0
\$\begingroup\$

cable and switch http://ibb.co/dDctkk

http://ibb.co/dDOAWQ)

Quick preface, I have nothing but rudimentary EE education, only what I may call "common sense" sort of knowledge.

I have 3x serial (or parallel, actually) switches, used for printers at a now defunct call center, the entire centers' equipment which I inherited via Craigslist last summer.

Two switches are 25PDT switches in a big box and the third switch is a fancy PCB switch, shown in pictures.

I have an FTP server running in Windows 7 on one PC and Debian Jessie on a separate system, and I am getting tired of switching the USB plug when I want to switch systems, so I thought I could use the parallel switch with USB wired in to accomplish this task. All wires and all circuits have shown connections, as tested with a *-rigged circuit tester I made - - the Led lights up.

The problem is that there is no connection when I put the switch into the loop. Is anyone able and willing to help me troubleshoot?

That is about all I know to inform you with. If there is any further pertinent information, please allow me to update my question with said information.

Thank you if you have read this far, as I know I can be verbose to put it

enter image description hereenter image description here

\$\endgroup\$
14
  • 1
    \$\begingroup\$ Side note, this is what VNC and other remote desktop applications are for. \$\endgroup\$
    – Passerby
    Apr 30 '17 at 7:28
  • \$\begingroup\$ I use TeamViewer, but to keep it short, it doesn't work for what I intend. I had thought maybe it was the voltage in the circuit causing the issue, but your answer may just be the only approach. I shall try other methods then. Thank you. \$\endgroup\$
    – Ben
    Apr 30 '17 at 7:44
  • \$\begingroup\$ And I hope your lines aren't tied into the lines with leds on them. \$\endgroup\$
    – Passerby
    Apr 30 '17 at 7:48
  • \$\begingroup\$ There do exists comercial USB/KVM switches. I would suggest you buy one. \$\endgroup\$
    – winny
    Apr 30 '17 at 7:49
  • \$\begingroup\$ If I could afford even a cheap kvm, I would do so right now, however, I am stuck with what I have for the moment. Thanks anyway. \$\endgroup\$
    – Ben
    Apr 30 '17 at 7:50
2
\$\begingroup\$

USB is a system that uses what we call "differential pairs", because it is a very high speed system.

A differential pair is designed to be coupled together at a very specific coupling. Like in UTP cables, the signal wires are put into the cable at a very specific recipe, so that the impedance seen by signals of those data wires fits an exact specification.

Once that impedance changes more than 10% over more than a very small section problems occur with the signals. They start echoing and bouncing up and down. It'll be like trying to communicate a very delicate series of instructions in the most echo-y room you can imagine.

Adding a few loops of wires or such very likely creates a deviation of more than 50% at a local point. That point will be the hard brick wall in my analogy that is the surface those signals echo against/from.

If you don't know specific recipes for keeping USB signals perfectly coupled, or close enough to it, you'll create such perfect brick walls that nothing can use your system to properly communicate, because all they will hear is echoes.

\$\endgroup\$
10
  • \$\begingroup\$ The poster is talking about the lowest speed class of USB device however, which means a lot more tolerance for bad cabling than trying to access a flash drive or hard drive at USB 2 high speed rates or more would likely have. \$\endgroup\$ Apr 30 '17 at 7:27
  • 1
    \$\begingroup\$ if you cut the PCB traces ans solder the usb wires directly to the switches you have a chance that it might work, \$\endgroup\$
    – Jasen
    Apr 30 '17 at 7:40
  • \$\begingroup\$ I do also have the two 25pdt physical switches, but soldering the cables directly to that hasn't worked either. I must be doing something wrong. I suppose I should test each solder point individually. \$\endgroup\$
    – Ben
    Apr 30 '17 at 8:15
  • \$\begingroup\$ @ChrisStratton If you stand in a perfectly echoing chamber, you'll find it's extremely hard to properly talk at all. Sure USB 1.0 is a bit more forgiving, but it's not a world of difference. Reflection is still reflection and once the signal encounters close enough to perfect reflection it might as well be a a morse code, you'll not get it out properly anymore. Apart from that if devices identify as a certain protocol through their resistors, it'll soon default to minimum 1.1 or other higher standards and then all bets are off anyway. \$\endgroup\$
    – Asmyldof
    Apr 30 '17 at 8:44
  • 1
    \$\begingroup\$ @ChrisStratton And this KVM makes all devices suddenly USB 1.0, yes? That's what 1980's parallel port KVM switches are known for, I suppose. You are thanked for your pointless pedantry on things I have seen happen while inspecting USB2.0 busses for reviewing device behaviours, specially cheap mice and keyboards, but this answer is now nicely confusing to the beginners most likely to find it. \$\endgroup\$
    – Asmyldof
    May 1 '17 at 7:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.