On Tue, 17 Jan 2012 14:47:08 -0700, Jim Thompson
wrote:

On Tue, 17 Jan 2012 11:23:48 -0800, John Larkin
wrote:

On Tue, 17 Jan 2012 11:44:48 -0700, Jim Thompson
wrote:

On Tue, 17 Jan 2012 10:33:20 -0800, John Larkin
wrote:

[snip]


John, You really do need to see a mental health professional.

Another unsupported claim.

Nope. By your personal illustration :-)



Hey, why not arrange as many led's in series as you want, to get the
lighting you like? After all, we do have a current source.

John

But we don't... the current tails off significantly as you raise the
output voltage. If you'd bother to run the Magdowski model into real
loads you'd know that... but I don't think you know how to even run
LTspice.

I use LT Spice, probably not as often as you do,

I rarely use LTspice, I use it for running netlists generated by other
programs that lack the speed-up needed for things like crystal
oscillators. But LTspice lacks the presentation-level graphics needed
for design reviews.

bacause I don't
design by fiddling.

You know I don't design by fiddling. Why do you keep persisting on
that lie? How many times have I said I design by pad and paper, then
_verify_ with PSpice. One might assume your lie is intentional to
kill off business that might come my way otherwise. My lawyer calls
that tortuous interference. Knock it off or meet my lawyer.

Oh dear, he's given up on a fist fight, and on a road race, and is now
siccing his lawyer on me. Whatta man.



You always have component values available because
you do "design" by Spice fiddling.



...Jim Thompson

The corner frequency, the speed where the
constant-current-versus-speed region starts, will obviously move up as
the load impedance increases. But it's certainly not a constant-power
device in general. At any given speed, it's an AC voltage source
behind an impedance, and all the normal circuit theory applies.

You clearly haven't analyzed, or Spiced, the model. I have.


If, at some speed, you plop the load impedance at the peak of the
power transfer curve, sure you'll be on the flat spot. Nothing unusual
about that, but that's not the general case. The point is that, at
reasonable bike speeds, there's more power available than the lamps
normally consume, and you can extract it if you match better. If
that's not true, your buck switcher makes no sense.

If you disagree, say something substantive.

John



Show us the "design" with values, so it can be verified.

If you keep up this nonsense I will sue you for tortuous interference.
If you think not, try me.

If you don't want people to think that you are incompetant, say
something substantive about the power transfer issues here.

John