AF6AY wrote:
On Mar 21, 4:31�pm, Joerg
wrote:
AF6AY wrote:
"Joerg" posted on Fri, Mar 20 2009 6:06 pm
ken scharf wrote:
JIMMIE wrote:
Relays are ok in those applications, where there are actual currents
flowing. What I meant was pure signal switching with no DC currents.

That's called "dry" switching in the electronics industry, has been
called that for over a half century.

Sealed relays, Reeds, mercury-wetted and such work quite well there. But
non-sealed versions have issues and a snap-on plastic cap ain't a seal.

Didn't say that that it was a seal. But...there are many kinds of
'seals' and there are many kinds of contact alloys which few hobbyists
investigate.


Most must make do with whatever places such as Digikey offer at
affordable cost.


Those problems became really nasty after we moved across an ocean and
all this stuff was in a sea container for two months. After that almost
all the band switching relays had problems, had to clean all of them.
Now I am down to one sticky relay every couple month or so.

My wife and I own a nice 2005 model Chevrolet Malibu MAXX that we've
driven in both driving rain with some intermittent ice to Wisconsin
from California and back, to Washington state from California and
back. It has two little boxes of many small-signal relays, only a
very few qualifying for large-signal types. Never a problem since we
got it in late June of 2005. Those relay boxes all have lids and the
small relays have little covers and that auto has definitely been
exposed to the environment many times in the last 4 years.


Sure, but in a car relays always switch something that draws a serious
current, more than a few milliamps. That works for a long time. Well,
until a "weld wart" shows up, like it did on a relay in out Genie garage
door opener.


I can sympathize with your bad sea container shipping experience but
consumer-grade radios (such as for amateur radio) were never designed
to be exposed to sea evnvironments. Ask yourself how all those off-
shore made radios made it to the USA? Inside standardized container
boxes.


Usually they are carefully packaged and silica gel is included. When you
move that usually ain't the case. Plus the radio was quite old by that time.


I didn't make my comment lightly or pretend that I know everything
there is to electronics. I do know, by a rather large set of
experiences that 'dry' circuit relays (hermetically sealed OR by
reasonably-good individual covers) will work without having to be
'operated many times' in order to 'clean their contacts.'


Hmm, I've had that happen even in $xxxxx lab gear, not quite
consumer-grade. Never really with mil gear though but those guys can
often design to "the sky's the limit" cost goals.


I'll cite one application that is military, the US AN/PRC-104 manpack
HF transceiver. About one-third of that backpack radio is an
automatic antenna tuner so that one whip length can be optimized for
best electrical characteristics. It does that with a rather
conventional microprocessor control driving two banks of binary-
sequence inductance and capacitance values switched by relays. It has
been in operational status with the US Army since around 1984 and is
expected to be phased out soon in favor of more modern HF-to-UHF
transceiver designs. It's been a while since I've seen the guts of it
but I don't recall that it had any hermetically-sealed relays in it.


But I bet no expense was spared to pick the very best parts for that radio.


Right now I'm beginning to start cutting holes for a rebuild of an
'ancient' HF receiver once made for my late father wanting to tune to
some SW BC stations, principally Radio Sweden back in 1964. I've been
fortunate to get a large collection of North Electric small sealed
relays dating back to about 1955 production which I've already
breadboarded for bandswitching use. Very familiar with relay testing,
I found NO problems from 'dry' contact switching. The low capacitance
to ground and minimal series inductance from the contact set do NOT
upset any of the L-C circuits being switched.


Sealed relays are ok. I also have a few really old ones here and so far
they have never let me down. But I don't have enough to equip a filter
bank with 15-20 in- and outputs. For that I've got a stash of HSMP-3810
PIN diodes.


replaced with PIN diode circuits and that, of course, made the issues
completely go away. But it's always a hassle to do in an exisitng circuit.

Yes, I have that in the filter board of my two decade old Icom IC-R70,
all switched with 'RF switch' diodes (no registry number). If needs
be, that entire filter board could be enclosed to prevent any problems
from the environment. Icom used that sort of semi-conductor switching
for ease of overall parts cost along with reduced labor costs to
Icom. Such work fine at the LOW impedances involved (50 to 75 Ohms)
but I've tried to duplicate that at 10 KOhms on a breadboard and have
run into problems with diodes' own impedances affecting circuit
operation. Those could be solved, I'm sure, but I didn't care to
spend weeks fiddling with them for my rebuild project. The relay
contact set had only the shunt capacity and series inductance to
contend with and those were very low values and easily compensated for
alignment.


Yeah, anything above 1K or so is less suitable for diodes. That's where
JFETs come in which, in turn, are not so hot for anything low impedance.
Unless you use the fat expensive kind and they've got too high
capacitances. Many fine FETs like the P8000 have gone over the rainbow
bridge by now. I guess they only found homes in ham gear and that wasn't
a large enough market.

--
73, Joerg