Wetting current

Any idea if the current could be ac?

I have the possibility to inject ac current, with more than 1mA, a lot easier than DC current

you can hear them ring mechanically too when you pull a magnet away from them.

maybe you need some lovely mercury wetted relays.

We use lots of Fujitsu FTR-B3G relays... almost 100K so far.

DPDT, surface mount, small and reliable. We use the latching version at thermocouple levels. They are good up to 3 GHz too.

Reeds are expensive junk.

They are nice but you do have to mount them so they stray vertical. Orientation-insensitive mercury wetted reeds did get touted from time to time, but never lasted.

Of course Bell Labs develped them because they needed something more reliable than traditional relays for the telephone network.

They were invented in Russia in 1922, but Bell Labs developed them into something useful in the 1930s - there's a 1941 US patent.

If John Larkin doesn't like them, it probably means that he didn't use them as carefully as he should have.

If your open circuit ac voltage is in the 5-15V range I guess that should work. The worst switch contact oxidation film I saw took about

12V before reliable contact was made. Since you don't know what contacts your circuit may be exposed to I would steer on the high side (10-15V?). I think once the current is above a few hundred micro amps you should be OK.

Is there a reason why you cannot see the relay coil voltage instead?

piglet

For a reliable bulletproof general solution, I think it would be best to use something like 24V/10mA and pulse the current for the measurement to minimize the wasted power.

You can probably get away with less..

No Gray-or-compatible encoding? That's too bad.

Best regards, Piotr

Yes, the relay is in another product, so we run wires from our box to the relay contacts, to detect the state of the external product

Cheers

Klaus

The solution we have now is a capacitor charged to a small voltage. Then when the relay closes, high current runs to discharge said capacitor

In this case the voltage is low, but peak current is high (for less than 100ns)

Cheers

Klaus

It's hard to do that with relays switching resistors, especially relays that bounce a lot. I guess it could be done, but it would take too many relays and resistors.

We do resistor simulation electronically. That can be kept monotonic.

You might look into the BSTJ archives. I recall chatting

20 years ago with someone from the Labs who had studied it intensively.

A shortcut to them is

Well it's not 1930 anymore and the applications for reed switches are few and far in between. The wiki article even mentioned reed switches being used as sensors in old brushless motors. Awful.

Bosch are still using reed switches to make their dishwashers unreliable

So much for "German engineering"

CH

I've never buy anything from bosch but a few power tools.

european appliances are just pure trash when it comes to design and repairability. Oh, the cars too.

That's why one 'powers' a relay coil, instead of signalling it. The motion of the contacts means the device is an electric motor, with all the back-EMF you expect, generated by moving magnetic parts in the field of the coil.

Do your FTR-B3G things work while wet with solvents, hot, and in a corrosive atmosphere? Can you position them to serve as coil-current or B-field detectors? How's the explosion-proofing on those make/break events, in case you need 'em in a petroleum-handling facility? Did you ever pot an assembly with relays?

Reeds have their uses. I'm puzzled how there's a thermocouple problem; other than length of the capsule, there's nothing to encourage thermal gradients, and both glass-seal metal elements ARE the same alloy. Contact plating is a few microns, not a problem at all.

They work great in our products. DPDT, nanovolt offsets, 3 GHz bandwidth, $1.32 for the non-latching surface-mount version.

Coil current, sure. They have coils. But I want signal switching, not mag field detectors.

Reed relays are fun B_field sensors. If one is energized, it can keep an adjacent one from dropping out. The open coil field leaks all over the place.

Nice in theory, but they have terrible thermocouple offsets in real life. The coil gets hot and the leads are not the same metal as the reeds.

Why do some people love and defend reeds? They are expensive, unreliable, usually just SPST, twang, have big thermals, are giant compared to a modern telecom relay, and have to be kept apart from one another.

You and sloman love reeds and love China. You're probably the same person.

The leads are part of the magnetic circuit, they ARE the same metal in all the examples I've seen. Are you talking about RELAY leads, or reed leads? Coil dissipation is less of an issue in latching relays, but you can make those with reeds if you care to. It's an easy design with a cruciform pole piece.