SATA power connector: which moron invented it?

[Agent24]

Re: SATA power connector: which moron invented it?

Most people don't hot-plug SATA. And those that do, most likely have properly-designed drive bays\caddies where the issues with the connector flimsiness are irrelevant, eg: in servers.

The problem with the connector lies with the average Joe building a home PC and snapping the plastic off.

[Behemot]

Re: SATA power connector: which moron invented it?

I have seen few server bays which did not have any special intermediate connector (but many do), but it would not be any more expensive or so. Mostly the problem with the bays is they are very hard to get alone, you basicaly need to buy second hand, that's what's expensive on them. Or buy drives with enclosures together from the server manufacturer; these drives do not even have special label stickers these day, they are common WD, Seagate, Hitachi drives with 60 % price premium.

[Agent24]

Re: SATA power connector: which moron invented it?

The one in my HP Proliant server doesn't use any intermediary connector between the drives and the backplane.

[tmiha71]

Re: SATA power connector: which moron invented it?

One of the main reasons why the sata power connector was introduced was missed by pretty much everyone here : low insertion force, required for "hotplugging"

I hope you were joking, with this statement...
Nobody with common sense (my thinking) can not guarantee "hotplugging" with mechanical connector (they toggle, from few to few tenth of times) only, as far as I know there's no mechanical connector/switch which does not toggle...

Low insertion force -> lower quality of mechanical contact...
Low insertion force and more wires -> troubles again...
Keeping same diameter of wires with low insertion -> troubles again...

Title of this topic is, sadly, VERY TRUE, SATA power connector is The Fail...

Never seen real implementation of hotplugging, but always presumed that there's some MOSFET switch which should be turned on - to allow power to go to hotplugged device (if that's true, than power connector MUST be on motherboard, or at some backplate - but again connected to motherboard)...

And that to function there should be capacitor near new load :-)))

[mariushm]

Re: SATA power connector: which moron invented it?

You work around the "toggling" by making sure the ground leads make contact first and by adding a "delay" on the hard drive which makes the drive only turn on a few ms after power detected, or when the system sends the start command through the sata channel or the special ground wire in the sata power.

If things were like how you said, we wouldn't have USB sticks today. USB connectors are much looser than sata connectors.

[tmiha71]

Re: SATA power connector: which moron invented it?

Pretty sure that "toggling" has accepted workarounds, I simply stated my opinion on "hotplugging" and low insertion force related to SATA power connector, to repeat myself -> designers missed their target...
---------
To USB.
USB connectors are easy to connect, USB equipement do not have wires, if it has, wire is more flexible than wire on SATA connectors (less strain on both connectors) ...

USB ports have current limiting, do not know if SATA power cable from PSU have current limiting ...

USB connectors are easy accessible, SATA power connector is placed on the left side of hard disk - it's harder to access it /cases have similar internal order/...

USB devices /usually/ don't have mechanical parts, hard disk rotates at 7200 rpm, or more, have moving heads, so any connector has to be tightly in position, and if it's designed with hotplugging in mind - easy to connect ...

USB equipement costs less than 10$ .
And to emphasize USB connector is better designed than SATA power connector, or simply better suits presumed usage ...

Anybody tried hotplugging SSD disc, or normal hard disc?

[Behemot]

Re: SATA power connector: which moron invented it?

Sure, you think I turn my PC off each time to check disk? I would must be mad to do that. On the other hand I have no problem hotplugging UATA drives either. They just require restart to get from system…I don't care and just remove them first in device manager and than physicaly

[mariushm]

Re: SATA power connector: which moron invented it?

Well, usb connectors are actually much worse in some respects and good at other things.


USB connectors are easy to connect,

yes, but they don't have key, you can put your hand behind the computer and shove the connector inside some jack and if it doesn't work you reverse it 90 degrees and try again.. it has lousy key-ing system.
Also... the standard usb connector is so nice it actually fits inside the network connector. That's dumb.

USB equipement do not have wires,

yes it has

if it has, wire is more flexible than wire on SATA connectors (less strain on both connectors) ...


it's only more flexible because of low current allowing you to use thinner wires. SATA is supposed to allow 2.5 amps per voltage rail, and power supplies usually chain 2-4 sata connectors on a strip and since they use awg18 for the other cables that can handle 5-8 amps safely, why bother using thinner cables?
There's nothing stopping psu manufacturers from using awg22-24 for... let's say 100-150w power supplies for miniITX cases, with just one or two sata connectors.


USB ports have current limiting, do not know if SATA power cable from PSU have current limiting ...

Not really, the hub usually controls 4 ports and gives 2 amps to all four ports, so each port could give 0.5 amps but nothing stops you from pulling up to 2a usually. SATA according to standard has to give 2.5 amps on each voltage but makes no sense to have a limit.

USB 3 allows up to 90w or maybe more, i forget now. Standard is 5v 0.9amps I think, but controllers can negociate higher currents or even switching to 12v or higher.


USB connectors are easy accessible, SATA power connector is placed on the left side of hard disk - it's harder to access it /cases have similar internal order/...

But usb is supposed to be for peripherals, sata is for mass storage. You're not supposed to insert or remove hard drives daily.

USB devices /usually/ don't have mechanical parts, hard disk rotates at 7200 rpm, or more, have moving heads, so any connector has to be tightly in position, and if it's designed with hotplugging in mind - easy to connect ...

yes, but hard drives are usually screwed to the case, so the vibrations and the fact that it's a motor doesn't affect the connector much.

And to emphasize USB connector is better designed than SATA power connector, or simply better suits presumed usage ...


So for the reasons above you can't really say it's better designed or that it's better suited for the usage.

One thing that for example is a problem is the number of insertion cycles.

The USB standard says that the old large usb connectors should support about 1500 insertions before the connector goes bad. However, for the miniUSB connector the original designs only allowed about 1000 insertion cycles but later designs improved the connector to handle about 2500-5000 cycles.

The microUSB finally added those latching pins that also press onto the tabs so this connector officially handles up to about 10000 insertion cycles.

With SATA, you have 5000 insertion cycles for the eSATA connector but the actual connectors on sata cables for data are only supposed to support about 50 insertion cycles. They won't fail after 50 insertions , but after 40-60 hard disk changes I would change the SATA cable or I may experience corrupted packets or retransmissions due to improper contact.
The actual physical connectors that are supposed to be soldered on pcbs can usually handle at least 500 insertion cycles, which is a lot considering you're not supposed to plug hard drives daily.
The most expensive connectors can handle up to 22000 insertion cycles

See for example this document for various sata connector examples with their mechanical data.

[Behemot]

Re: SATA power connector: which moron invented it?

USB 2.0 supports 100 mA when device can negotiate 5x100 mA or 0,5 A, that's the standard.

[tmiha71]

Re: SATA power connector: which moron invented it?

I am not sure, did you wrote that current limiting is bad /unwanted/ feature at USB (hotplugging in general) ?

[Behemot]

Re: SATA power connector: which moron invented it?

All those garage implementations counting on something the standard does not allow are. Than you have fried southbridge because of hard drive trying to suck >1 A from USB. It just does not start in the better scenario.

[mariushm]

Re: SATA power connector: which moron invented it?

The standard does say something like "a device should not pull more than 100mA when connected, then it has to negociate for more" but i don't think there's anything in the standard saying the hub/controller MUST limit or disconnect the device if more current is pulled.

The hubs do monitor the current consumption or at least report how much power was negociated by the device but again, not sure if they'll limit.

If you want, I can make a test and check, I have a usb dc-dc converter that pulls about 900ma from usb input to charge a supercapacitor. I could monitor how much current the individual connector will give.

[Behemot]

Re: SATA power connector: which moron invented it?

OK guys if you are interested in other premium stuff, I have 5 pcs of these SATA power females on the way as samples. If they are of sufficient quality, I will order 1000 pieces of them.

[eccerr0r]

Re: SATA power connector: which moron invented it?

What I'm wondering is how prevalent the female SATA power connectors are that have the drive activity light pinned and ported out... That would be interesting (and that I may need some in the near future to repair my hot swap bay power connectors).

[Behemot]

Re: SATA power connector: which moron invented it?

What is that?

[SteveNielsen]

Re: SATA power connector: which moron invented it?

I don't like SATA connectors either. The are weak and subject to frequent loosening due to vibration and therma creep. Horrible design.

[shovenose]

Re: SATA power connector: which moron invented it?

There is one computer at work we use for testing hard drive bays I don't even wanna know number of sata

[Sparkey55]

Re: SATA power connector: which moron invented it?

I don't like SATA connectors either. The are weak and subject to frequent loosening due to vibration and therma creep. Horrible design.

While ripping apart junk dvr's to salvage sata and pata hdd I liked the combo power and data connectors. All one piece, just solder new power wires to the short ones that went to the board.

[SteveNielsen]

Re: SATA power connector: which moron invented it?

More friction holding the combo connector on? SATA cable connector clips are useless, if they even exist, there's rearely anything to clip onto at the receptacle.

[Rulycat]

Re: SATA power connector: which moron invented it?

The problem I have with SATA power connectors is the nasty bend you get on them on good PSUs due to the thick inflexible wires. You can tell it is easy to snap it. I'd much rather they had a proper clip system. And as for the 3.3V line, that's pointless. Would be better to carry 12V on that line.