Yes, please do. I can find no such computation.
John
relays
Nov 13, 2008
360 Replies
OMG! I have finally made a mistake. I admit it.
JF got it *close* on his *fourth* try.
I am sorry, truly humbled. Imagine confusing 3 with 4.
John
No, you don\'t.
As usual, you\'re just being a troublemaking, sardonic ass who doesn\'t
know when to quit.
Remember this? snipped-for-privacy@4ax.com
Are you suggesting that the math is wrong, as GM and JKK seem to feel, or do you want to quibble over whether an unboundedly large real number can be called "infinite"?
If the former, correct it.
If the latter, get a life.
The "coward" claim is absurd, desperate even. If you have any guts, try to keep your cool and stick to facts... and check your work before you post. Good engineers do that.
But "bourgeois"?
John
It\'s hardly quibbling when you flat-out refuse to admit that a latching
relay _can\'t_ have infinite gain regardless of how large the numerator
becomes or how small the denominator becomes, since, by definition, it\'s
a mathematical impossibility.
You, who claims to be _so_ mathematically astute, should respect that
and stop the little: "Well, it\'s real big number so it\'s infinite to me"
song and dance.
If the word is meaningless, I confiscate it and use it, in daily work, to mean "big and unbounded." It's my definition.
I never claimed to be mathematically astute. I do calim that I check my arithmetic and ususlly wind up with meaningful, correct-units, engineering-accuracy results. Not because I'm smart, but because I'm careful.
Because I see no mistake.
Facts are facts. I wish the innumerates here would get over their fear of math; everybody would be happier.
What a heap of words, cut-and-paste dictionary silliness, all to avoid a little arithmetic. Again, if you don't agree with my math, correct it.
John
In you find a properly defined word meaningless, then that\'s your
failing, and if you want to redefine it in a peculiar way and use it in
daily work, then that\'s fine.
But this isn\'t your daily work, and using your peculiar definitions
where people are used to the rather more rigorous formal definition can
only result in hostility, particularly if you keep insisting that your
definition is better.
You calculated the gain of a simple non-latching relay as 78, 280,
1418, and finally 2836. But you haven't disputed my calculation of the power gain of a latching relay, much less tackled this simple math yourself.Quit yapping and do some math.
John
--- This isn't about me not being able to do arithmetic, John, and you know it. What it's about is your dishonesty in trying to get out of owning up to your mistakes by dancing around like a madman using all sorts of diversionary tactics in order to divert the focus away from yourself while trying to make everyone else look like the heavy.
Goodbye.
JF
Cool. I can shut up any number of people here by asking them to do simple math.
John
And, I might add, he specified *power* gain, which means that the the numerator isn't all that large! It's the *energy* gain that gets larger without bounds. The *power* gain is finite and fixed.
JL -- a person who likes to call people "innumerate" when they get the math right and he gets it wrong should also stop his little "Well, I can convert power plus time to energy so they are the same thing" song and dance. They are *not* the same thing. I can convert distance (miles) plus time (hours) to speed (miles per hour) but that does not mean that speed and distance are the same thing. How many MPH are there between California and Arizona? How fast does a horse run in units of miles?
Guy Macon
More blather. No numbers. All hat and no horse.
John
JF. Walk away, JL will never admit sloppiness, let alone error. Fortunately Google preserves this all for posterity (so far).
Good advice. I am now killfiling this thread.
Guy Macon
As noted, but apparently never understood, is that I specified average power gain over some stated time interval. In that case, energy gain and power gain are identical, since both the numerator and denominator of Eout/Ein can be divided by the time interval, to produce the numerically identical Pout/Pin.
So the latching relay power gain goes up as 1/(coil duty cycle). What's wrong with that? Couldn't be much simpler. You guys have got your panties in a bunch over a simple, whimsical observation.
John
???
What you originally stated was:
"We use a cute little latching relay that will pull in in 500 usec,
using about 125 microjoules, just once. After it\'s pullec in, it will
easily dump 50 watts, which is 50 joules per second. Measured over one
second, its power gain is 400,000. Over one year, it\'s 1.3e13. There\'s
no limit on the power gain if you\'re patient enough."
Such being the case, and considering the equal input and output times
you now describe as being necessary to determine power gain, that 50
joules in one second will turn into 25 millijoules in 500 microseconds,
so the power gain of the relay is:
Wout 2.5e-2J
PG = ------ = ---------- = 200
Win 1.25e-4J
>So the latching relay power gain goes up as 1/(coil duty cycle).
>What\'s wrong with that?
I did not specify that the load observation time and the coil actuation time were the same. I stated the coil drive ("500 usec... just once") and load observation times (1 second and 1 year) exactly. To use a latching relay efficiently, you only pulse the coil for as long as it takes to flip the contacts. But the resulting load stays on or off indefinitely.
More mathematical nonsense. Why *do* you keep doing this, flinging numbers around until you get something that satisfies your expectations?
If you reduce the observation time to 500 microseconds, sure, the power gain of the latching relay is the same as it would be if the relay didn't latch. IF you ignore the contact closure delay. Actually, the power gain here is closer to zero, since the contacts may not even be closed at the end of the 500 us coil pulse... they are still in flight.
But the point about the latching relay is that it continues to deliver load power long after you quit driving the coil. My calculations are correct for the time intervals I stated. And that's why latching relays are used in all sorts of low-power applications, like battery-powered thermostats; they have a huge power gain when driven at low coil duty cycles.
John
Short attention span?
Sure you did. Just look above and you\'ll see that you wrote:
"As noted, but apparently never understood, is that I specified average
power gain over some stated time interval. In that case, energy gain
and power gain are identical, since both the numerator and denominator
of Eout/Ein can be divided by the time interval, to produce the
numerically identical Pout/Pin."
It sure is if it makes the power gain zero as measured over 500 usec. Which it will, with the Fujitsu relays I mentioned. If you're willing find a couple of numbers and divide, but aren't interested in whether they mean anything in real time, go for it. My example calculation doesn't have this problem.
Now you are arguing definitions. Of course any measurement or calculation of "power gain" must have a context. I have defined the power gain in a manner that's useful in real equipment, where low duty cycle of coil actuation confers real advantages. If you have your own definition, which never considers time, be happy with it.
We use these relays, and low average coil power consumption matters to us. In fact, it makes several of our products possible.
John
It\'s not an issue because it\'s understood that there can be no gain
_until_ charge starts flowing through the load, making the time of
flight immaterial.
What _does_ make sense is that the observation interval of the output be
the same as the width of the input pulse, once charge starts flowing
through the contacts, which makes time cancel out of the equation and
then a simple division will yield the power gain of the relay.
Join the Discussion
Have something to add? Share your thoughts — no account required.
Didn't find your answer?
Ask the community — no account required