Re: Einstein's Doppler equation wrong?
From: Androcles (dummy_at_dummy.net)
Date: 10/27/04
- Next message: AllYou!: "Re: A definition of time."
- Previous message: Ken S. Tucker: "Re: The GR metric in LeSagian Exponential Form..."
- In reply to: sal: "Re: Einstein's Doppler equation wrong?"
- Next in thread: sal: "Re: Einstein's Doppler equation wrong?"
- Messages sorted by: [ date ] [ thread ]
Date: Wed, 27 Oct 2004 11:28:21 GMT
"sal" <pragmatist@nospam.org> wrote in message
news:pan.2004.10.27.03.26.12.109147@nospam.org...
: On Mon, 25 Oct 2004 13:58:50 +0000, Androcles wrote:
:
: >
: > "sal" <pragmatist@nospam.org> wrote in message
: > news:pan.2004.10.25.01.38.45.967816@nospam.org...
: > : On Sun, 24 Oct 2004 14:10:18 +0000, Androcles wrote:
:
: > : Zeros between the "bumps" in the waves...
: >
: > Sal, a mass hanging on a spring doesn't do anything by itself...
:
: I'm obviously not going to convince you that the fields are in phase
: unless I can show you that the energy has "someplace to go" at the
zero
: crossings. Since I can't do that (those are zeros in the energy --
the
: distribution is lumpy) I might as well give up on this one.
Yep, you might as well. Another reason you might give it up for is the
'j' in Maxwell's equations. It seems I agree with Maxwell, you and
Andersen do not. Ask Sue about, you don't listen to me.
:
: On the other hand, you are certainly not going to convince me that
the
: fields are NOT in phase,
I know I can't. I couldn't convince moortel that a stopped clock has
zero frequency either, he has an equation that says f = 1/t so a
stopped clock has infinite frequency. Relativists are like that, they
think math drives reality instead of describing it. It's a waste of
time talking to them.
unless you can show me how you deal with the fact
: that curl(E), curl(B), dE/dt, and dB/dt are all at maximum magnitude
at
: the point where the fields cross zero.
When you integrate dx/dt, you get the distance travelled, x, in the
time, t.
Has it occured to you that you need to add how far you've already gone
to know where you are? This is called "adding a constant".
When you integrate dE/dt, you get the voltage gained, E, in the time,
t.
Has it occured to you that you need to add how much voltage you had to
begin with?
That is how I deal with it. It is also how Maxwell dealt with it.
: Since the curl of each is equated
: to the time derivative of the other,
Whatever it had to begin with is lost when you examine only the
change, dE/dt.
: the maxima must occur at the same
: point in space.
If and only if both B and E begin at zero.
: Since I don't think you're going to convince me otherwise
: (or convince me Maxwell's equations are wrong), you might as well
give up
: trying to convince me I'm wrong about this (actually, I think you
already
: have).
I find no fault with Maxwell's equations.
Say "uncle".
:
: > : * * *
: > :
: > : I liked the bit about the SAT resistor, by the way. By the time
I was
: > : starting out that sort of thing was really frowned on -- if the
board
: > : required that kind of "elf work" in order to function then the
design
: > : was considered NG. But that was digital, of course; I've never
even
: > : seen the inside of an analog computer.
: >
: > It was the RC of a timer that had to meet better than 0.1%
accuracy,
:
: Omigod you don't mean you actually used one-shots, do you?
In the early 70's, yes, of course we did. How long would you wait
before you
were certain the "weight-on-wheels" signal was solid enough to retract
the undercart, and how would you time it? It isn't good practice to
let the captain
or first officer bring the gear up before you are off the runway. Not
many people realize it, but to fly a commercial plane all the captain
has to do is taxi
to the runway, dial up the destination and press "go". We put things
like manual controls in to pacify the passengers and give the crew
something to look at. After the plane has landed (weight on wheels) he
can play with the air-brakes and drive to the gate. He is permitted to
override certain functions, of course, so I do exaggerate, but not all
that much. I don't recall what that particular one-shot was used for,
it was too long ago and I was a rookie, as I said.
:
: You can get fired for using those things these days. Everybody
knows that
: they don't work, evidence provided by the flights of the Concorde
: notwithstanding.
:
:
: > and the design engineer was not comfortable with resistor,
capacitor and
: > voltage tolerances combined.
:
: I'll bet he wasn't! :-) I'm surprised he didn't specify that the
circuit
: be packed in ice or otherwise temperature-controlled while in use,
as well.
Actually it was temperature controlled. Passengers won't fly if they
see the pilot wearing a leather jacket, scarf round his neck, helmet,
goggles and a parachute on his back. It makes them nervous, and the
flight attendants aren't too happy about it either. Commercial planes
have a flight deck, military planes have a cockpit. That doesn't mean
the electronic components are not mil-spec, though.
:
: > The "tuning" was accomplished by a parallel (SAT) resistor, so the
: > circuit did actually work but the time interval was too great. You
would
: > probably have designed it with a potentiometer to minimize
production
: > costs.
:
: Not if I've understood you so far -- I wouldn't trust a trim pot to
hold
: its setting well enough...
Nor did Marconi Avionics.
:
:
: > 10k - 1% capacitor
: > |----/\/\/\/\----|____o---| |----o
: > |----\/\/\/\/----|
: > SAT - perhaps 470k
: >
: > Remember this was for Concorde, not a mass-produced domestic toy.
2.5%
: > of altitude at 40,000 ft is 1000 ft. Lives matter, elf-work was
the
: > ideal solution. I wasn't permitted to solder any components, that
went
: > to a rework wireman who had an inspector sitting beside him.
:
: Yeah, I'm sure they didn't want "flux fade" bringing down the plane
in six
: months as a result of imperfect elf work.
Everything has to meet specification, and tolerances were tight. One
antiquated piece of test equipment used to simulate self-test by
stuffing
step voltages into the machine was giving us a lot of grief, it
operated on
old telephone relays (yeah, solenoids) and was nicknamed "the chunter
box"
as it clunked its way through the test. I was asked to fix it. I
redesigned it instead, using TTL and one 256 x 8bit EPROM. 1 MHz
clock, counters, comparators, LEDs and a liquid crystal display. That
was state-of-the-art then. 4 bits were used for time and 4 bits for
signal address. With the address counter set to zero (reset) a signal
was decoded from its address and the timer started. When it matched
the time preset in the lower 4 bits the comparator fired and
incremented the program counter to the next step.
The output of the unit under test (analogue) was then compared with a
fixed voltage and give a no-go (green LED). If it gave a go, the fail
light came on
and the process stopped. This was the low end tolerance. the program
counter went on to the high end tolerance and looked for a go. Thus
the signal was bracketed in time for the correct voltage. I was
failing perfectly good computers that were passing all other tests.
Why? Because my timing
was to the nearest microsecond, and the test spec had been written for
one-shots and tolerance in the test equipment had not been considered.
I was then asked to broaden my tolerance, the political ramifications
of changing the written test procedure were enormous, it had already
been approved based on the old chunter box. So I did. Ces't la vie.
:
:
: > The heart of the analog computer had become the op-amp on a chip,
8
: > legs, but while I was there the Chinese placed an order for some
old
: > design (BA 1-11 ?) planes and the original designs were brought
out of
: > mothballs. These used transistors that were obsolete, and had to
be
: > specially made. It would have been cheaper to redesign, using
latest
: > technology, than to retrain the technicians who were having all
kinds of
: > difficulty testing and debugging an operational amplifier on a
card
: > instead of replacing a chip, but politics of the era required the
: > non-release of current technology.
: > The digital equivalent would be building a half-adder out of
separate
: > NAND gates.
: >
: > Digital computers were just coming in, and I recall a double
: > Euro-card (about the size of two motherboards back to back) that
was the
: > ALU (arithmetic and logic unit) of the processor. The register
(yes,
: > just one for data) was on a separate card,
:
: I've seen that technology. My roommate had an old PDP - uh - darn,
now
: I've forgotten which one. Anyway, a PDP-something which he kept in
the
: corner, and every register was on a separate card.
Yep, and a wire-wrapped back-plane. To boot from disk, write a program
on the binary switches on the front panel. Press 'enter' for each
instruction. That switch was different to the others, you had to push
it up instead of down to avoid accidentally entering an incorrect
instruction to early. It also incremented the instruction address
register. Mind you, the PDP-11/35 and -11/55 was far more advanced
than the one I'm talking about here, which was based on an Elliot 9000
that I doubt you've ever heard of. Google doesn't seem to either.
:
:
: > all instructions to
: > AND/OR/ADD/NOT were between RAM and the register, through the ALU
and
: > back to register or RAM.
:
: And _that_ sort of design didn't go completely out of style until
RISC
: processors came in.
Physically combining an 8-bit chip and a 4-bit chip to create a 12-bit
register is a tad ancient, but that's how it was :-)
I recall a CMOS counter, resettable but non-settable, internally
cascaded.
If you clocked it at 1 MHz, it wouldn't overflow for 100 years.
CMOS would baulk if the frequncy was too high.
I can't imagine why anyone would design that, but someone did. :-)
:
: > The D-to-A and A-to-D converters had their own cards, and if the
program
: > didn't complete it's cycle in 30 milliseconds the instruction
register
: > was reset, equivalent to a reboot. The program was stored in
E-PROM. No
: > mass-storage was used. I designed the test procedures for 80% of
the
: > cards in 6 months, then I left to go to flight simulation with a
: > different company.
:
:
: --
: I can be contacted through http://www.physicsinsights.org
:
- Next message: AllYou!: "Re: A definition of time."
- Previous message: Ken S. Tucker: "Re: The GR metric in LeSagian Exponential Form..."
- In reply to: sal: "Re: Einstein's Doppler equation wrong?"
- Next in thread: sal: "Re: Einstein's Doppler equation wrong?"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
