# [sf-lug] SF-LUG meeting notes for Sunday December 3, 2017

Michael Paoli Michael.Paoli at cal.berkeley.edu
Wed Dec 6 20:35:17 PST 2017

```~600 VAC, not DC.  Don't f*ck with sh*t like that unless you very much
know what the heck you're doing - can be fatal if you f*ck up.
I used a power transformer from an old not solid state color TV
console has various higher output voltages to power tubes and B++
voltage (go look it up - pro'lly ought be on Wikipedia).
Anyway, I don't recall precisely what value resistor I used,
I did calculate to be high enough on resistance to make the current
quite low, and also to not overheat the resistor at a dead short
at full voltage.  So ... let's see ... was probably 1/4 W resistor,
AC voltage - I don't remember, but likely something in the ballpark
of 600 VAC RMS.

And AC would be *slightly* preferable - higher peaks (more likely
to achieve ionization).
And lowish (e.g. 60 Hz) frequency also relatively preferable - give
contacts potentially half a chance or so to "rest" between half cycles.
With 60 Hz, if there was anything especially "interesting" happening
at the contacts (actively arcing / burning anything off), that would also
be more likely to be detectable (seen/heard/felt) than DC - so better to
be able to monitor progress/activity - at least if used/needed for more
than some second(s) or less.

So ... W=A*V, A=V/R,
W=V^2/R
R=V^2/W
R=(600V)^2/(1/4)
\$ echo '600*600*4' | bc -l
1440000
\$
Resistor tolerance probably 20%, so ... standard values ...
\$ echo '1.8*.8' | bc -l
1.44
\$
Bare minimum would be 1.8M - but that's before we take into account
accuracy of the voltage measurement - so I probably went with something
>> 2.2M that was handy and kept the power low.  Figure max power at the
switch contact points would be with 1/2 the voltage on the resistor and
the other half on the switch (if/when it happened to match impedance - at
the applied voltage) ... but I didn't want 1/4 or 1/8 W or anything near
that possibly concentrated between the switch contacts - lest they
potentially get spot welded together.  So I probably grabbed something
quite a bit higher in resistance, that would still allow enough current
to potentially and likely burn off whatever was preventing the contacts
making electrical contact, with negligible risk of welding them
together.  So I likely did something in the 5.7M to 22M range.
I'll leave max power to the switch contact points and max current for
those resistor values as an exercise.  ;-)
The basic methodology was close the contacts (mechanically - but they didn't
close electrically), apply high voltage limited current,
release the contacts, remove the voltage and retest the switch - if it
worked, done, if not, maybe retry ... possibly slight more
power/current/voltage if it didn't work ... and if it's stuck closed,
already did too much power/current/voltage - and time to buy a new
switch.

Also, 600 VAC - that's what I recall, at least approximately (was
about what I was aiming for - enough to arc a very small closed but not
conducting gap with approximately air and some presumed metal oxides
between).  I may have done something else to step up the voltage if I
didn't get ~600 VAC straight from that transformer's output - using
whatever highest voltage output taps it offered.  But I may be
misremembering.  I may have possibly used an underdriven flyback
transformer, or an ignition coil (yet another type of transformer).
In any case, I recall having something around 600 VAC, and I'm about 80%
certain it was straight off the highest voltage taps of power transformer
from and old non-solid state color TV console chassis.

Random mini-fact: rough rule of thumb for dilectric (electrical breakdown)
voltage of air: about 10,000 V/in.  This can however, be altered by
various factors.  E.g. gas mixtures, temperature, humidity, dust.

Firefighters actually have guidelines when using water on energized
electrical equipment - what type of water delivery (fog/spray/stream),
what voltage, and required safe distance.  F*ck up and you're dead.

Hmm, maybe exactly what I did (or nearly so) might even be in my old
hardcopy hardware logs.  Have about 100 or 200 pages of such "data" ...
before I switched to going electronic (with regular backups accessible
in various ways ... much more searchable, also much more convenient to
record diagnostics and precise digital data information).
Alas, ye olde hardcopy - not very searchable ... but does still function
when (all the) computer equipment is down (like when I really only had
exactly and only one computer available to me).

> From: jim <jim at well.com>
> Subject: Re: [sf-lug] SF-LUG meeting notes for Sunday December 3, 2017
> Date: Thu, 7 Dec 2017 03:43:13 +0000

>
>     600VDC? with a 10MOhm resistor. Great
> tip!
>     Where'd you get the 600VDC? Not from
> an everyday receptacle.
>
>
> On 12/07/2017 03:23 AM, Michael Paoli wrote:

>> oh, and I did have one key on a Cromemco 3102 terminal which I'd received
>> which didn't work (magnetic reed relay switch ... and ... how'd I repair it?
>> switch physically closed contacts, but not electrically - sealed in a glass
>> envelope - apply a rather high voltage (about 600+ VAC) very limited current
>> (about a 10+M Ohm resister in series) ... burnt off whatever (oxide?) was
>> preventing electrical contact, without welding the contacts together - has
>> worked fine ever since)

```