Posted by
wayne burdick on
Feb 12, 2006; 6:16pm
URL: http://elecraft.85.s1.nabble.com/K2-Relay-contact-resistance-tp386534p386540.html
Stewart,
The small relays we use all have wiping action. This was one of our
selection criteria. We also ran them through millions of switching
cycles just to be sure, and not a single one failed, which bodes well
for relay life in our products. Gary, our customer support engineer,
has noted that relay failures in the field are also extremely rare, and
in most cases can be traced to damage during initial installation
(e.g., melting a corner of the package with a soldering iron).
Even if a relay did develop a problem, you should only replace that
relay, not an entire group. All of our manuals include signal-tracing
procedures that should be helpful in identifying a failing relay. In
the case of our internal ATUs, you can exercise each relay individually
using the rig's ATU menu entry. The T1 allows individual relay testing
using its switch/LED user interface.
If you ever do have to replace a relay, it's wise to remove *all* of
the solder using a large (not wimpy!) vacuum desoldering tool, followed
by solder-wicking all pins to clean things up. Even still, you might
have to pry a bit with a jeweler's driver to get the relay off.
Since all of our PC boards are double-sided, there's a small risk of
pulling out the plating on one or two plated-through holes when you do
this. But we anticipated this, so we make all PC board connections to
relays on the opposite side of the board. The loss of through-plating
is then a non-problem. We apply this rule to all large components that
obscure their top-side traces. The rule is harder to follow in dense
areas around ICs, but in this case you can solder the pins both top and
bottom if plating is lost.
We also strive to maximize the size of PC board pads and traces
whenever possible. This simplifies possible rework, lowers trace
impedance and improves reliability. Our PC layout methodology is in
sharp contrast to most of the electronics industry, which seems to use
the smallest trace widths they can get away with, even when PCB
real-estate isn't an issue.
73,
Wayne
N6KR
-----
www.elecraft.com
_______________________________________________
Elecraft mailing list
Post to:
[hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft
Help:
http://mailman.qth.net/subscribers.htmElecraft web page:
http://www.elecraft.com