2
\$\begingroup\$

I am trying to put together the requirements for a cable and I was wondering what would be the most suitable wire size to load the DUT with 25Amps @ 9VDC. I will have few other high current lines 13Amps @ 5VDC. There will also be 2 DC lines 160V @5Amps each.

I was wondering if all of this can be bundled (or what would be the key things to keep in mind) with some other wires carrying RF signals without causing any signal integrity issues.

I plan to use shielded Coax cables (at least 20) and bundle all of it along with some other signal lines (like DIO, SPI, CAN bus, etc) in one cable. The RF signals are not going to have a frequency higher than 15MHz but I will need to sample it quite fast to detect a time difference between 2 signals in 5 nanosecond range.

Would anyone suggest:

  • The right wire size for this application
  • Would there be any issue in regards to RF measurements?

The overall cable length is going to be about 15ft.

\$\endgroup\$
6
  • 3
    \$\begingroup\$ How much voltage drop can you tolerate? 9V @ 25A over 15 feet is a bit extreme. Nearly identical issue with 5VDC @ 13A. Half the current so half the voltage drop, but half the voltage so also half the tolerable voltage drop. \$\endgroup\$
    – DKNguyen
    Jul 23 '19 at 20:41
  • 1
    \$\begingroup\$ You really should consider point-of-load regulators. Do all of the long-distance runs at the highest voltage available (160V) and then use a DC-DC converter at the destination to produce each of the low-voltage supplies. What on earth is this for, anyway? \$\endgroup\$
    – Dave Tweed
    Jul 23 '19 at 20:48
  • 1
    \$\begingroup\$ Bundling the 160V DC lines is a bad idea - all the low voltage signals in this cable would have to be treated as if these carried lethal voltages, for safety reasons. You usually don't want that. Oh, and there is also cross talk... \$\endgroup\$
    – Turbo J
    Jul 23 '19 at 20:55
  • 1
    \$\begingroup\$ @DKNguyen, I plan to connect the sense lines of the Power supplies if the Voltage drop is going to be an issue. \$\endgroup\$
    – Rav
    Jul 23 '19 at 21:00
  • 1
    \$\begingroup\$ @TurboJ, thanks for the feedback, I planned to twist wires (for example, sense lines, any differential pair line etc) to eliminate cross-talk. I am looking for a resource on cable design...in case anyone knows a good reference. \$\endgroup\$
    – Rav
    Jul 23 '19 at 21:04
3
\$\begingroup\$

You have three things to worry about: heat, voltage drop, and interference.

Interference between your DC and RF wires is probably the least of your problems, unless your high-current lines dump noise onto the RF lines, or the RF lines are really powerful, and dump enough energy into the DC lines to cause problems for whatever you're powering.

For the heat problem, you can probably size the wires as if they were in a car. Find a table of recommended wire sizes, and maybe do a test to see if the wires get hot.

But -- voltage drop is going to be your enemy. Just nine volts implies that you're going to need to restrict the voltage drop to 1/4 or 1/2V (or less -- you know what you can stand). The general way to deal with this is to calculate the maximum resistance that you can stand (i.e., 0.25V / 25A = 0.01\$\Omega\$), look up the resistance per meter (or foot, or whatever) of various sizes of copper wire, then choose a wire diameter that will result in the entire length of your wire run (there and back) to have less resistance than your specified amount.

I suspect that once you get wire that fat, you'll be well above the size you need to be safe from heating.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you for the prompt response. Very helpful. \$\endgroup\$
    – Rav
    Jul 23 '19 at 21:10
1
\$\begingroup\$

Let's assume you can tolerate a 5% drop in voltage on your nominally 9V supply.1 0.45 V / 25 A = 18 mΩ total resistance (wire + connectors). Ignoring connectors for the moment, and with 30 feet of wire (supply + return), that's just 0.6 mΩ per foot, or 600 mΩ per 1000 feet. AWG 8 copper wire is 612 mΩ per 1000 feet, just barely good enough.

You need to go through this calculation for each supply. 5V @ 13 A (5% drop) will also require AWG 8.


1 If your power supplies support remote sensing, then you can tolerate much more drop in the wires, limited only by the "headroom" of the supplies themselves.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks, very helpful. I will connect the sense lines of all available power sources and twist the sense lines in the cable. \$\endgroup\$
    – Rav
    Jul 23 '19 at 21:10

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.