Re: relay coil inductance

[life imitates life]

Quote from life imitates life,
" I am quite certain that any 3.3V relay you choose to examine will prove
to be faster than any of the numbers that more than 10 year old document
measured. Even the diode."

Just doesn't seem like you saw a lot of wisdom in the 10 your old document.
Tell me, what part of it do you think is wise?
I didn't morph and I'm not a retard,
I think your losing your mind.
Mike

Do you think the back EMF generated on an unsuppressed 3.3V relay
solenoid coil upon excitation removal is -750 Volts?

 

[amdx]

Do you think the back EMF generated on an unsuppressed 3.3V relay
solenoid coil upon excitation removal is -750 Volts?

Nope, anything else you need help with?
Mike

 

[life imitates life]

Nope, anything else you need help with?
Mike

Then the ten year old PDF is meaningless. It would appear as if it is
you that needs help.

 

[Archimedes' Lever]

Yes you can. Try it in a real relay and see. A small value capacitor
will make the contacts open sooner. I will let you flounder about
trying to come up with why.

The cap will be at the excitation voltage when excitation is removed,
which means that excitation potential will be on the coil LONGER,
regardless of what happens as the flux field collapses, which will not
occur until AFTER the cap discharges, AFTER the excitation stimulus was
removed. ergo LONGER times.

 

[Archimedes' Lever]

Boy! that was a worse flounder than I expected. Using upper case in
your replay make you look even sillier.

You're an idiot. There are only a few ways to show emphasis with mere
text, and no, I am not going to learn the gang boy retarded texting realm
that wastes so much of the world's time to please a pussified **** like
you.

 

[John Larkin]

guess that is the idea with clamped fets, let the voltage fly up to as
much
as you can live with and limit it with the fet.

still need to know the inductance to figure out how much power will
go into the fet when limiting the voltage.

-Lasse

A relay coil voltage won't fly up to infinity; circuit capacitance and
eddy-type losses will limit the peak voltage. So use a fet that's good
for more, and it won't avalanche. Of course, the substrate diode can
complicate life.

It's hard to build a single-inductor boost converter that steps up as
much as 100:1. The coil and fet capacitances will getcha.

Another option is to just slow down the gate drive.


John

 

[John Larkin]

That's interesting. I'll have to think about that. But since
AlwaysWrong disagrees, I know it's true.

John

 

[boB]

Do you think the back EMF generated on an unsuppressed 3.3V relay
solenoid coil upon excitation removal is -750 Volts?

It's not beyond the realm of possibility.

boB

 

[Archimedes' Lever]

That's interesting. I'll have to think about that. But since
AlwaysWrong disagrees, I know it's true.

John

You just got done posting that one could slow it down. In ANY case, the
unsuppressed coil opens the contacts it has faster than any of the
suppressed methods.

 

[life imitates life]

It's not beyond the realm of possibility.

boB

Given the 12 V figures, and the difference in core mass between the two
form factors, I would say that it is decidedly easy to contend that it IS
beyond the realm of possibility to get that level of flyback from a 3.3
Volt stimulus into a tiny, low mass, low turns count, open ended coil..

 

[Jasen Betts]

Diode+resistor or diode+zener are both a lot faster than just a diode.

Or even just a series or shunt resistor, if you can spare the power.

No clamp at all is even faster; just let it fly up and ring, with
suitable precautions.

The untimate would be to apply a large reverse voltage to the coil
until the current goes to zero. That's what happens automagically in a
linear amp driving an inductive load with current feedback.

you can probably come close to that with just the right RC snubber
(just enough for critial damping)

 

[John Larkin]

The capacitance isn't the only thing making the field continue after
the contacts open. The eddy currents in the core are also at work.
The contacts start to open when the field going through the bit of
iron that can move drops below some value. Some relays are made with
an effective shorted turn on purpose to keep the back EMF within
reason.

I know the effect is real. I assume that lowering the SRF down to
nearer the time frame of the decay of the eddy currents causes the
field at the point where it matters to fall below that magic value a
little sooner because the current in the coil is backwards at just
that instant.

What complicates things is that L is changing, usually severely, with
time, as the armature moves. Less for reeds than regular relays.

This would be a bear to Spice.

John

 

[ehsjr]

A capacitor across the coil makes the contacts open faster. Try it,
it works.

I have, it doesn't, for the cases I've tried. Your statement
has to be incorrect, if you meant it to apply to all cases of
a cap across the coil. Perhaps you have a specific example in mind?

Ed

 

[Archimedes' Lever]

One could go a little bit faster than open-circuit unsupressed, but it
wouldn't be worth the considerable trouble. The point is to get rid of
all armature magnetization as fast as possible.

The shunt capacitance thing might just work.

John

Riskier is to find the voltage that closure occurs at on say 50 units.
Just a tad over that threshold voltage is where you want your drivers to
fire at.

Then, release will be as immediate as one can physically make it as the
closure plate will detach even before the voltage drops to zero.

In such a case, one could design a "soft release" in the driver pulse
signature that negates any back-emf spike since there is no high slew
rate fall-to-zero stimulus removal, which is the cause of the spike to
begin with.
The soft release will not slow the relay much if at all, because it is
already close to losing its closure.

Perhaps they should design a relay that has a coil to close it, and a
coil to aid in turn off velocity, and soak the back emf as well.

Likely too expensive, but I wonder if there are any piezo stack
actuated relays out there...

 

[life imitates life]

Jim designs custom IC's,

All IC designs are "custom". Duh!

 

[Archimedes' Lever]

Cool. 5% won't pull in at all,

You pulled your numbers out of thin air, and the drive voltage would
not be set to a point that would have such a result.

and 25% won't pull in when the
temperature goes up.

Why? Magnetic curie point is like over 400°F, so I don't think we'll
have any worries there.

Great idea.

Depends on the application. Hell, a feature that tweaks the voltage up
on a channel that has displayed errant operation wouldn't be all that
hard either.

 

[life imitates life]

And, Jim designs them, with the help of some of the members
here at A.B.S.E., and I imagine some of the students at the
University. ;-))) ( ducking )

Bill

Where are your relay coil driver stimulus release spike abatement ideas
at then, asswipe (slings shit your way)? I fart in your direct
direction!

 

[life imitates life]

That was three fecal references in 24 words, 12.5%. I think 10% is the
threshold for nether-region obsession.

John

Don't count 'em up and attribute them to a single person, you shit for
brains dumbfuck!

I made ONE reference (slings shit your way). Not that a retarded wuss
like you could grasp the fact that he wrote (ducking), much less what it
refers to.

An "asswipe" is a wipe for the ass. There is no fecal reference there
at all. The fart reference is "bathroom humor" not fecal, you pathetic,
cringing, wussified milksop.

So **** OFF yet again with your retarded claims that you make
incessantly. You're a goddamned idiot, as well as being stupid for
thinking that some number validates your retarded claim.

It doesn't get any more pathetic than that, Johnny. Somebody should
drop a burning paper bag on your doorstep.

"I think 10%... blah blah blah." You're an idiot, Johnny. We only
needed ONE occurrence of your stupid shit in the groups to validate that
FACT. Give it up, you shit for brains dumbfuck.

It's like this... If I am "obsessed with it", then you are made of it.
So make up your mind. If your claim is valid, you are made of it. If
you are a lying idiot that has zero psychological knowledge, much less
qualifications, then... you are STILL made of it.

You lose, either way, Johnny boy.

 

[ehsjr]

Then you'd better drive it to at least the mfr's guaranteed minimum
pullin voltage.




The copper coil winding has a positive resistance tempco of 0.4
percent per degree C. Between ambient changes and self-heating, it's
another opportunity to screw up.




One could take advantage of the pullin/dropout hysteresis: power the
high side of the coil from an R-C off the main supply. That will let
you overvoltage the coil on pullin - fast close - and then droop the
current into the hysteresis zone, setting it up for faster dropout.
This can seriously reduce self-heating, too.

All of which is sort of academic.

Just putting a resistor in series with the coil will speed up
everything.

John

Just the drop out - it'll slow down the pull-in.
As you mentioned, kicking it with higher voltage with
a cap at pull-in, and holding it energized below pull-in
current with a resistor does the speedup job both ways.

Ed

 

[life imitates life]

I think most people derive more enjoyment from the input end of their
digestive system than from the output. Buy hey, whatever turns you on.

John

AGAIN, you retarded ****, NOBODY said a goddamned thing about enjoying
ANY fucking input(s) or output(s) of any kind on any part of anyone, you
goddamned won't fucking let it go sub-human pond scum!

You do know what pond scum predominantly is, right?