Personal computing discussed

Moderators: renee, mac_h8r1, Nemesis

 
ShadowEyez
Gerbil XP
Topic Author
Posts: 348
Joined: Wed Dec 03, 2003 12:31 pm

Cableing question for all the Electrical Engineers out there

Wed Jan 02, 2013 1:42 am

Working on a machine, got me thinking - why can't computer companies design a "combo" cable that transports both data and power, ie; molex + SATA or USB in one cable with one connector. Is there a signaling issue, difficult to transport digital data/voltages along with electron power flow in one cable?

I'm not thinkging a long distance cable; think around 1-3 meters with high bandwidth, for internal components, not a cat5 networking replacement. Seems like it would be handy, one cable to any given device. Why not?
The finest tools are forged from the hottest fires
 
Captain Ned
Global Moderator
Posts: 28704
Joined: Wed Jan 16, 2002 7:00 pm
Location: Vermont, USA

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 1:45 am

How much power are you looking to transmit? Thunderbolt passes 10W.

High-speed serial comms lines like really thin conductors, mainly for skin effect purposes at the high transmission frequencies, whereas high power transmission needs thick conductors for ampacity purposes. Requirements are mutually incompatible.

As an example, the highest specified speed for Home Power Line Networking (using the house wiring instead of Cat6) tops out at 500 megabits per second. That's just nowhere near the speeds needed for SATA or PCI-Express links, and those speeds are only achievable in a house with completely new wiring installed by an anal-retentive contractor and crew. Average test speeds are more around 100-150 megabits/second and that's for read only.

Oh, and for the record, I'm no engineer. I'm just a geek.
What we have today is way too much pluribus and not enough unum.
 
Rübenschwein
Gerbil In Training
Posts: 7
Joined: Tue Dec 11, 2012 4:28 am

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 5:22 am

USB and Thunderbolt go into that direction. The largest issue with closely packing different, electrical (high frequency) signal-lines into one cable is interference, i.e., the signals would effect each other via their electromagnetic emanations and thus get corrupted. Such cables and especially the signal driving components would have to be carefully designed, read expensive, to be robust to whatever possible signals could occur on them. The more signal lines the more complex it becomes.

USB 2.0 for example basically uses a 50 MHz differential two-signal line and packs a 5 V and ground line into the cable. Differential signaling over a drilled line is a method already required to pass such high frequency signals. The power supply over that line is quite limited as more Watts would make it quite a bit harder, on the driving components to pass a readable signal.

Thunderbolt pushes the bandwidth a bit further than USB and they already plan to switch to optoelectronics hoping to avoid some of the EM interference for higher bandwidth.
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 10:56 am

@ShadowEyez - The first two replies seem to be assuming that you meamt power and data over the same pair of copper conductors. Is this really what you were asking, or were you thinking along the lines of a single cable with multiple separate wires for power and data inside of it, more like how USB works today?

Assuming you actually meant the latter, the main reason we don't do this for internal components is due to power distribution issues. For example, if we were to combine the SATA power and data lines into a single cable (but with separate power and ground wires within the cable), then the power for all of the hard drives would need to come through the motherboard instead of directly from the PSU. We'd need an additional +12V PSU power connector down by the SATA ports (like the ones we already have near the CPU socket and on video cards), and the SATA connectors would need to be much larger to accomodate the extra power and ground wires. This would take up a lot of space on the motherboard; the resulting cables would also be thicker and less flexible, making cable routing more difficult.

Rübenschwein wrote:
USB 2.0 for example basically uses a 50 MHz differential two-signal line and packs a 5 V and ground line into the cable. Differential signaling over a drilled line is a method already required to pass such high frequency signals.

USB 2.0 "Hi-Speed" mode switches a lot faster than 50 MHz. It's up in the 100s of MHz.

Rübenschwein wrote:
The power supply over that line is quite limited as more Watts would make it quite a bit harder, on the driving components to pass a readable signal.

That makes no sense; USB 2.0 uses separate pairs of conductors for power and data.
Nostalgia isn't what it used to be.
 
superjawes
Minister of Gerbil Affairs
Posts: 2475
Joined: Thu May 28, 2009 9:49 am

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 11:49 am

just brew it! wrote:
...
were you thinking along the lines of a single cable with multiple separate wires for power and data inside of it, more like how USB works today?

Assuming you actually meant the latter, the main reason we don't do this for internal components is due to power distribution issues. For example, if we were to combine the SATA power and data lines into a single cable (but with separate power and ground wires within the cable), then the power for all of the hard drives would need to come through the motherboard instead of directly from the PSU. We'd need an additional +12V PSU power connector down by the SATA ports (like the ones we already have near the CPU socket and on video cards), and the SATA connectors would need to be much larger to accomodate the extra power and ground wires. This would take up a lot of space on the motherboard; the resulting cables would also be thicker and less flexible, making cable routing more difficult.

There you go.

Your components don't really care where the power (or data) comes from as long as it is usable, but the path that your power (or data) requires some system-level thinking. You could develop a PCI protocol that passes enough power for a beefy video card, but you would still need to get that power onto or through the motherboard. Frankly, it makes a lot more sense to focus your motherboard on delivering reliable data between components and routing power to each individual component as needed.
On second thought, let's not go to TechReport. It's infested by crypto bull****.
 
UberGerbil
Grand Admiral Gerbil
Posts: 10368
Joined: Thu Jun 19, 2003 3:11 pm

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 12:29 pm

just brew it! wrote:
Assuming you actually meant the latter, the main reason we don't do this for internal components is due to power distribution issues. For example, if we were to combine the SATA power and data lines into a single cable (but with separate power and ground wires within the cable), then the power for all of the hard drives would need to come through the motherboard instead of directly from the PSU. We'd need an additional +12V PSU power connector down by the SATA ports (like the ones we already have near the CPU socket and on video cards), and the SATA connectors would need to be much larger to accomodate the extra power and ground wires. This would take up a lot of space on the motherboard; the resulting cables would also be thicker and less flexible, making cable routing more difficult.
And distributing all the power through the motherboard would probably require additional layers and even more careful design; it was already a big change when the PCI spec jumped to 150W. And for what net benefit? So that you'd have one less wire to plug in when you added a component? For most machines, which are assembled once and rarely -- or never -- have a component added, that would be a lot of extra engineering work for not much of a gain. And if you are in habit of adding and removing, say, hard drives frequently, there are docks that combine the two connections so you can essentially "plug in" the drive in a single convenient step.
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 12:56 pm

UberGerbil wrote:
And distributing all the power through the motherboard would probably require additional layers and even more careful design; it was already a big change when the PCI spec jumped to 150W.

That's why I said we'd probably need another PSU power connector down by the SATA ports. That way you wouldn't need to run a +12V power plane all the way down to that corner of the board from the CPU area (where most of the available +12V currently enters the motherboard). And even if you did tap into the same power plane that feeds the VCore regulators you'd probably need to double the size of the CPU power connector to handle the worst-case load from a full set of hard drives spinning up all at once.

No matter how you slice it, it's a bad idea.
Nostalgia isn't what it used to be.
 
RealWolf
Gerbil In Training
Posts: 3
Joined: Thu Dec 20, 2012 4:03 pm

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 2:07 pm

On the subject of routing power through the motherboard for drives, this is of course already being done in laptops. I've also run into some Compaq desktops that use a laptop style DC input directly into the board instead of a power supply, and then feed power to the drives off of the motherboard (blurry picture). You're not going to run a six drive RAID 5 off of that, but power for just a couple of drives does not seem to be an issue. I'm not sure what problems would be involved in changing the board design to move those power connectors down by the sata ports, but the power is already running all the way across the board.

I would tend to agree that this is not a good idea, but then again wouldn't be too surprised if some OEM actually did it.
 
NeelyCam
Gerbil First Class
Posts: 150
Joined: Fri Oct 16, 2009 12:25 pm

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 2:15 pm

ShadowEyez wrote:
Working on a machine, got me thinking - why can't computer companies design a "combo" cable that transports both data and power, ie; molex + SATA or USB in one cable with one connector. Is there a signaling issue, difficult to transport digital data/voltages along with electron power flow in one cable?

I'm not thinkging a long distance cable; think around 1-3 meters with high bandwidth, for internal components, not a cat5 networking replacement. Seems like it would be handy, one cable to any given device. Why not?



There are some SATA cables that have a single connector for the drive side, covering both power and signal connectors - i.e., signal and power wires are all running inside this "cable". In the other end, though, the cable is split into a SATA signal cable and a SATA power cable.. Is this what you were thinking..?

The signal wires are differential and shielded; noise from power wires is not an issue (the noise is at low enough frequency, too, that it won't be a problem)
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 2:30 pm

RealWolf wrote:
On the subject of routing power through the motherboard for drives, this is of course already being done in laptops.

Laptop drives use a small fraction of the power used by desktop/server drives.

RealWolf wrote:
I've also run into some Compaq desktops that use a laptop style DC input directly into the board instead of a power supply, and then feed power to the drives off of the motherboard (blurry picture). You're not going to run a six drive RAID 5 off of that, but power for just a couple of drives does not seem to be an issue. I'm not sure what problems would be involved in changing the board design to move those power connectors down by the sata ports, but the power is already running all the way across the board.

As you note, you're not going to run a ton of drives off of that. Most motherboards have 6 SATA ports these days; it would be rather odd to have 6 ports, but then tell the user they're not allowed to use all of them!

RealWolf wrote:
I would tend to agree that this is not a good idea, but then again wouldn't be too surprised if some OEM actually did it.

Yeah, I wouldn't be surprised either. But only for systems that are designed to run just 1-2 hard drives.
Nostalgia isn't what it used to be.
 
Captain Ned
Global Moderator
Posts: 28704
Joined: Wed Jan 16, 2002 7:00 pm
Location: Vermont, USA

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 5:51 pm

just brew it! wrote:
it would be rather odd to have 6 ports, but then tell the user they're not allowed to use all of them!

How soon you forget the past.
What we have today is way too much pluribus and not enough unum.
 
ShadowEyez
Gerbil XP
Topic Author
Posts: 348
Joined: Wed Dec 03, 2003 12:31 pm

Re: Cableing question for all the Electrical Engineers out t

Wed Jan 02, 2013 9:11 pm

Excellent answers as usual:

just brew it! wrote:
@ShadowEyez - The first two replies seem to be assuming that you meamt power and data over the same pair of copper conductors. Is this really what you were asking, or were you thinking along the lines of a single cable with multiple separate wires for power and data inside of it, more like how USB works today?


@just brew it! - You are correct, I was thinking in terms a single cable with multiple wires inside. Think a standard "data + power" cable with one connector; maybe the next-gen usb?

Thunderbolt is a step in the right direction, but it's expensive, and do we really need an "intelligent cable"? Why not just integrate the controller logic on the board and in the device? Also, 10w is great for rubber-duck/light up usb toys, but for things like monitors, printers, external hard drives, and speakers (not to mention external graphics cards...) you'd want a little more wattage, right? And like many things, it seems to be a tradeoff of higher capacity power transmission vs. higher capacity data transmission vs. flexibility vs. cost...

But if we did have one such standard, royalty free and reliable, wouldn't that be nice?
The finest tools are forged from the hottest fires
 
ludi
Lord High Gerbil
Posts: 8646
Joined: Fri Jun 21, 2002 10:47 pm
Location: Sunny Colorado front range

Re: Cableing question for all the Electrical Engineers out t

Thu Jan 03, 2013 2:14 pm

For the power side, there are multiple tradeoffs at stake, and most of them come out of Ohm's law: (V) = (I) * (R), where

(V) = potential, in volts
(I) = current, in amps
(R) = resistance, in ohms

...and its sibling power equation, (P) = (V) * (I), or by algebraic substitution, (P) = (V^2) / (R) = (I^2) * (R), where

(P) = power, in watts.

Presently, the power consumed within a PC is driven by potentials of 12V and under. To get more power, either (V) or (I) must be increased. If we increase (V), we quickly run into safety and durability issues because DC makes for some really prolonged, fat arcs when it is interrupted or finds a short circuit path, and higher voltages aggravate the problem. So, we have generally kept (V) low, and increased (I) when necessary.

However, since the conductor carrying (I) has its own value of (R), and (P) losses in the conductor are proportionate to (I^2), we have to make the cable length (L) shorter and/or make the conductor thicker, because losses in the cable cause heating and a (V) drop across its length. Likewise, the magnetic field (B) generated around the conductor is proportionate to the magnitude and rate of change in (I), and any interference effects transmitted to other conductors are proportionate to the strength and rate of change in (B) and the parallel (L) of the conductors. Although DC should theoretically have a continuous current flow, switching devices actually deliver and draw power in dozens of high-frequency pulses. These pulses are buffered on both ends, but a substantial ripple current cannot be avoided, so (B) is changing to some degree, and this can be transmitted to data-path conductors, modifying their signal. The catch is that (B) decreases exponentially with distance, so two conductors in the same cable jacket may have significant interference problems while two conductors lying a couple centimeters from each other might be perfectly fine.

A different-but-related set of problems is being encountered on the data conductors at the same time, as other posters have already detailed. And there's also the inevitable question of what happens if a power and a data pin short to each other -- can both devices survive the initial fault and shut down quickly enough to avoid damage?

The tl;dr is that as power capability increases on the power conductors and data bandwidth increases on the signal conductors, there are multiple tradeoffs being played against each other. Sometimes it makes sense to combine both functions in one cable and sometimes it does not.
Abacus Model 2.5 | Quad-Row FX with 256 Cherry Red Slider Beads | Applewood Frame | Water Cooling by Brita Filtration
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Cableing question for all the Electrical Engineers out t

Thu Jan 03, 2013 3:51 pm

ludi wrote:
The tl;dr is that as power capability increases on the power conductors and data bandwidth increases on the signal conductors, there are multiple tradeoffs being played against each other. Sometimes it makes sense to combine both functions in one cable and sometimes it does not.

...and in general, it comes down to "If the device is external, combine them if the amount of power is sufficiently low to be carried over thin and flexible wires; if the device is internal, don't bother."
Nostalgia isn't what it used to be.

Who is online

Users browsing this forum: No registered users and 1 guest
GZIP: On