Blown Motherboard from ATA-100 Cables? 30
Dragan Lazin asks: "I recently bought a couple of rounded ATA-100 cables from an online store; very ingenious actually and they have a nice color: blue ;-) Problem is, when I installed the cables, 16 capacitors on my motherboard blew - right between the CPU and the parallel port header. This is an Abit KA7-100 mobo. What the hell causes this kind of damage? I'm trying to get a refund and a new mobo from the company. Did anyone ever experience this?"
A short? (Score:2, Interesting)
Re:A short? (Score:2, Informative)
who did you get the cable from? (Score:1)
What kind are they? (Score:1)
Re:What kind are they? (Score:4, Interesting)
The reason ide cables have traditionally been ribbon-shaped is to minimize cross-talk. Perhaps the round cables use some sort of pair-twisting scheme, or maybe they use shielding. Or perhaps they just decided that cross-talk wasn't really as much of a problem as the engineers originally thought.
Re:What kind are they? (Score:2, Informative)
Re:What kind are they? (Score:1)
(An explanation: The clock line carries a square wave at a fixed frequency - 33/66/100MHz for IDE - and is used to keep everything in sync. Devices at both ends of the cable look for a rising edge (when the voltage on the clock line goes from 0V to 3.3V) on the square wave, and read data off the cable then. Crosstalk tends to result in 'glitches' on adjacent lines, e.g. a rising edge on one conductor can induce quite a high 'spike' on adjacent conductors. If you get a positive glitch on a clock line, sometimes it is treated in the same way as a rising edge, and the devices get confused).
How this applies to IDE cables:
"Standard" (old-style) IDE cables have 40 conductors, without many grounded lines. Presumably they have something to protect the clock line though...
"New" (for ATA-66 or ATA-100) IDE cables have 80 conductors, and every second one is connected to ground (0V). These acts as a 'shield', absorbing the aforementioned glitches (the glitches happen on the grounded lines, which don't carry any data, rather than on the signal and clock lines).
If you twist up / fold a 40-conductor cable, you'll probably end up with crosstalk problems. Your mileage should be *much* better with an 80-conductor cable, because of all the grounded lines. Alternatively they may be twisting a grounded line around each signal line, which should be even better (Ethernet does that BTW -- that's the 'T' in UTP and STP).
One of the hardware guides did a 'do-it-yourself rounded cable' tutorial a while back - not sure which one though. I think they used 80-conductor cables, with a bit of masking tape
Bad Cable, Bad Board? (Score:2, Informative)
KidA
Re:Bad Cable, Bad Board? (Score:3, Informative)
How do we blow up caps? 1) power up a tantulum cap backwards. Not possible with a bad IDE cable for the reason above... no caps connected to the lines. 2) apply a huge voltage to a low voltage capacitor. Likely the caps on a motherboard are around 6.3V caps, so this is possible. But again, there are no caps on the IDE lines. Not gonna happen.
Hence, blowing caps off of the motherboard (assuming that it actually did happen...) was not caused by the cables. The only was to possibly do this was to plug in your cables from the powersupply to the motherboard wrong. In this case, it is possible to get GND and +V backwards in this case.
Don't blame the cables, blame user stupidity.
Actually, I doubt that this is even a real question. Sounds more like a troll that got posted by Cliff.
Re:Bad Cable, Bad Board? (Score:3, Informative)
Re:Bad Cable, Bad Board? (Score:1)
Last I saw, most of the caps around a processor are electrolytics (tall purple cans). They are polarized and you can blow them up if connected backwards. Enough current on the ground plane could do this if the board is poorly designed, but it doesn't seem to make sense in the end, although I have seen stranger things happen.
KidA
PS Yeah could be just a troll..
How would a cable make capacitors blow up? (Score:4, Informative)
Capacitors blow up from too much heat, which for a DC power supply filtering capacitor implies too much voltage. Capacitors are in parallel with the power supply. Something put too high a voltage across the capacitors.
Unless I'm really confused, the highest voltage on the IDE connector is only 5 volts, and all of the pins on the IDE connector are either ground or are compatible with 5 volts. You can hurt the logic chips and the power supply by shorting stuff on the IDE connector, but you won't blow up the capacitors.
The Cables caused the Problem?!? (Score:1)
Personally it sounds more like a power surge to me than a problem with the cables...
Re:The Cables caused the Problem?!? (Score:1)
Re:The Cables caused the Problem?!? (Score:2)
[1] The pin/socket style used on most HD power connectors isn't that good. It tends to get weaker with every use and fails after only a few hundred cycles. It will also fail faster if it isn't aligned right and you force it. This could cause the ground and +5VDC lines to not connect up properly. Thus the drive would be biased at +12V.
[2] Removing, knocking off the voltage control jumpers or putting them into an undefined state may make the switching power supply that feeds the caps to go over voltage. This could also destroy the cpas.
Re:The Cables caused the Problem?!? (Score:1)
6 or 8 maybe. Maybe. Personally I don't think 16 caps blew. Not without a fire that would make it impossible to tell caps from resisters from diodes, etc.
Rounded cables probably not the problem (Score:1)
Not sure if this would help... (Score:2)
One thing to consider, though, is whether the parallel port controller chip also is the IDE controller as well - sometimes motherboard manufacturers use these "super"-chips as a means to cut costs. I am involved in a group "hacking" the Acer NT-150 set-top box, and the controller chip for the parallel port on it also has some IDE controller functions, as well as floppy drive functions. Basically, they made it so you could build a "funky" parallel cable, hook it up to a floppy drive, plug the other end into the parallel port, and that becomes the floppy port - otherwise it is for a printer.
There may be a connection in your case - who knows?
Re:Not sure if this would help... (Score:1)
This sort of thing just happened to me (Score:2)
Not certain on either one, but it made me buy a new board and processor (and ram... DDR is god) and now I dont use my printer any more because I really dont want to fry this one.
More or less, the things that others have said about the rounded cables not being the problem are probably right. Not nearly enough voltage to blow all those caps.
Oh well. I havent tested out my processor to see if it is still okay yet. I want to use it for a server so I hope it is.
Re:This sort of thing just happened to me (Score:2)
It went in OK, but would not switch to VR mode (it has goggles). I tweaked the drivers and rebooted, it wouldnt boot. Safe mode and VGA mode died soon after, and then it wouldnt boot from Win2k CD. So I figure bad card, replace it. When I open the case I see one of the caps near the DIMM slots was on an angle, I figured I kicked it plugging something in. But when I pull the card (w/ everything unplugged, myself grounded), I hear a short hiss. Examination with a flashlight shows 3 caps near the CPU have popped (off base at angle with burnt stuff below, curved top), and the one I saw earlier has completely blown off its base and theres burnt black rubber stuff around it and the one next to it.
I was looking for an excuse to get a KG7, and this works, although it wasn't what I had in mind
If anyone cares, the blown ones are:
By CPU, 1500uF, rated at 6.3v, 105C: 3 blew, cant see markings. From the top, 2nd, 3rd and 6th in the row next to the SlotA
And by DIMMs, a 2200uF 6.3V 105C, right next to DIMM1, top in a group of two.
All are JPCON
Who did Abit buy the capacitors from? (Score:2)
Seeing as how we've got at least 3 Abit owners posting this problem so far, and not all are using ATA-100 round cables, I'd suspect that Abit got a bad run of caps from their subcontractor, although some sort of voltage regulator problem that fed those caps too much voltage or let a reverse polarity spike get through to them could also be at fault. Are all of y'all using the same brand power supply?
Gnnn. Can't... resist... suggesting... (Score:2)
...that this is an "undocumented feature" of WinXP hardware product verification. >>> ;-) <<<
Power Supply (Score:1)
Had this happen to me (Score:1)
New M.B.?? (Score:1)