[GreenKeys] Why this stuff matters

[Jeffrey Angus]

On 9/10/19 5:50 PM, John Nagle wrote:
> Good analysis.
Thank you John.
> I've done some analyses like that.
The most important part of these shows the di/dt of current
through the selector magnet.
12v loop 48 mA at 40 mS, 120 loop 48 mA at 10 mS.
Bear in mind, that at 60 WPM, the optimum time is around 11 mS
when the selector sets the first Mark data bit and 33 mS for the
second data bit.
> In particular, a diode across the selector magnet will make it
> drop out too slowly, causing typing errors. 
There's that pesky 4 H coil inductance again getting in the way.
The same reason a snubber is needed to deal with the resultant
voltage spike as the magnet field in the coil collapses when the
current is switched off.

Also, di/dt becomes even more important when you put a second
machine in series in the loop.

Once again, the initial voltage across the selector magnet is equal
to the full loop voltage, until the current increases causing the
voltage to drop across the series limiting resistor.

More of that "pesky science stuff": Time constants.
t = L / R.
With 4 H and 200 ohms, t = 20 mS.
With 4 H and 2000 ohms, t = 2 mS.

As a side note: Loop lengths.
Using 18 AWG wire, it takes a little over 27 miles of wire pair to
equal 2K ohms of resistance. Using 24 AWG wire, it's a little over
7 miles. So worrying about wire size in the typical "in your shop"
loop lengths is sort of pointless.

In case you want to play with this yourself, LTSpice is free and
available at:
<https://ltspice-iv.en.lo4d.com/windows>

-- 
"I am a river to my people."
Jeff-1.0
WA6FWI
www.foxsmercantile.com