Simulating long cable/wire lengths

btmull@yahoo.com wrote in news:e79bbf17-d127-41a5-ad82-
22a8a7fbff3f@d21g2000prf.googlegroups.com:


Lork Kelvin derived the cable equations for the TransAtlantic Cable.  
Google up "Cable Equations" and see what happens.  You'll likely find many
Neuroscience and biophysics links.  If you can find a paper by Rall, it
will probably cite the engineering papers.

--
Scott
Reverse name to reply

Scott Seidman wrote:

Wasn't that done by Michael Pupin?

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Wikipedia on "Transmission Line"

"Mathematical analysis of the behaviour of electrical transmission lines
grew out of the work of James Clerk Maxwell, Lord Kelvin and Oliver
Heaviside. In 1855 Lord Kelvin formulated a diffusion model of the
current in a submarine cable. The model correctly predicted the poor
performance of the 1858 trans-Atlantic submarine telegraph cable. In 1885
Heaviside published the first papers that described his analysis of
propagation in cables and the modern form of the telegrapher's
equations."


--
Scott
Reverse name to reply

Scott Seidman wrote:

   ...


I guess I was thinking of loading coils. They probably were in the first
transatlantic telephone cable.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

You might get a kick out of the Rall Handbook of Physiology Chapter, if
you can find it: Core conductor theory and cable properties of neurons
W Rall - Handbook of Physiology, 1977

He used Cable Theory to model just about all of the passive properties of
the axon and dendritic tree!  This was my first exposure to hyperbolic
trig functions.  I knew what they were, and I knew there were keys for
them on my calculator, but I had no clue what they solved.



--
Scott
Reverse name to reply

Scott Seidman wrote:

I'm curious about how closely the model conforms to reality. Signal
speed in real neurons goes up, linearly I think, with diameter. Much of
the early work was done at Woods Hole because of the availability there
of the large giant sea squid that made experimenting easier. In higher
animals, signal speed is increased greatly by the myelin sheathing that
surrounds the axon. The sheathing is not continuous, and the nerve
impulse jumps from gap to gap without needing to propagate chemically
inside. The whole subject is fascinating.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

I teach a physiology course specifically aimed at engineers who know
circuits and diffeqs.  

For unmyelinated axons (vel. proportion to sqrt(radius) ) and the
dendritic tree, the equations do extremely well (assuming you can measure
the parameters accurately).  The equations still work well for myelinated
axons ( vel proportional to radius, as you've said), but the model gets a
little complex, as the membrane resistance can't be considered constant
along the length.  There are some assumptions made at the endpoints and
bifurcations, but they do OK.

The nerve impulse itself isn't entirely regulated by this, but by active
pores that open and shut based on membrane voltage.  A little current
goes in, and if the membrane voltage exceeds a threshold, then a positive
feedback process starts on that patch of membrane.  The current then
propagate passively down the axon, following the cable equations, and if
all is working, it brings the adjacent patch above threshold, and the
process continues down the axon (in both directions, actually, if you
inject current into the middle of the axon!)

The Nobel work by Hodgkin and Huxley was done on squid giant axons (if it
were done on giant squid axons, there wouldn't be much to work on, but
I'm always making that spoonerism myself).  The nifty thing about that
prep is you can actually run wires down the axon, or you can roll the
goop out with a squeegee and substitute your own goop.


--
Scott
Reverse name to reply

Scott Seidman wrote:

While this groundwork was going on ay Woods Hole, the foundations for
continental drift were being assembled at Lamont-Dougherty, and both
were frequently written up in _Scientific_American_. Those were
interesting years, even for a bystander.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

On Jan 4, 8:31 am, btm...@yahoo.com wrote:

Easiest is to buy 4000ft of cable and hook it up.
ed

On Fri, 4 Jan 2008 10:13:55 -0800 (PST), Ed Prochak


Indeed. Four of these, for example:
http://h30094.www3.hp.com/product.asp?sku 85518&jumpid=ex_r2910_performics/k17401/DDI%20Link

Heck, for the small difference, why not get five and test with some
margin?

Best regards,
Spehro Pefhany
--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

I agree, simple is better.  For any alternative you have to design,
build, and verify; then justify with papers.
It might be reasonable if the test fixture was going to be
manufactured but not in a one-off.  Try ebay, you might make a profit
when you sell it back.

Ray

This is going to be ugly because you have to unwind the cable. Otherwise the
capacitive coupling would be the predominant effect.

Vladimir Vassilevsky
DSP and Mixed Signal Consultant
www.abvolt.com

btmull@yahoo.com wrote:

There are line simulators for just such tests available from test
equipment makers.  OTOH, 100 m is pretty short.  Why simulate at all?
Just put up a 100 m spool of cable in some unwanted corner and be done
with it.