relay coil inductance

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?

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

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

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

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

FET's are sub-nanosecond

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

the

just

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

=A0 =A0 =A0...Jim Thompson

ed.

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.

Put a known capacitor across the coil and grid-dip it?

Jim

Nope, anything else you need help with? Mike

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.

...Jim Thompson

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.

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

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.

=A0 =A0 =A0 =A0...Jim Thompson

lated.

e is

or

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 f*ck like you.

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

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

John

I wouldn't be too fast to hang my hat on it. The relay coil already has a self-resonant frequency due to stray capacitance in the winding. Adding capacitance would lower the resonant frequency. Doesn't make sense, but maybe it's a good capacitor versus a lossy distributed one?

...Jim Thompson

It's not beyond the realm of possibility.

boB