2
\$\begingroup\$

I have a 16m x 0.8m grid of neopixel-based leds, ziptied to a metal grid of the same size (matrix of 8x120 LEDs).

The LEDs have been waterproofed by inserting them into clear heatshrink, and a glue gun used to seal ends. This means that the LEDs are completely insulated from the metal grid.

Unfortunately a metal connection on the grid has snapped during transport, and we have not yet inspected the rest of the grid, but more may have snapped. To fix the metal grid we will be using an arc welder.

Further misfortune is that the process of attaching the LEDs to the grid is very long, so removing the LEDs from the grid is not feasible.

Question finally is, would it be safe to weld the grid back together with the LEDs still attached? We can of course clear about a half meter radius around the welding spot, to ensure heat isn't an issue. The bigger concern is whether the high voltage of the welding will cause havoc with the LEDs?

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Consider silver brazing instead - works at 650C (medium-bright red) with a plumber's butane/propane torch if the mass of metal isn't too high to heat effectively. (Need : Plumbers torch, Borax flux, silver brazing rod) \$\endgroup\$
    – user16324
    Commented Nov 29, 2017 at 17:19
  • \$\begingroup\$ It depends on welder type, grounding of weld, proximity of target LEDs and if open circuit and absorption of stray EMI into a grounded workbench. \$\endgroup\$
    – D.A.S.
    Commented Nov 29, 2017 at 20:11

1 Answer 1

Reset to default
2
\$\begingroup\$

Arc welding uses high current not high voltages, however those high currents, 100s of amps, can induce high voltages in the wires of your LED array. Unfortunately, even if you clamp a work lead on either side of the weld spot, a considerable amount of current will still flow through the metal frame, which will induce currents in the LED wiring.

As such, I think you would be better not arc welding the frame back together. You would be better to use oxyacetylene welding, or any other non-electrical method instead.

\$\endgroup\$
7
  • \$\begingroup\$ And I assume connecting ground to the grid wouldn't help? Probably make things worse... \$\endgroup\$
    – Makoto
    Commented Nov 29, 2017 at 17:27
  • \$\begingroup\$ @Makoto it would depend on how the welder is configured. Either way, it does not sound like you should be taking a chance it might be ok. Especially not an any advice from some guy on the internet ;) \$\endgroup\$
    – Trevor_G
    Commented Nov 29, 2017 at 17:30
  • 1
    \$\begingroup\$ But Trevor, you're not just some guy :p but true, don't think I'll be taking that chance \$\endgroup\$
    – Makoto
    Commented Nov 29, 2017 at 17:33
  • \$\begingroup\$ @Makoto :) thanks. After all.. you only get one shot at this.. why risk it when there are other easy to obtain methods. \$\endgroup\$
    – Trevor_G
    Commented Nov 29, 2017 at 17:35
  • 1
    \$\begingroup\$ I might, but others not. I once had to measure a shunt current in the 1~100 mV range under 10kA arcs from a diffusion arc welder which ran about 5V under load or 50kW of continuous arc. It worked but the long coax had to be 90 deg to loop. also ... iaea.org/inis/collection/NCLCollectionStore/_Public/30/032/… \$\endgroup\$
    – D.A.S.
    Commented Nov 29, 2017 at 20:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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