Electrical friction?

With friction (in its ideal form), there is no movement until the applied force reaches some threshold, and then there is acceleration that is proportional to the applied force minus the threshold force.

If we consider the voltage to be the force, and current to be the velocity, then the behaviour is nothing like resistance.

It is, however, like a simplified diode (constant, but non-zero, forward voltage drop) in series with an inductor.

Sylvia.

In the sense that heat is produced as a result of both, it is not "nothing like resistance".

I posted an answer to George a few hours ago, didn't see it.

What I had was ...one model of motion is force is voltage mass is inductance damping is resistance drag from turbulence is ??? stored energy, spring [from memory also height] is capacitance friction is a reverse voltage, subtracted from applied force stiction is a spike reverse voltage

A diode is allegorically more like 'slop' or backlash, because NO force is applied until the friction is overcome. Friction is a bit more like voltage of the opposite polarity to whatever is being applied.

Wait. same, forget it.

me thinking about other types of damping. And I was wondering if there is an electrical equivalent of friction? It seems to me that it could be a c onstant voltage loss. So maybe just some diodes in an LC circuit?

What would be the point of knowing the equivalent to friction in electronic s?

Regards

Klaus

re losses got me thinking about other types of damping. And I was wondering if there is an electrical equivalent of friction?

--- I believe it's called 'resistance'. -- JF

Hi John, friction is actually a little different from resistance. I tend to think about it in terms of simple harmonic motion, so if you'll f orgive the digression for a moment. For an LCR circuit the voltage going a round the loop is, (Forgive me if I get the some signs wrong.. I mostly wa nt to look at the units involved.) C * q(t) + R * dq(t)/dt + L dq(t)^2/dt^2. And for a mechanical system (with no friction, but with a velocity dependen t ?dash pot?. (gamma)) (mass M on a spring k.)

k * x(t) + gamma * dx(t)/dt + M dx(t)^2/dt^2

So both give harmonic motion. And we make the mapping of charge to distanc e, So x ->q velocity -> current

Now friction gives a loss that is proportional to the distance traveled, wh ere as resistance goes as current, which equates with the velocity.

So I was thinking that ?electrical? friction would look like some loss that went with the charge. (Sorry my thinking here is still a bit fuzzy.) I didn?t know how to do that, but the voltage on the cap is proportional to the charge... So I was thinking of a constant voltage loss.. an ?idea l? diode. (And by ?ideal? I mean the freshman level of ideal.. no co nduction till some voltage and then perfect conduction after that.)

If you scroll down this page till almost the bottom, you get to damped osci llations, you?ll see three plots, the first for friction, then velocity d ependent damping (LCR) and the v^2.

The diode LC circuit looked like the friction damping... the decay is linea r in time.

George H. (Sorry for the long relpy)

Yeah something like that... I had two diodes in parallel (one reversed)

George H.

e thread by Vlad about the Johnson noise of magnetic core losses got me thi nking about other types of damping. And I was wondering if there is an elec trical equivalent of friction? It seems to me that it could be a constant v oltage loss. So maybe just some diodes in an LC circuit? > > > > It seems t o work, here?s a spice file. > > Any thoughts or other ideas are welcome. > > (Something that worked at less than 100 V would be nice.) > > [Snip] W hat would be the point of knowing the equivalent to friction in electronics ? Regards Klaus

Ahh, Well I can think of a few reasons (mostly self-centric). I guess the simplist is that it would help with the mechanical and electrical analogies .

Hey! All those analog computers that were used in the past for simulating mechanical systems.. they must have know how to do friction... I'll try goo gling that.

George H.

No. What you are talking about is viscoscity.

If you think of resistance as viscoscity

Current in a resistor is proportional to voltage

Force in a viscous coupling is proportional to velocity

(the classic machanical analog of an RLC uses a dashpot for the resistior)

then the analog of friction is a constant current. A practical constant-current resistor is a pair of depletion fets back-to-back. Above a volt or two, anyhow.

If you put one of these across your resonant tank, you'd get the Christmas-tree (linear) amplitude decay like you see with the diodes... not exponential decay.

I think.

Supertex has LT Spice models of their depletion fets. They are cool parts, on account of the friction analogy.

Nope. John F is correct, the "hanging on" until suddenly breaking away is called "stiction".

"viscoscity" must be some strange thing they do in cities like San Fransicko >:-}

Viscosity...

re losses got me thinking about other types of damping. And I was wondering if there is an electrical equivalent of friction? It seems to me that it c ould be a constant voltage loss. So maybe just some diodes in an LC circuit ? If you think of resistance as viscoscity Current in a resistor is proport ional to voltage Force in a viscous coupling is proportional to velocity (t he classic machanical analog of an RLC uses a dashpot for the resistior) th en the analog of friction is a constant current. A practical constant-curre nt resistor is a pair of depletion fets back-to-back. Above a volt or two, anyhow. If you put one of these across your resonant tank, you'd get the Ch ristmas-tree (linear) amplitude decay like you see with the diodes... not e xponential decay. I think. Supertex has LT Spice models of their depletion fets.

They are cool parts, on account of the friction analogy.

OK thanks, I'll see if I can make those work too.

George H.

-- John Larkin Highland Technology Inc

What's with the non-paragraph, non-wrap?

Using Outhouse Excuse ?:-} ...Jim Thompson

You're using Agent. If you press "H" you will see the headers and see that it's GoogleGroups. Every day it becomes more broken...

core losses got me thinking about other types of damping. And I was wonder ing if there is an electrical equivalent of friction? It seems to me that i t could be a constant voltage loss. So maybe just some diodes in an LC circ uit? If you think of resistance as viscoscity Current in a resistor is prop ortional to voltage Force in a viscous coupling is proportional to velocity (the classic machanical analog of an RLC uses a dashpot for the resistior) then the analog of friction is a constant current. A practical constant-cu rrent resistor is a pair of depletion fets back-to-back. Above a volt or tw o, anyhow. If you put one of these across your resonant tank, you'd get the Christmas-tree (linear) amplitude decay like you see with the diodes... no t exponential decay. I think. Supertex has LT Spice models of their depleti on fets.

Sorry Jim, f*ing google groups I'm afraid. So when I use chrome google pu ts in all these extra lines. Today I tried with IE. No extra lines.. but it globs the whole thing toget her.

George H.

I think he is confusing static friction with classic friction.

Classic friction is a constant force between two moving bodies that is independent of velocity for any v > 0. It is equal to the normal force times K, the coefficient of friction.

Stiction is an initial resistance to motion at V = 0, above the frictional force.

Don't you remember the early EE courses where an RLC was shown to have the same differential equations of a mass-spring-dashpot system? A dashpot is a liquid or air filled shock absorber, a viscous damper.

The oscillation of the spring-mass-dashpot thing decays exponentially, like an RLC. If you replace the dashpot with a friction device, it looks like George's Christmas-tree linear decay.

Viscoscity is force that's proportional to velocity, the analog of resistance. The analog of friction is a constant-current device. You could Spice that as an ideal bridge rectifier driving a current source. I think you could back-to-back two current sources that are qualified to be sink-only, which I think LT Spice allows.

Try an ideal-diode bridge rectifier driving a negative current source, so that the diodes conduct by default. That is an ideal bipolar current sink, the equivalent of friction. Even a real diode bridge and a current source would be good down to a couple tenths of a volt, and go ohmic below that.

Lost interest in the topic?