[GreenKeys] Snubbing of selector magnet

[Chris Elmquist]

On Sunday (07/21/2013 at 08:37PM -0700), John Nagle wrote:
> On 7/21/2013 7:55 PM, greenkeys-request at mailman.qth.net wrote:
> >> > Mostly the snubber affects what happens on the transition from Mark to
> >> > Space. And quite frankly, unless you've put an enormous value of C in the
> >> > snubber, the Selector magnet is going to pretty much going to let go when
> >> > the loop opens up.
> >> > 
> >> > However, with too small a value, the over shoot and ringing issues are going
> >> > to beat the hell out of the loop keying circuit.
> 
>    Right.  Fortunately, modern optoisolators like the Omron G3VM-401B/E
> can handle sizable voltages, like 320V.  Enough snubbing to get the
> spikes down into that range isn't hard.

yes-- so if we choose an opto that can handle anything as big as the
spike, we probably don't have to snub it at all.

>    There's an argument for an R/C snubber instead of a diode.  A diode
> effectively shorts the selector magnet during the MARK/SPACE transition,
> so the current drops more slowly than merely opening the circuit.

When I mentioned a TVS diode, I was thinking one of the actual transient
surpression devices that have something like 1 or 2 picosecond switch
times--  the thought being that this would shunt the spike down to the
breakdown voltage, effectively turning it from a spike to a small bump
in short (pun intended) order.

-- 
Chris Elmquist