On Sun, 06 Sep 2009 14:52:38 -0700, Jim Thompson wrote:

On Sun, 06 Sep 2009 16:43:15 -0500, Tim Wescott
wrote:

On Sun, 06 Sep 2009 11:50:43 -0700, Jim Thompson wrote:

On Sun, 06 Sep 2009 13:33:45 -0500, Tim Wescott
wrote:

On Sat, 05 Sep 2009 14:08:54 -0700, Jim Thompson wrote:

On Sat, 05 Sep 2009 12:43:20 -0700, Jim Thompson
wrote:

On Sat, 05 Sep 2009 14:01:59 -0500, Tim Wescott
wrote:

On Sat, 05 Sep 2009 11:03:51 -0700, life imitates life wrote:

On Sat, 05 Sep 2009 12:48:41 -0500, John Fields
wrote:

On Sat, 05 Sep 2009 09:41:06 -0700, Jim Thompson
wrote:

I haven't done a relay driver circuit for, literally, eons.

Where do I look for reed relays that work on 3.3V supplies with
specications for the coil inductance?

Thanks!

---
That's a toughie.

I can't recall ever seeing inductance data on a spec sheet; I
think you'll probably have to go to the manufacturer for that
one.

JF

Would not the real question be why would someone concern himself
with the solenoid inductance of a miniature relay? I could see
it if it were huge.

Jim designs custom IC's, so (I assume) he's concerning himself with
it so he can make the output stage of his circuit both economical
and robust. He may even care about making the relay turn off
quickly.

It seems like a perfectly valid concern to me, even if the circuit
in question _isn't_ custom -- what if you're powering the relay
from logic, and want to insure fast & safe turn-off?

Dead-on, Tim!

...Jim Thompson

Like this...

http://www.analog-innovations.com/SE...elayDriver.pdf

In the discrete world, with more voltage tolerance available, you
can scale this to larger controlled flyback potentials, for fast
turn-off.

...Jim Thompson

Clever.

I first came up with this scheme for a light-weight flyback switcher
built on a bipolar process, circa 1975.

I assume that the gate threshold of MP1 guarantees that the drain of
MN1 never exceeds your max voltage?

That and the ratio of the 10K to 33K resistor.


I could see a potential for this to oscillate on turn-off, with the
inductor working against the delays in the two FETS,

FET's are sub-nanosecond

but if it did it may
well not matter, and your relay coil is probably sluggish enough to
keep it from happening anyway.

Do you get to tune the gate thresholds of the individual parts, or is
that usually locked in by the process?

Usually locked by the process though, occasionally, I run onto
processes that have a bunch of different device _types_, some even
offering depletion mode MOS devices... nice for kick-starting
micro-power stuff ;-)

...Jim Thompson

When I was in school in the early '80's all the IC design stuff was
digital, and we were all told that no one was going to get a job
designing an analog IC ever again, 'cause they were all going to go
away.

Silly me, I listened and stayed away from IC design entirely. But I
have fun with what I do...

There still seems to be ample analog functions needed. All I'm seeing
is more digital controls added around my analog, to do such things as
auto-zero, auto-cal, etc.

The mistake that Academia made (and industry is still suffering from) is
stating "all the processing is going to be done digitally" without ever
answering the question "but how is the information going to get in, and
the commands out?"

Because you can design the best damn DSP in the world, but if you feed it
inputs that are filled to the brim with gallons of ground bounce,
egregious thermal noise, nonlinear effects and timing jitter with maybe a
teaspoon full of real signal, you aren't going to have a good system.

--
www.wescottdesign.com