If memory serves: a 0.2" wide track of 2oz copper will take 10A for
something like a 50C rise. I'd have to look up the reference to be sure.
That's dc current of course. If you want to bring the tracks closer to make
them wider, you can apply an insulating compound. I'm not in that business
any more, so, I can't tell you what compound, but silicone glue might do it
(the type without the vinegar smell). Can't remember if this is called
supplementary insulation or reinforced insulation for UL purposes. You
still have to take care of the pcb on the other side, ie to stop debris
settling between the relay pins, so, you could mount them a little off the
board to ensure your board washing clears it all away from the underside of
the relay.
David
G3UNA
> Jim, W8ZR recently mentioned his preference for the RTB140XX 12A relay
> over the RTD140XX 16A relay for switching legal limit RF. His
> preference was because the RTB relay only uses 1 pin per contact
> connection and staggers them resulting in a slightly larger pin spacing
> than the RTD relay which uses 2 pins per contact connection. The
> assertion was based not on the relay itself but on the PCB traces
> needing to be closer together to connect to the 16A relay. I believe I
> have paraphrased Jim's assertion correctly.
>
> I decided to do a quick PCB test. I made a small PCB with various
> trace widths and hi-potted them. The traces were about 2 inches long and
> parallel to each other. Also I live at 5400' above sea level and the
> humidity is low here.
>
> Trace spacing and voltage breakdown:
>
> 0.115" = 3000VAC rms = 4243V pk
> 0.085" = 2300VAC rms = 3253V pk
> 0.055" = 1800VAC rms = 2546V pk
> 0.040" = 1600VAC rms = 2262V pk
> 0.025" = 1100VAC rms = 1556V pk
>
> The voltages are very hard to measure accurately so the voltages shown
> are slightly below the breakdown points. Also the traces did not have
> bends but slightly curved to the hi-pot connection points.
>
> Using 0.200" trace widths with the center trace centered on the relay
> center contact pins and the outside 2 traces offset by 1/4 of the trace
> width away from the center trace (to increase the trace spacing) would
> result in a trace spacing of 0.050". By offsetting the outside 2 traces
> so that the hole is .03" from the trace edge would yield PCB trace
> spacings of 0.070". Using this offset and increasing the PCB trace
> width to 0.240" would still allow a PCB trace spacing of 0.05".
>
> Running legal limit into a 4:1 SWR will produce a maximum peak voltage
> of sqrt( 2 * 1500 * 50 * SWR ) = 775V pk
> Even running 3000W into a 6:1 SWR is still a maximum peak voltage of
> sqrt( 2 * 3000 * 50 * SWR ) = 1342V pk
>
> Obviously the above maximum peak voltages occur at only 1 impedance for
> the given maximum SWR but we need to be sure about the voltage
> breakdown. I haven't done any tests to see if 0.200" or 0.240" traces
> on either one or both sides of the PCB can handle the current necessary
> but this discussion was intended to be about voltage breakdown. Has
> anyone done/seen an analysis on RF trace currrent capacity in the HF
> region?
>
> From this analysis it appears that there is plenty of PCB trace spacing
> for all amateur powers provided the relays are not between an antenna
> tuner and the antenna where the voltages could be even higher. Comments
> as to any problems with the analysis?
>
> 73,
> Larry, W0QE
>
>
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|