Re: MEMS again

On Thu, 14 Nov 2024 00:10:51 +0000, Cursitor Doom <cd@notformail.com>
wrote:

On Wed, 13 Nov 2024 14:12:29 -0800, john larkin <jl@glen--canyon.com>
wrote:
>

On Wed, 13 Nov 2024 19:00:29 +0000, Cursitor Doom <cd@notformail.com>
wrote:
>

On Wed, 13 Nov 2024 10:16:54 -0800, john larkin <jl@glen--canyon.com>
wrote:
>

Menlo Micro is making some grand claims for their power and RF relays.
They appear to be electrostatic MEMS relays. Lots of people have tried
that in the past couple of decades, but the parts haven't been
reliable.
>
https://menlomicro.com/images/general/MM5130_Datasheet.pdf

>
On-State Insertion Loss is only stated at 6Ghz. I'd be interested to
know what it is closer to the top end. A graph would be better,
though.

>
All the parts can only cold switch. That would seem to be a problem,
especially for the power parts. How do you cold switch a circuit
breaker?
>
The two killers of MEMS relays have always been contacts frying and
contacts sticking.

>
Holy crap, I didn't spot that! Big downer.

I'd love a good small-signal MEMS relay. No semiconductor switch can
touch its Ron*Coff time constant, and free isolation is cool too. Even
a 90 cent telecom relay wins there.

Those Menlo things can probably hot switch low level analog stuff.

I recently used a bunch of Analog Devices switches, ADRF5024, to
switch some pulses around. Had to characterize them ourselves for DC
operation... ADI was no help.  They work fine, for $108 each.

The signals come in at right angles, which dominated the whole PCB
layout. Some really tiny SPST switches would be cool... put them
anywhere. I like single opamps for the same reason.