Image Map Image Map
Results 1 to 10 of 10

Thread: Cheap y-splitters, chip fans, and long power runs oh my

  1. #1
    Join Date
    Jan 2013
    Location
    Marietta, GA
    Posts
    2,853

    Default Cheap y-splitters, chip fans, and long power runs oh my

    Is there some law that says newer y-splitters are not allowed to make good contact between connectors? And cheap connectors on the power supplies are to blame too.

    On top of that, why can't they make chip or case fans that last more than a couple of years? When each one costs ~$10 at Microcenter you would think they would last a while.

    I recently wound up buying a new power supply for one of my machines because it was only a few bucks more than buying just a replacement power supply fan. What a load!

    Speaking of load... why is it that when I have to run a power consuming device through one of these cheap splitters, even if it seems to make contact, the device invariably behaves intermittently as if it is not getting enough power?

    Has anyone else ever run in to a problem with hard drives developing "bad sectors" because other devices such as chip fans have to run off the same power supply cable? (not even the same connector with a y-splitter, that is just asking for trouble!) The odd thing is these "Bad sectors" may even disappear from the smart status once the power problem is sorted out.

    You would think a newer (allegedly) 600 watt power supply would be able to put out enough power to run some relatively simple hardware, but somehow it never is.

    I've been up to my neck it problems with a particular system, and always seems to stem back to something somehow not getting a good power connection, or not getting enough power somehow. There are a lot of splitters for case fans, and most of the devices use "molex" plugs instead of SATA type power plugs. It is just insane, if I connect things one way, then one thing acts like it is not getting enough power and if I wire it another way something else acts that way. WTF.

    Also, do newer power supplies still have 4-pin "molex" connectors? It seems like the last time I went looking for one I had hard time finding one. Also had a hard time finding one that wasn't lit up like some consumertard's xmas tree.

    I have a bad feeling I may have to break out the soldering iron and cut off some 4-pin molex plugs from some old power supplies made back when connectors were real connectors. (And they were usually white instead of pitch black so I could actually SEE them when I was working on them!)

    Although this is a bit of a rant, any serious suggestions to address these are more than welcome.

  2. #2
    Join Date
    Jan 2007
    Location
    Pacific Northwest, USA
    Posts
    29,180
    Blog Entries
    20

    Default

    Why scavenge old Molex when you can buy them new? Wobbly pins in Y-splitters have been a curse for years.

  3. #3
    Join Date
    Sep 2003
    Location
    Ohio/USA
    Posts
    7,159
    Blog Entries
    2

    Default

    Case fans last a while if the lube doesn't leak/evaporate/get dirty, and they are easy enough to fix if they do (before the bearing goes anyway).

    Brand new PS still have 4 pin model power connectors (lots of them) and even the floppy drive power plug.

    600W supplies run some decent heavy loads. Of course most of that power is on the 12V side so if you connect them to an older system that is 5V heavy you will have problems (look at the label they tell you what the max load for each voltage is).

    CPU fans run off their own connector on the motherboard, not sure why that would interfere with a HD that gets its power directly from the PS.

    If a PS fan costs the same as a new PS you are buying crap PS's or somebody is over charging you on the replacement fan. Those fans either just need lubed or the pin holding the shaft in place pops off and the fan blades fall off (easy enough fix). Try using a decent case that has a filter screen on the PS air intake (bottom mounted PS that sucks cold air from outside the case).
    What I collect: 68K/Early PPC Mac, DOS/Win 3.1 era machines, Amiga/ST, C64/128
    Nubus/ISA/VLB/MCA/EISA cards of all types
    Boxed apps and games for the above systems
    Analog video capture cards/software and complete systems

  4. #4
    Join Date
    Jan 2013
    Location
    Marietta, GA
    Posts
    2,853

    Default

    Well, I've wasted this entire holiday weekend trying to fix some of these problems. I put a spare unopened unused hard drive in the system and the stupid thing already claims to have developed new bad sectors! It turns out there are several variety of SATA power to 4-pin molex connectors that I had never noticed before - but I think it was trying one of those that pissed off the hard drive. I've tried everything including a different power supply, with no improvement.

    The worst part is I tried a spare motherboard that I used to use %100 reliably in my old development system and it turns out that at some point it developed a fault. No, not the kind of fault where it doesn't turn on, or gives a failure message at boot, or lets out the magic smoke... oh, no, it loooookd like it worked perfectly until I started observing some intermittent file corruption where every here and there a bit would get flipped. Of course I had to go through EVERYTHING to narrow it down - no, surprisingly it's not the new hard drive, not the power supply, oddly not the cables, not the port the drive was connected to, not any of the other devices, leaving MEMTEST running for HOURS said everything was freaking perfect, swapped RAM anyway, switched around the CPUs because I had formerly had a problem with a flaky CPU, and all that was left was the motherboard. Finally put the old one back, so now I only have the problems I was already familiar with! :P.

    Everything... I... ever... do... is... a... FAILURE!

    At least through all this I MIGHT (maybe... possibly... probably not) have found a power plug arrangement that makes the components more or less happy, and I now know of some specific combinations (anything extra on the same cable from the power supply as the hard drive, and anything at all that uses the third SATA only cable from the power supply that I had never used before) that will cause things crap out. Also managed to eliminate several of the y-splitters I was using.

    Oh, and one of the chip fans I didn't replace is making noise now!

  5. #5

    Default

    Quote Originally Posted by SomeGuy View Post
    The worst part is I tried a spare motherboard that I used to use %100 reliably in my old development system and it turns out that at some point it developed a fault. No, not the kind of fault where it doesn't turn on, or gives a failure message at boot, or lets out the magic smoke... oh, no, it loooookd like it worked perfectly until I started observing some intermittent file corruption where every here and there a bit would get flipped. Of course I had to go through EVERYTHING to narrow it down - no, surprisingly it's not the new hard drive, not the power supply, oddly not the cables, not the port the drive was connected to, not any of the other devices, leaving MEMTEST running for HOURS said everything was freaking perfect, swapped RAM anyway, switched around the CPUs because I had formerly had a problem with a flaky CPU, and all that was left was the motherboard. Finally put the old one back, so now I only have the problems I was already familiar with! :P.
    Yeah, this is a sign of bad caps on the motherboard. I had the same problem with data corruption and the only clue was that the drive reported CRC errors in one of the SMART attributes (I don't recall the exact name of the attribute). So that might be something to watch out for with aging motherboards I guess.
    Looking for a cache card for the "ICL ErgoPRO C4/66d V"

  6. #6
    Join Date
    Jan 2013
    Location
    Marietta, GA
    Posts
    2,853

    Default

    On top of this, why is it that the fake bad sectors ONLY appear in the middle of the Windows registry? I have seen this EXACT same behavior on perhaps a dozen Western Digital drives ranging at last 320GB to 1TB and on various versions of Windows. Even after I go to the trouble to make sure the registry gets moved elsewhere on the disk.

    I'm honestly beginning to think there is something inside the hard drive's embedded CPU that is spying and intentionally crapping itself.

  7. #7
    Join Date
    Jan 2013
    Location
    Marietta, GA
    Posts
    2,853

    Default

    I needed to replace a large 140mm fan... the only options at Microcenter were a single, likely whining-sounding, Kingwin brand model or dozens of consumertardastic fans lit up with blazing-blue, pink, or rainbow LEDs. Grabbed a sane unlit one off of eBay. What a world.

  8. #8
    Join Date
    Oct 2008
    Location
    Kamloops, BC, Canada
    Posts
    5,427
    Blog Entries
    41

    Default

    I have a DVR that thankfully uses an external DC power brick but ships with a 4A supply when it REALLY needs a 5A supply and externally powered cameras
    The excessive draw on the power brick caused the obvious overheating of the brick and intermittant DC drops that would cause the hard drive to corrupt and recover itself automagically by overwriting the corrupt segments.
    Of course, the high current draw was also associates to power hungry major IC's that shipped without heatsinks so the system only really worked reliably in the Winter anyways until I installed heatsinks on about a dozen IC's and installed two small fans at the bottom of the DVR and let convection and undervolted fans to take care of the rest.
    = Excellent space heater

  9. #9
    Join Date
    Jan 2013
    Location
    Marietta, GA
    Posts
    2,853

    Default

    Is there any good, and not too expensive, way to measure the stability of +5 and +12v power?

    Just this morning my new hard drive developed yet another "bad" sector again (right in the middle of the registry as always) for no explicable reason other than it hates me. I don't even see how that could be caused by anything outside of the hard drive itself. I mean, a motherboard or CPU crashing and halting an I/O operation or sending gibberish should not create a bad sector on IDE/SATA style drives.

    I think I managed to overwrite the bad sector, so I'll have to check and see if the smart status dropped back down to normal again.

  10. #10
    Join Date
    Jan 2013
    Location
    Marietta, GA
    Posts
    2,853

    Default

    Yep, it went back to reporting perfect health.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •