[Prev][Next][Index][Thread]

Re: That secondary behaviour, E-Tesla5, and Corum's thing...

Hi Ken,

	I did it!!  I guess I was just really lucky, but I grabbed a single sine
input pulse and the resulting output from a scope probe connected directly
to the top of the coil.

http://users.better-dot-org/tfritz/site/misc/Ken05.gif

A close up:
http://users.better-dot-org/tfritz/site/misc/Ken06.gif

I also ran MicroSim for a transmission line model of the same effect:
http://users.better-dot-org/tfritz/site/misc/Ken07.gif

If the output signal starts 1/4 wave after the input signal, looks like we
have a transmission line as shown in Ken07.  if it starts sometime earlier
as in Ken03, we have a lumped system.

http://users.better-dot-org/tfritz/site/misc/Ken09.gif

Shows a close up of my test along with simulations of a transmission line
and lumped model.  The output waveform is obviously starting far before the
90 degree point.  However, it is not quite in the lumped part either but it
is difficult to see exactly.  I also worry about the scope probes loading
effects.  I can get a 2.5/100Meg  100X probe from work Monday that should
help.  The resonance of the coil was retuned for the probe being in place
and all that.  It dropped from 350kHz to 214kHz suggesting the capacitance
went from 9.36pF up to 25.0pF.  The probe is supposed to be a 16.6pF/10M so
that works out but the wire to the probe probably messes with stuff too.

Of course, it still gets worse :-))

I ran MicroSim models with the probe inserted.

http://users.better-dot-org/tfritz/site/misc/Ken10.gif

Yep!  The probe makes the lumped model later and the T-line model sooner
making it difficult to tell them apart experimentally. The probe impedance
is almost making this experiment impossible ;-(

The models are at:
http://users.better-dot-org/tfritz/site/misc/Probe.gif
http://users.better-dot-org/tfritz/site/misc/test.sch

I may be able to reattenuate my plane antennas to get a better signal.
They don't load the secondary in any way but I'll have to be sure I don't
mess up their bandwidth in the process.  This may be a key lumped vs.
T-line experiment if we can get the details worked out.

So Ken,  Are you nuts yet! :-))  Perhaps the "slow" refers to how long it
takes for this stuff to drive one insane! ;-))

Cheers,

	Terry




At 11:03 AM 04/15/2000 -0700, you wrote:
>Terry & the rest-
>
>More wonders of the computer age!  I printed out your Web image & have it
>in front of me & it looks oh-so-familiar.  Except...the image I got on my
>scope just this a.m. doesn't have the fuzz--hah!, hah!, hah!
>
>What I did:  My ancient Wavetek 115 will emit a single sine cycle, at the
>>end< of its gate input from the HP 8005.  I made it do that, then synced
>the Tek 7904 from the beginning of the gate signal.  Thus, by choosing a
>long enough sweep time, I could get the 1 cycle event-time fully to
>appear on the screen.
>
>Next I adjusted the Tek's 7B92A dual time-base plug-in to display that
>whole 1-cycle time & a bit more all across the screen.  I displayed a)
>the input 1-cycle to the bottom of the secondary and b) the signal at the
>top via the 10:1 probe.  I left the probe galvanically connected to avoid
>phase shift due to unknown series capacitance were I to just lay it near
>(your point!).
>
>Now here's the fun part:  My 7904 incorporates a 7D11 digital delay
>plug-in.  It can make a little pulse & an intensity change appear on the
>screen, digitally positioned anywhere along the display.  So, I set that
>up & with it made the following measurements at resonance:
>
>A.  Commencement of the 1 input cycle, 0 us, normalized.
>B.  1st input peak, 2.7 us.
>C.  Input z.c. (zero crossing), 5.6 us.
>D.  2nd input peak, 8.3 us.
>E.  Input end, 11.0 us.
>
>F.  1st output z.c. (each z.c., eyeballed by choosing the place on the
>wave equidistant between 0 or a preceding peak and the following peak),
>3.2 us.
>G.  1st output peak, 5.3 us.
>H.  2nd output z.c., 7.8 us.
>I.  2nd output peak, 10.5 us.
>J.  3rd output z.c., 12.9 us (the coil is, of course, continuing to
>ring).
>K.  3rd output peak, 15.5 us.
>L.  4th output z.c., 18.0 us.
>
>>From the above:
>
>1.  Successive output vs. input time delays, z.c to z.c & peak to peak: 
>F-A=3.2, G-B=2.6, H-C=2.2, I-D=2.2, J-E=1.9.
>
>2.  Free-ringing 1/4-cycle times, K-J=2.6 & L-K=2.5...& on & on.
>
>3.  Input 1/4-cycle times, B-A=2.7, C-B=2.9, D-C=2.7, E-D=2.7
>
>So what's it all mean?  It's straining at gnats, is what I think.  Sure,
>there's an itty-bitty extra delay right at the beginning, & then the
>output wave seems to speed up a tad just at the end of the input cycle
>(anticipating the end, do you suppose?--a new phenomenon previously
>unknown to science!).  Of course, I use not-necessarily-finely-calibrated
>equipment but still, one can compare measurement to measurement pretty
>accurately.  I think I find nothing of interest here:  The output wave is
>going to settle down to being 90 degrees behind the input pretty damn
>soon & after that, that's always the same.  I don't see anything that's
>"slow".
>
>Why am I still doing this??  I'm supposed to be retired!  Someone save
>me!
>
>Ken Herrick
snip...