Small Signal Relays

FUJITSU FTR-B3GA4.5Z-B10 non-latching

FUJITSU FTR-B3GB4.5Z-B10 latching

These are under $1 in quantity.

They are damned good RF switches.

Oh dear that's much to complicated for me. :^)

Thanks again, GH

But 10 uA (which I saw as some minimum current) into

How does that work?

Saturated output turns off automatically?

Cheers

Klaus

You don't have to get far into relays to realise that it's a good idea to differentiate between open contact relays and reed relays

John Larkin hasn't bothered, and it isn't obvious from the Fujitsu datasheet either, though they probably aren't reeds.

Mercury-wetted reeds are pretty nice - they do have to be mounted vertically, but they offer a lower, and more constant contact resistance than regular reeds, they last ten times as long, and they don't bounce on closing.

Thermal offset voltages can be problem - the coil is a heat source and the reed material has an appreciable thermocouple voltage versus copper.

K0 is the desired relay state. Set that high or low, then pulse KA low for a few milliseconds to set the relay.

This board has 16 latching relays, driven from an FPGA. K0...K7 are an

Nice :-)

Thanks for the explanation

Regards

Klaus

Going to take a wild guess here and speculate that mercury-wetted reed relays aren't going to be RoHS-compliant, as the Fujitsu parts claim to be. Most reed relays (mercury-wetted and otherwise) don't look like these, being longer than they are wide.

These do seem like nice parts for RF, as long as they're mounted under a shield. Even at this late date, solid-state RF switches are either terrible parts that have a long list of caveats, conditions and constraints like the Analog Devices MEMS parts, or way too expensive like the PE42020. Usually both. :(

-- john, KE5FX

Reed relays are awful: big, expensive, unreliable, sole-sourced, bouncy, twangy, power hogs, and have horrible thermal offsets. If you put them too close together, the field from one can keep the other one latched.

Why buy a flakey SPST reed when you can get a nice DPDT relay for less?

MEMS relays have been the wave of the future for, what, 20 or so years now?

At least. Nice low capacitance, but of course you can fuse the contacts by discharging a 100-pF capacitor. ;)

Cheers

Phil Hobbs

Plus IME the housings are usually nylon-6 or 66, which is hygroscopic and leaks like the proverbial sieve.

Cheers

Phil Hobbs

The twanging is fun. After they (eventually) quit bouncing and close, the reeds vibrate in the mag field, and make millivolts of ringy signals that last many milliseconds. The waveform resembles a bell whacked by a hammer.

I guess that the average reed relay contacts make a giant loop with the traces on a PC board, said loop circling around half of the coil. Yuk.

Reed relays are hermiticly sealed. The mercury (and there isn't a lot of it) can only get out if you break the glass encapsulation, and that's buired inside the coil and some kind of squared off package.

Reed relays can work well at RF frequencies, but getting the packaging right to maintain constant impedance (when the relay is closed) is tricky, but has been done.

They aren't unreliable - they were invented to provide a relaible relay for the telephone system.

The contacts aren't exposed to the atmosphere, which can be important.

You do have to pay attention to contact ratings (which is the kind of detail work that John Larkin doesn't like).

They certainly aren't sole-sourced. They do bounce (if the contacts aren't mercury-wetted). "Twangy" would appear to be the same complaint.

Latching reeds aren't power hogs (if you can find them) and consequently don't have horrible thermal offsets - which aren't all that big anyway.

If you must put them close together you can use soft iron to shield one from another (which is the kind of detail work that John Larkin doesn't like).

Smaller contact reistance, much lower leackage current, smaller "off" capacitance.

At least that long.