[R-390] Isolation transformer question

[Gary Gitzen]

Hi R-390,

Paul: many thanks for your thoughts on the problems using
isolation transformers (IT). Please consider extending
your thoughts to the following.


Paul suggested a few issues and/or problems with using
isolation transformers, as well as a possible "solution".

His first comment was that in theory the floating secondary
could be floating at KV levels above ground due to static
electricity. A second, unmentioned but much more dangerous
possibility, is a power supply fault in the powered device
causing possibly high voltages to appear on the IT secondary.

Paul further suggested using small 110V "globes" (presumably
light bulbs for us here in the US) from each secondary
leg to ground. These would glow dimly under normal operation,
and one would glow brightly if either secondary leg shorted
to ground. I think this is a Very Bad Idea, and will get to
that shortly.

First, let's take the case of static electricity on the secondary.
This is easily dealt with by connecting 1-10 meg 2W resistors
from each leg to ground. Static problem solved.

Second, Paul's suggestion of a "leg shorted to ground" indicator
is an excellent idea, but might I suggest instead two neon bulbs,
each in series with say 220K/2W from each leg to ground?
Under normal conditions, they would both be dark, and conduct
zero current. If either turns on, one leg is shorted to ground.
If they both turn on, some fault in the powered equipment
is putting high voltage on the secondary.

Paul further claimed that the two "globes" would create
a virtual center tap on the secondary, with "safe" voltages
of 55-60VAC on each leg. I must respectfully disagree with
"safe". My understanding is that under the wrong conditions,
as little as 10-20V at 10MA or so can prove fatal. 
Paul further claims that any fault condition, such as a shorted
leg, would cause at least one bulb to glow brightly. That would
be true for a shorted-to-ground leg, but what happens if
A: That bulb burns out?
B: A fault puts high voltage on the secondary, blowing both
   bulb filaments?
In either case, both bulbs would be dark, but that might not
be noticed when the "normal" condition is a dull glow.
The neon bulbs appear to address both issues. If either/both
neons come on, a fault exists.

Paul's suggestion of a virtual center tap on the secondary,
resulting in 55-0-55 is excellent, but there may be a better
way of getting it, and then adding even more protection for us
mortal users.
If you take a small transformer with a 120/240 split or
center tapped primary, connect each end to one of the IT legs,
then ground the middle of the winding(s), you have created
a 60-0-60 voltage source, ideal for powering an R-390X rcvr.
But we're not done yet. For added protection, add a GFI
to the secondary. Ground the GFI and we now have a 60-0-60
voltage source with GFI protection in case the user happens
to connect him/herself between either leg and ground.
If you happen to have a pair of 130-150V varistors and connect
them from each leg to ground, you've just added protection
in case a power supply fault in the powered equipment ever
tries to put high voltage on its AC power feed from the IT.

Does anyone else have any opinions, insights, or suggestions
for further improvements?

Aside: a 60-0-60 AC feed is now popular with the professional
audio crowd (not to be confused with audiophools) because
it reduces hum.

Just my $0.02

Gary, San Jose CA

 O  / 
  >< ------------------------- cut here ----------------------
 O  \ 

> From: "Paul Galpin" <galpinp at absamail.co.za>
> To: <r-390 at mailman.qth.net>
> Date: Sun, 8 Jul 2007 21:22:54 +0200
> Subject: [R-390] Isolation transformer question
> 
> Certain assumptions are being made about the secondary of the I.T. In fact,
> there are several unknowns. What is the secondary live-to-earth resistance,
> neutral-to-earth impedance? How do these change when a piece of equipment is
> connected? The secondary is floating and could even be standing up at kVolts
> of static (very unlikely, but theoretically possible in a very dry climate)
> 
> IMHO, it would be best to relate the secondary to earth in a known way.
> 
> My suggestion is to connect small 110V globes from each side to earth,
> giving visual indication that the supply is now 55-0-55, very safe if you
> insist on connecting yourself to the power cables!  If either lamp goes out
> and the other goes bright, then you've got a short somewhere between one leg
> and earth. You might want to put a shorting switch across one leg to earth
> to create this condition, maybe, for certain tests
> 
> Of course, down here it would be a 110-0-110 supply, but that's still safer
> than the full 220.
> 
> Incidentally, ELT's (GFI to you) also go out on a low impedance
> neutral-to-earth short.