left could end up carrying more than that. The DC peak current on hard drives is actually a bit variable, and not all that easy to predict (short of finding the detailed drive spec sheet that has an actual oscilloscope trace). Â*Â*Â*Â*Â*Â*Â* SATAÂ* ------ SATAÂ* -------- SATAÂ* ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \_______ SATAÂ* ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \____ SATAÂ* ________ SATAÂ* ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \ Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â* -------- SATAÂ* ---- SATA 15p Molex has a 6 to 10 amp limit, partially influenced by the wire gauge crimped into the connector, partially influenced by how many pins are next to one another. I might split the difference and claim it can carry 8 amps or so - but I have had a Molex "burn" with only 5.5 amps or so (video card). If a Molex works itself loose, so the connection becomes bad, that's how it burns. Actual authentic Molex brand would never do that (they have extremely high insertion force), while all the knockoffs don't have nearly the same insertion/retention characteristic. The loom can use various gauges of wire. From little surprises I've received here, around four drives running off a single cable run from the PSU is about the limit. At some point, there is too much DC voltage drop along the cable. You might need to use three or four Y's in a row, to encourage "bad electrical hygiene" when dealing with HDD loads. Hard drives have two voltage sensors. The drive is not put in running order, until the 12V wire reaches a bit more than 11V. If the wire is beginning to "sag" under the load, the DC voltage gets uncomfortably close to 11V. The HDD senses this, and thinks the power has just come on. It resets the drive in the middle of what you were doing. This is... not good. I've had a drive here go into a "reset loop" where it keeps spinning up, over and over again, and sloppy Y cable use was the root cause. This doesn't seem to hurt the drive, so that's not an issue. If that happens, re-distribute the drive electrical loads on two cable runs from the PSU, rather than continuing to run all those drives off one originating cable run. The run I had trouble with, was powering both hard drives and fans. The fans don't draw too much current, but they encourage making the "chain" of molex Y cables too long. Then, sticking a hard drive on the very end one, gets you below 11V again. With a little common sense, you shouldn't have a problem. Â*Â* Paul Thanks to all who replied. I knew you wouldn't let me down. Experience and having-done-it is the best teacher. Ed |
#7
|
|||
|
|||
Power line splitter
Paul wrote:
Ed Cryer wrote: I've googled for this, and got mixed advice. So I turn to the experienced people in this group who've probably done this many times. I want to add an internal HD. I have a bundle of wires coming from the PSU; one of them has a female Molex plug on it, and a line with male converter goes to the existing HD. Can I change this to a split line, and use the second line for the new HD? This splitter https://goo.gl/Gt58ri Ed If you need to build a "tree" off one cable, and need a large fanout, you should use Molex to Molex Y cables for the initial fan. The idea here, is this might resist burnt connectors a slight bit better than using SATA. Â*Â*Â*Â*Â*Â*Â* Molex ------ Molex -------- Molex ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \_______ Molex ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \____ Molex ________ Molex ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \ Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â* -------- Molex ---- SATA 15p Using Molex to SATA Y as you propose, is perfectly good for *direct* usage, because the SATA is on the end of the run when doing this. If the loom had three Molex on it, you could power six SATA this way, but running six HDD off one cable would likely be a mistake. The connectors won't burn in this picture, but the drive at the bottom could be getting close to the 11V voltage mark. At some point, the skimpy cabling on modern PSUs, has too much voltage drop. Â*12V | Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* ------- Â*Â*Â*Â* +- Molex ------ SATA 15pÂ*Â*Â*Â* ^ Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* | Â*Â*Â*Â* vÂ*Â*Â*Â*Â*Â*Â*Â* \____ SATA 15pÂ*Â*Â*Â* | Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* * 4 drives is probably "enough" Â*Â*Â*Â* +- Molex ------ SATA 15pÂ*Â*Â*Â* | Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* | Â*Â*Â*Â* vÂ*Â*Â*Â*Â*Â*Â*Â* \____ SATA 15pÂ*Â*Â*Â* v Â*Â*Â*Â* |Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* ------- Â*Â*Â*Â* +- Molex ------ SATA 15p Â*~11VÂ*Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* --- Check for voltage drop if Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \____ SATA 15pÂ*Â*Â*Â*Â*Â* adding too many... The thing to be avoided, is chaining a bunch of SATAs in a row. If you find a male SATA to two female SATA, then that would be an "ampacity fail" in terms of a design. SATA has a 3 amp limit, and some SATAs on the left could end up carrying more than that. The DC peak current on hard drives is actually a bit variable, and not all that easy to predict (short of finding the detailed drive spec sheet that has an actual oscilloscope trace). Â*Â*Â*Â*Â*Â*Â* SATAÂ* ------ SATAÂ* -------- SATAÂ* ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \_______ SATAÂ* ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \____ SATAÂ* ________ SATAÂ* ---- SATA 15p Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* \ Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â* -------- SATAÂ* ---- SATA 15p Molex has a 6 to 10 amp limit, partially influenced by the wire gauge crimped into the connector, partially influenced by how many pins are next to one another. I might split the difference and claim it can carry 8 amps or so - but I have had a Molex "burn" with only 5.5 amps or so (video card). If a Molex works itself loose, so the connection becomes bad, that's how it burns. Actual authentic Molex brand would never do that (they have extremely high insertion force), while all the knockoffs don't have nearly the same insertion/retention characteristic. The loom can use various gauges of wire. From little surprises I've received here, around four drives running off a single cable run from the PSU is about the limit. At some point, there is too much DC voltage drop along the cable. You might need to use three or four Y's in a row, to encourage "bad electrical hygiene" when dealing with HDD loads. Hard drives have two voltage sensors. The drive is not put in running order, until the 12V wire reaches a bit more than 11V. If the wire is beginning to "sag" under the load, the DC voltage gets uncomfortably close to 11V. The HDD senses this, and thinks the power has just come on. It resets the drive in the middle of what you were doing. This is... not good. I've had a drive here go into a "reset loop" where it keeps spinning up, over and over again, and sloppy Y cable use was the root cause. This doesn't seem to hurt the drive, so that's not an issue. If that happens, re-distribute the drive electrical loads on two cable runs from the PSU, rather than continuing to run all those drives off one originating cable run. The run I had trouble with, was powering both hard drives and fans. The fans don't draw too much current, but they encourage making the "chain" of molex Y cables too long. Then, sticking a hard drive on the very end one, gets you below 11V again. With a little common sense, you shouldn't have a problem. Â*Â* Paul In a previous thread I opened about installing an SSD, you warned me about the bundling of wires, and recommended a SATA data cable with a 90° clip on the end. I found out in doing it just how correct you were. The SSD sits facing the front panel, with very little room, and the 90° clip was the answer. In addition to that I had to unbundle the wires from the PSU that some nimble-fingered, far-eastern woman had so neatly passed around the box edgings and tied up so neatly. I took out a cable I needed, wired it into the SSD, tied up the remaining wires as best I could with a piece of twisty wire, and put the box edgings and panels back on. But when I rebooted there was a clankety-clank sound, then silence until I powered down, when another slight clankety-clank sound came. I think I knew straight away what it was; and, yes, I was right. A wire hanging down in front of the fan in the fascia casing. And if you google for this sound, you get thousands of hits for similar things. Ed |
#8
|
|||
|
|||
Power line splitter
Ed Cryer wrote:
In a previous thread I opened about installing an SSD, you warned me about the bundling of wires, and recommended a SATA data cable with a 90° clip on the end. I found out in doing it just how correct you were. The SSD sits facing the front panel, with very little room, and the 90° clip was the answer. In addition to that I had to unbundle the wires from the PSU that some nimble-fingered, far-eastern woman had so neatly passed around the box edgings and tied up so neatly. I took out a cable I needed, wired it into the SSD, tied up the remaining wires as best I could with a piece of twisty wire, and put the box edgings and panels back on. But when I rebooted there was a clankety-clank sound, then silence until I powered down, when another slight clankety-clank sound came. I think I knew straight away what it was; and, yes, I was right. A wire hanging down in front of the fan in the fascia casing. And if you google for this sound, you get thousands of hits for similar things. Ed And the funny thing is, power supplies with modular wiring, really don't make things all that much better. There's still stuff hanging down where you don't want it. At least with SSDs, there's less current draw, so not nearly the same potential issue with max_current. Paul |
#9
|
|||
|
|||
Power line splitter
Paul wrote:
At least with SSDs, there's less current draw I've not generally found SSDs take less current than 2.5" HDs |
#10
|
|||
|
|||
Power line splitter
Andy Burns wrote:
Paul wrote: At least with SSDs, there's less current draw I've not generally found SSDs take less current than 2.5" HDs The ones I've tried so far, seem to draw less than 500mA. But that's not all of them. I don't own any which use a Sandforce controller (which has a compressor in it). The SSD I just tested, used 5V @ 340mA at the very beginning of an HDTune read benchmark run, and then settled down to 200mA during the majority of the run. The idle current is 60mA. The test was done in Win7. 512GB SSD 5V @ 60mA idle 5V @ 200mA read bench (a couple seconds of 340mA just before the bench curve starts rendering) The laptop drive that used to be in my laptop, draws 5V @ 180mA idle, but rises to 5V @ 450mA during the body of the HDTune read benchmark. Compared to the 200mA the SSD draws. 320GB HDD 2.5" 5V @ 180mA idle 5V @ 450mA read bench The SSD is winning by a little bit, on this relaxed benchmark. I would guess under some other circumstances, the SSD would be drawing the 340mA continuously. In any case, for the SSDs here I've tested (4), they generally stay under 500mA. My statement from a connector perspective, is that 0.34 amps is less than the 2 amps on 12V a 3.5" drive might draw at spinup (your mileage may vary). There's less danger of a chain of drives burning a connector. I could chain a fair number of 2.5" SSDs at the 0.34A level. The 3.5" HDD, once spinup is complete, generally drop back to a lower current flow level (in the old days, around 12V @ 0.6 amps). So if you want to chain a zillion SSDs across your benchtop using Y cable, be my guest. There's less danger with those. There's at least one Sandforce drive that used a peak of 7W (so 5V @ 1.4A on write or so), but not all SSD controllers are designed exactly like that. Kingston brand drives seem to use a lot of Sandforce chips, if you wanted to track down samples. The worst part about SSDs, is they don't tend to show the 200mA or 340mA number. That's left as a mystery. Instead, they show some idiotic "devsleep" number on the tin. The marketing department simply takes the lowest number seen anywhere on the datasheet, and that becomes the "figure on the box". If you're worried about running the SSD drive bus-powered on a USB2 enclosure, you'll have to measure the drive when you get it, to see if it's less than 500mA. (The idea being, that USB2 is the lowest common denominator in the computer room, and yes, I've had to do backups over that interface.) Paul |
#11
|
|||
|
|||
Power line splitter
Paul wrote in news
My statement from a connector perspective, is that 0.34 amps is less than the 2 amps on 12V a 3.5" drive might draw at spinup (your mileage may vary). There's less danger of a chain of drives burning a connector. I could chain a fair number of 2.5" SSDs at the 0.34A level. The 3.5" HDD, once spinup is complete, generally drop back to a lower current flow level (in the old days, around 12V @ 0.6 amps). So if you want to chain a zillion SSDs across your benchtop using Y cable, be my guest. There's less danger with those. There's at least one Sandforce drive that used a peak of 7W (so 5V @ 1.4A on write or so), but not all SSD controllers are designed exactly like that. Kingston brand drives seem to use a lot of Sandforce chips, if you wanted to track down samples. Paul I have five 3.5" drives in a stack in my tower case. I originally had mostly direct runs from the PSU, with a couple of Ys, but that made a really dense mass of cables. I found a kind of splitter that one drive side connector and four cable side connectors strung together along a set of separate wires in a ladder type structure. It also had some 2200 µf caps in the head connector, supposedly to stabilize the +12 and +5 lines. It has been working fine for three years now, and definitely cleaned up the cable mess that used to be there, and freed a couple of connectors for my DVD drives. Silverstone Tek Four-in-One SATA Power Connectors with Power Stabilizing Capacitors (CP06) |
#12
|
|||
|
|||
Power line splitter
Tim wrote:
Paul wrote in news My statement from a connector perspective, is that 0.34 amps is less than the 2 amps on 12V a 3.5" drive might draw at spinup (your mileage may vary). There's less danger of a chain of drives burning a connector. I could chain a fair number of 2.5" SSDs at the 0.34A level. The 3.5" HDD, once spinup is complete, generally drop back to a lower current flow level (in the old days, around 12V @ 0.6 amps). So if you want to chain a zillion SSDs across your benchtop using Y cable, be my guest. There's less danger with those. There's at least one Sandforce drive that used a peak of 7W (so 5V @ 1.4A on write or so), but not all SSD controllers are designed exactly like that. Kingston brand drives seem to use a lot of Sandforce chips, if you wanted to track down samples. Paul I have five 3.5" drives in a stack in my tower case. I originally had mostly direct runs from the PSU, with a couple of Ys, but that made a really dense mass of cables. I found a kind of splitter that one drive side connector and four cable side connectors strung together along a set of separate wires in a ladder type structure. It also had some 2200 uf caps in the head connector, supposedly to stabilize the +12 and +5 lines. It has been working fine for three years now, and definitely cleaned up the cable mess that used to be there, and freed a couple of connectors for my DVD drives. Silverstone Tek Four-in-One SATA Power Connectors with Power Stabilizing Capacitors (CP06) Be careful not to exceed the stability criterion. Too many caps spells trouble. The control loop has phase margin. Adding caps degrades margin. If you go too far, the supply could oscillate at a low frequency. This is from some old Intel guide on how to design ATX power supplies. "3.3.5 Capacitive Load The power supply should be able to power up and operate normally with the following capacitances simultaneously present on the DC outputs. Output: +12VDC +5VDC +3.3VDC -5VDC -12VDC +5VSB Capacitive load (uF): 1,000 10,000 6,000 350 350 350 3.3.6 Closed Loop Stability The power supply shall be unconditionally stable under all line/load/transient load conditions including capacitive loads specified in Section 3.3.5. A minimum of 45 degrees phase and 10 dB-gain margin is required. The power supply vendor shall provide proof of the units closed-loop stability with local sensing through the submission of Bode plots. Closed-loop stability must be ensured at the maximum and minimum loads as applicable. " That doesn't mean each supply design honors those numbers. Those numbers might be considered minimums. And the values aren't typically shown in power supply datasheets either. At work, we would have to phone people to get that kind of information. The temptation to use caps is... overpowering :-) It's a temptation you have to fight. HTH, Paul |
Thread Tools | |
Display Modes | Rate This Thread |
|
|