relays

Although there are windmills that use mecahnical feedback loops to keep the windmill facing the wind (smaller set of blades that turns a gear that aims the windmill until the smaller set of blades is at right angle to the wind and the main plade faces the wind) The windmill itself is just a power supply, not a gain device.

The paragraph you replied to ("Water power was used to lift and turn things a long time before the steam engine was invented.") simply shows a deep confusion between a source of power and a device that modulates that power with a control signal, thus amplifying the control signal. You can add that to not knowing the difference between power and energy and not knowing the difference betweenor between infinite and unbounded.

Some early water clocks had a feedback loop where the water falling below a certain level opens a filling valve (a sort of flap, actually), which raises the level and closes the valve. That would be a water-powered system with power gain configured as an oscillator.

And, of course, there is no rule saying that a human cannot be part of a feedback loop. My use of water is a free-running oscillator; thirst causes me to drink water, which reduces my thirst, which causes me to stop drinking. Not drinking water increases my thirst, which causes me to drink water, etc.

"Hey Ogg! The cave is too cold! Throw some more wood on the fire! ... That's good. you can stop adding wood now."

You keep doing these things, John. You make some fundamental mistake, then come to a silly conclusion that you have to try to defend.

I've been trying to figure out where you made it in this thread, I it looks like this is the place.

The relay requires so many watts to activate. The contacts are rated at so many volts and amps.

The ratio between these two is meaningless. If you do not supply the required energy to the coil, the relay won't activate. So the driver has to be designed to supply that energy.

The relay contacts can handle so many amps at so many volts. Thus, it can supply so many watts to a load. If you draw more than the rated power, the contacts will fail.

The ratio of these two powers is a constant. It does not change with time.

The contacts have the same power rating at the end of a year as they had at the start.

You cannot integrate the ratio between the coil input power and the contact rating over a year and claim the ratio has changed.

Your statement, "There's no limit on the power gain if you're patient enough" is false.

The contacts do not change size, and the coil remains the same. The power gain of the relay doesn't change with time.

The total energy delivered to a load increases with time. But that is not power gain. That is merely your electric bill.

Regards,

Mike Monett

You're missing the math here, but yes, if I get the units right, I certainly can.

Damn, am I up against a girls' school debating team?

Suppose you had to design a domestic electronic thermostat, like the one in my house. It runs on one AA battery.

Assume it will switch 24 volts at 1 amp.

Suppose you had to design such a thermostst. I mean, imagine that somebody trusted you to do it. Suppose you used regular, non-latching relays to switch the furnace and a/c loads. The battery would last weeks, maybe. And you'd be looking for another job.

But my thermostat uses latching relays. I change the AA battery once a year, every Christmas, just to make sure it's OK. For all I know, I could change it every 5 years.

That's a concrete example. Assume some reasonable numbers and do the math.

John

"Have to"? Sorry, you don't get to make the rules for equipment that I design. Let's see here... yup, we've used 3593 latching relays in the last year or so. The cute little Fujitsu surface-mount things. Their average coil operating power must be in the nanowatts. Maybe picowatts.

Never heard of dimensionless numbers? I suppose not.

If switching takes 125 microseconds,

Then there's no reason to ever use latching relays. Boy, is Fujitsu going to be embarassed.

As I said before, if you divide energy by time you get power.

John

message

That'd

I clarified what's-his-name's obsfucation of my use of the word "infinite". Since he can't understand the electricity, he wants to play word games with math definitions. As I read the Wikipedia article on infinity, my association of limitlessly large isn't an unreasonable use of "infinite."

But that's a silly distraction. We were talking acout the power gain of a latching relay. If you guys insist that limitlessly large and unbounded isn't "infinite", ok, that's a definition.

No damage done, and I don't give a rat's ass what you think my "reputation" is. My customers matter, and you don't. They buy my gear, some of which works very well because I chose to use latching relays, and I chose them for their nanowatt coil power consumption.

Only for being wrong, but mostly for being fatheaded about being wrong.

I thought I was being helpful by pointing out the power-saving virtues of latching relays.

I don't need to be relevant. This is a discussion group, not a girls' school debating society. No lady teachers are going to vote you a blue ribbon for your debating skills. Electronics is about designing stuff that works, and getting people to pay you for it.

John

Starbucks pays something like $1.43 per pound for their coffeee.

Tastes like it.

John

Again the basic confusion between power and energy. The latching relay saves energy, and a battery goes dead when you extract too much energy from it. The latching relay does not reduce the power required; if it did you would be bragging about being able to use a smaller battery that puts out less current (the voltage is determined by the chemistry, so less current means less power), not about the battery lasting a long time.

This has been explained to you several times. Let us not forget that you are the person who said that I was unqualified to design even trivial circuits because I made an error about double side band suppressed carrier modulation. Not knowing the difference between energy and power -- and worse, engaging in name-calling rather than learning from your error -- is a far more basic flaw in an electronics engineer than getting DSB-SC wrong is.

BTW, if I designed a thermostat, it would extract energy from the current through it during the on state, from the voltage across it during the off state, or from both. And it would be a lot cheaper to produce than something that has an expensive AA battery in it. This, BTW, *is* in my area of expertise; high volume (up to 100,000 units per hour) low cost consumer goods such as toys -- and thermostats.

See that pot over there? What color is it? Well, how about this kettle? Really? Thae *same* color? Well, how about that!

Next time, try pointing out the energy-saving virtues of latching relays. Unlike the power-saving virtues that you pointed out, the energy-saving virtues *actually exist*.

Never heard of a dry dock? How much energy does it take for one guy to open a gate, and how much energy does it take to lift a ship?

Ever hear of hydraulic mining? It takes a small effort to aim a nozzzle which rather precisely moves a good chunk of mountain.

A sluce gate that controls a water-wheel mill has a lot of power gain.

Most water-powered systems share the property of the latching relay, in that the control input can be intermittent but produce a sustained power output. If you were the guy operating the gate for the dry dock or the mill, you'd appreciate this... a little work, now and then, does a lot. The engineering units involved in these situations seem to confuse a lot of people. The energy output is the time integral of the control input.

I designed the timing/trigger and beam modulator systems for the world's biggest laser. And some fraction of the world's ICs are exposed by a MOPA eximer laser controller I designed. This week it's Fourier Transform Mass Spectroscopy and picosecond-exposure ICCD cameras and some sonar stuff. And you think I can't tell the difference between power and energy?

Stop being a silly ass and go design something. Post a schematic.

John

That's not power gain, John.

The power drawn by the load is constant.

The ratio between the power required to activate the relay, and the power delivered to the load is constant. There is no power gain, as you claim.

The total energy consumed by the load increases. The watt-hours increases, which is what you want.

But that is not power gain. That is energy gain. Your definition is incorrect.

And you can quit with the silly insinuations. That is a poor debating technique.

Mike Monett

False. The energy delivered to the load is the time integral of the flow and pressure at the output.

The energy consumed by the control is irrelevant in this case.

Mike Monett

OK, estimate the power dissipation of the non-latching and latching relay coils, averaged over one year.

I'm serious: make reasonable assumptions and post some numbers. No other time frame makes any practical sense, since the average power used is what kills batteries.

Show us the numbers.

If you're going to lecture, check your facts first. AM is pretty simple stuff, so lots of people will catch you if you don't.

John

I'm defining it as (load power) / (coil power) averaged over some practical operating interval. You know, the thing that will matter in an actual piece of equipment.

I'm wrong fairly often, but not about stuff as simple as this. When designing really complex systems, especially in the conceptual stages, it makes sense to be wrong now and then. In posting to Usenet about simple stuff, it's easy to check your facts, or simply don't make unqualified statements about stuff you don't understand.

So, what's your equation for the power gain of a latching relay?

John

Shape up or I'll cut off your beer.

John

Hey, where's that math?

This gets better and better. Now relays have no power gain. Short Tyco stock before word gets out.

My definition is a definition. It's apparently not the same as your definition, but it's not incorrect.

I define power gain as average load power out divided by average control power in. What's your definition?

Check the record for the trajectory of silly insinuations. After you post your definition.

Cool. Post a schematic. I assume it will work in a millivolt thermocouple loop, too.

John

Looks like I'll have to buy Safeway to cut off your beer.

John

Well, with a multiplicative constant of course. The "gain" thing. Hell, throw in the constant of integration while you're at it, if you need to be didactic.

In that case, you don't care about the thing called "gain." So go away.

John

As soon as you include time, you are no longer talking about power, you are talking about energy.

You state this in the above post: "As I said before, if you divide energy by time you get power." That is correct.

But you cannot make a statment about power gain over one year. That is meaningless. The relay does not change. It requires the same energy to activate, and is capable of handling the same power. So the ratio is constant.

What you are really talking about is energy gain. That can be very important.

So simply revise your statement and change it from power gain to energy gain.

Then you will agree with your own definition.

Mike Monett