Re: relay coil inductance

[life imitates life]

---
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.

 

[Jon Kirwan]

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.

Probably, Jim is designing an IC where every 'square' or 'hex' counts.
So maybe he doesn't want to use more than he has to? Just a thought.

Jon

 

[life imitates life]

---

L di
E= ------,
dt

so the diode damping the "spark" when the driver went open would slow
down the opening of the contacts, maybe?

Or maybe the magnetic field building up and breaking down might affect
some nearby component?

Dunno...

JF

3.3 V relay that requires a can over it due to the field it makes?

I suppose it is possible. I doubt much field is generated though.

The collapsing field gets pumped into the diode. One could design the
firing circuit to "slow stop" as well. reducing any collapsing flux to
minimal since it is created as a result of the slew rate at which it was
collapsed.

 

[life imitates life]

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?

Turn off can be immediate. The contacts open before the solenoid
completes opening to the gap it was at when off.

 

[Archimedes' Lever]

What would NymNoNuts know?

What would a retard that replies the way you do know?

 

[Archimedes' Lever]

Can't use a "flyback" diode... any "diode" will induce destructive
substrate currents.

You should not fire them directly from the chip. They should ALL
incorporate a driver AFTER the chip for each one. Diodes have been
placed across the coils of relays for decades. There is a reason for
that. The diode CLAMPS current, keeping it from the driver, idiot!

So I'm using an old design trick of mine from "eons" ago... active
device turn-on to limit "flyback".

The flyback occurs at turn off, NOT turn on, idiot, and the diode is
ONLY in play during turn off events, and is specifically for clamping
that flyback event.

So I need the inductance to estimate device sizes.

I think you are in overkill mode.

What would NymNoNuts know?

Likely far more than you.

Heaven help us, but I estimate he's a
technician in our "defense" industry :-(

That is another one of your problems. Your capacity to make a valid
estimation died twenty years ago, and your capacity to make one in your
senile condition is even less than it was when you were not senile. Even
then, it was marginal, at best.

 

[Archimedes' Lever]

Dead-on, Tim!

...Jim Thompson

Relay opening time on a mechanical relay is NOT electrically related.

It is mechanical. You can slow it down electrically, but there is
nothing you can do that makes the event occur any faster.

 

[boB]

Relay opening time on a mechanical relay is NOT electrically related.

It is mechanical. You can slow it down electrically, but there is
nothing you can do that makes the event occur any faster.

Do you think that, with a diode across the relay to reduce the flyback
voltage spike, that slow-er decaying current, and its associated
magnetic field holding the relay in, does NOT slow down the relay
opening ???

boB

 

[life imitates life]

Do you think that, with a diode across the relay to reduce the flyback
voltage spike, that slow-er decaying current, and its associated
magnetic field holding the relay in, does NOT slow down the relay
opening ???

boB

The coil's flux collapses and creates a spike. The diode clamping that
spike does NOT slow the spring loaded return time of the plate which is
attached to the contact(s) as it pulls away from the solenoid core end
face.

So, the answer is NO. The magnetic field is collapsing, as in NOT
"holding in" the relay any longer. The plate begins to pull away as soon
as the power is removed, and the clamping diode does nothing to slow that
process.

The field is collapsing, not being splayed out. There is no longer an
attachment force as soon as the power is removed. The collapsing flux
induces a current though the diode, but that diode does NOT slow the
collapse rate. That rate was determined by the slew rate of the voltage
change which was full voltage to zero in a practically square wave fall
rate.

The collapsing flux makes the back EMF. The diode eats that current.
The plate has already been released long before those events. Diode or
not, the relay opens at the same rate. The diode is there to kill the
spike, and that is all.

To slow the process, one needs to slow the rate at which the excitation
voltage falls. Once it falls below a certain value, however, the relay
will STILL "snap" open, so even that method does not "slow" things much.

Speeding one up, however, is what the engineers that designed it did.
I doubt you will be able to improve on their works short of adding a
solenoid to pull the relay off faster than the mechanical spring does.
A push-pull relay where there are two solenoids operating it.

Otherwise, you are simply tied to the mechanics of the system.

 

[life imitates life]

That sounds like a bureaucrat's version of "immediate".

More fucking retarded bullshit from someone with no real argument that
is based in fact.

A relay will
_always_ turn off with some delay,

No shit, you retarded ****. It was *I* that said that you can slow the
opening, but will fail to make it faster than it already is.

first for the magnetic field to die
down,

Wrong. The closure plate begins to pull away as soon as the excitation
voltage is removed. A static field pulls in and holds the plate, a
collapsing field does not.

then for the contacts to physically move.

That is a function of how far BEFORE the plate closure point they MAKE
contact. They will always make contact before the plate is fully
clasped. They will always remain in contact as the plate pulls away,
until it reaches a specific point of opening.

Even if you don't care about turning off the magnetic field as fast as
possible,

First it has to matter. Nearly all designers remove excitation voltage
immediately. That means the field is removed immediately. That is also
what causes the spike.

if you don't size your snubber circuit right the coil will take
out your output driver when you try to turn off the current.

That is what the diode is for. There will be no current any further
back in the circuit if the diode does its job.

 

[life imitates life]

doesn't have a clue about coil energy
that has to go somewhere.

...Jim Thompson

Lighting your retarded ass up with a few fibrillation inducing jolts is
well within the realm of my knowledge, and proves yet again that you have
no clue about that which you spew.

Of course shit is not all that conductive, so I might have to up the
juice quite a bit to have an effect on you.

pasticcio

 

[DarkMatter]

Lighting your retarded ass up with a few fibrillation inducing jolts is
well within the realm of my knowledge, and proves yet again that you have
no clue about that which you spew.

Of course shit is not all that conductive, so I might have to up the
juice quite a bit to have an effect on you.

pasticcio

Not to mention the hard as rock heart, if there is even one there.

 

[amdx]

Do you think that, with a diode across the relay to reduce the flyback
voltage spike, that slow-er decaying current, and its associated magnetic
field holding the relay in, does NOT slow down the relay opening ???

boB

Here's a pretty good pdf with explanation of different spike suppresion
methods.
It lists drop out times for the different methods, for one particular relay
and values.
http://www.kilovac.com/appnotes/app_pdfs/13c3311.pdf
PS. A diode is the slowest dropout time listed.
Mike

 

[life imitates life]

Here's a pretty good pdf with explanation of different spike suppresion
methods.
It lists drop out times for the different methods, for one particular relay
and values.
http://www.kilovac.com/appnotes/app_pdfs/13c3311.pdf
PS. A diode is the slowest dropout time listed.
Mike

Note also that the diode is also the best at suppressing the spike.

Note also that UNsupressed is the only way to get it fast. ALL the
other methods slow it.

ALL of the other methods violate his 'issues' list.

In many cases, that is the most important constraint (the spike).
Note also that all the figures are for 12V relays, which are not in much
use on modern low voltage circuit designs these days.

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.

 

[DarkMatter]

I thought that maybe Phil had assumed a pseudonym, but this guy is less
knowledgeable both in electronics and vile language.

All the pussy that you are is doing is looking to get a rise out of me.

That only amounted to me chuckling at the depth of your stupidity.

 

[amdx]

ISTR an appnote floating about on the net that recommends using a
combination of diode and zener to clamp the back emf, apparently if you
just clamp it with a single diode, the back emf drives a current round the
diode/solenoid circuit which causes hesitant unlatching.

Probably this one?
http://www.kilovac.com/appnotes/app_pdfs/13c3311.pdf
Note, life imitates life finds no wisdom in a ten year old document.
Mike

 

[John Larkin]

ISTR an appnote floating about on the net that recommends using a
combination of diode and zener to clamp the back emf, apparently if you just
clamp it with a single diode, the back emf drives a current round the
diode/solenoid circuit which causes hesitant unlatching.

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.

Reeds are usually pretty lossy, lots of coil resistance, so may be OK
with just a diode. But they are terrible gadgets in general. In my
experience, they are nowhere near as reliable as claimed, and they
bounce/twang forever.

John

 

[life imitates life]

Probably this one?
http://www.kilovac.com/appnotes/app_pdfs/13c3311.pdf
Note, life imitates life finds no wisdom in a ten year old document.
Mike

Show me where that was ever stated, you fact morphing retard!

 

[Archimedes' Lever]

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

You cannot make it open any faster than a non-suppressed version.

 

[amdx]

Show me where that was ever stated, you fact morphing retard!

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