I've been using coax cables with BNC connectors for over 40 years. I use them a lot for power connections to DC power supplies (convenient with a BNC female to dual banana plug adapter), but I wondered what I should limit the current to. Since I couldn't find any usable information on the web (it's probably out there somewhere; I just didn't find it), I made some measurements last weekend. Results in attached file (also at http://code.google.com/p/hobbyutil/).

Bottom line: I'll run my cables all day at 5 A, occasionally go up to 7.5 A, and once in a while when needed I might go to 10 A.

 

British UR cable designations have the following central core sections

UR \(\ \)\(\ \)\(\ \) Inner (in)

43\(\ \)\(\ \)\(\ \)0.032
57\(\ \)\(\ \)\(\ \)0.044
63\(\ \)\(\ \)\(\ \)0.175
67\(\ \)\(\ \)\(\ \)7/0.029
74\(\ \)\(\ \)\(\ \)0.188
76\(\ \)\(\ \)\(\ \)19/0.0066
77\(\ \)\(\ \)\(\ \)0.104
79\(\ \)\(\ \)\(\ \)0.265
83\(\ \)\(\ \)\(\ \)0.168
85\(\ \)\(\ \)\(\ \)0.109
90\(\ \)\(\ \)\(\ \)0.022


Sorry don't have the american RG series equivalents to hand.

 

A useful rule of thumb is 700 circular mils per ampere.

 

http://www.truepowerresearch.com/2011/02/coaxial-power-cable/

 

http://www.truepowerresearch.com/2011/02/coaxial-power-cable/

I looked at extrapolating from both NEC-specified current densities as well as some data I have from MIL-STD-975 (which is based on wire with 200 deg C insulation in a hard vacuum). But note MIL-STD-975 doesn't include convective cooling, so using it for wire in air will likely be conservative.

Fundamentally, the problem is a heat transfer problem, so things like orientation, coax cable insulator type, room temperature and airflow all come into the equation.

To avoid going to the library, I just did a relatively easy experiment and now I know what works for my cables.

 

It would be nice to know how your experimental data corresponds to the rule of thumb or other information. Nobody goes to the library anymore -- all the stuff is online now.

 

It would be nice to know how your experimental data corresponds to the rule of thumb or other information. Nobody goes to the library anymore -- all the stuff is online now.

Agreed; the public libraries suck for technical info, so I'd have to go to the local university library for a couple of technical books. But it's across town and I doubt I've been there in 15 or 20 years...

If the center conductor of RG-58 is about 19 AWG (0.9 mm), then the current density is 7.86 A/mm2 for 5 A of DC current. A few years ago I did a regression to NEC data (see attached plot; the data are for different insulation temperature ratings). Using that linear regression (the logs were base 10 logs), the regression line predicts about 12.5 A/mm2 allowed. I wouldn't used it at the extrapolated current density, especially because the insulation is thicker than the insulation of typical house wire, meaning increased thermal resistance and higher core temperatures than implied by NEC.

Papabravo, your 700 circmil/A is equivalent to a current density of 2.8 A/mm2. I've seen that 700 number quoted as a conservative design number (I think it was for transformer windings); the equivalent for 7.86 A/mm2 is around 250 circmil/A.

 

Agreed; the public libraries suck for technical info, so I'd have to go to the local university library for a couple of technical books. But it's across town and I doubt I've been there in 15 or 20 years...

If the center conductor of RG-58 is about 19 AWG (0.9 mm), then the current density is 7.86 A/mm2 for 5 A of DC current. A few years ago I did a regression to NEC data (see attached plot; the data are for different insulation temperature ratings). Using that linear regression (the logs were base 10 logs), the regression line predicts about 12.5 A/mm2 allowed. I wouldn't used it at the extrapolated current density, especially because the insulation is thicker than the insulation of typical house wire, meaning increased thermal resistance and higher core temperatures than implied by NEC.

Papabravo, your 700 circmil/A is equivalent to a current density of 2.8 A/mm2. I've seen that 700 number quoted as a conservative design number (I think it was for transformer windings); the equivalent for 7.86 A/mm2 is around 250 circmil/A.

Interesting. I thought the number was for wire bundles in a cable tray. I do quite a bit of industrial controls and wiring inside cabinets.