At 01:44 AM 10/19/2005, you wrote:
>I'm still on the hunt for a decent set of relays for a remote
>matching network. My goal? 1.5KW remote switching box. I went to
>a local surplus emporium in Minneapolis to scour the junque boxes
>for decent relays. Eureka! A boat load of 12v relays! 10A,
>breakdown voltage of 2KV between contacts, and 4KV between contacts
>and coils. 12v switching. DPDT and DPST styles. Visual inspection
>shows a good 0.050" to 0.060" air gap on the contacts. I selected
>an Aromat JC2aF-DC12V.
>
>Tonight, I decided to examine them in more detail. Aromat spec
>sheet says 0.1 ohm on the contacts. Contacts are silver
>alloy. Maximum switching power 2,500VA. MTBF 10^5. Kewel! They
>look perfect. They are cheap $4. They are small open frame styles
>with short RF paths of copper strap material.
>
>I soldered RG174 cables to the contacts to switch both shield and
>center. One cable went to a 50 ohm load, the other to the Autek
>VA1. Open! Switch the relays about 50 times and then they start to
>work. I can see that the contacts are closing alright. But no RF
>gets through. So I took a stray JA QSL card laying on the desk and
>ran it through the relay contacts. Then they start to work. SWR
>reads 1.12 :1. Close enough. But unreliable. They don't always
>connect--especially on the Normally Closed side of the
>relay. Repeating it with another relay yields identical results.
>
>Theory 1: These are designed to be hot switched. The tiny arch from
>a hot load fuses the contact. Stop worrying about it. Push some
>decent power through them. They will fuse and be just fine.
>
>Theory 2: My gut tells me that a robust design would include a self
>cleaning wiping action on the contacts, like on a band switch. I
>will not get reliable switching using traditional relays.
>
>So how does Ameritron get reliable connections? And Com-Tek with
>their phased arrays? Are relays unsuitable for this
>application? Should the contacts be gold? What am I missing here?
>
>Any help the engineer types can provide would be appreciated.
Ford--
We once had a similar problem with all of the Dow-Key coax relays used in our
radio astronomy observatory. These were the standard T-shaped relays with gold
contacts. We would use a thin strip of a 3x5 file card to clean the relays.
Even if the RX signal was getting through, within a week or so after a cleaning
the signal would have random phase shifts, a critical problem in an
interferometer.
After some testing we determined that there was a oxide buildup, even on the
gold contacts, that prevented them from making a good contact. But the oxide
layer was thin enough that the capacitance of the contacts would allow the
signal to pass, albeit with some large variations in phase
shift. We also found
that ohming these with an old style ohmmeter with VOLTS on the probes would
show they were OK, but a newer meter with only mV on the probes would show
them to be open.
Our short-term solution was to pass the DC for the preamps THROUGH all of these
contacts so they were hot switched by injecting the DC onto the coax before the
relays rather than after the relays. Our long term solution was to
use something
called a log-cell (if I recall correctly -- this was 30 years ago)
that was a mercury-
wetted relay in a glass cylinder. They were SPDT relays with the
surface tension
of the mercury keeping them engaged once an initial pulse of energy was applied
to the coil, making them inherently latching with no need to maintain power.
I believe that an open relay will show good continuity if handling some power
periodically, but this means that an antenna switching relay will never see
any power on the RX contact side. You can still hear through it due to the
capacitive coupling between the contacts that are only separated by a few
molecules, but the phase shifts may be surprising and the losses may be
significant.
In remote switching an antenna tuner, as long as the contacts get some RF
(or DC) power periodically they should be fine. But if there is no
power, then
critical applications such as phased arrays will show problems when only
used for receive signals.
Solutions: Vacuum relays. Or relays with better wiping action. Or running
a voltage through the receive contact side.
73 John W0UN
_______________________________________________
Topband mailing list
Topband@contesting.com
http://lists.contesting.com/mailman/listinfo/topband
|