HBT and more info for Doug
Original Poster: Doug Brunner <dabrunner-at-earthlink-dot-net>
Doug " *smacks self on head* I should have thought of insulating between the
and the core. Thanks for bringing it to my attention. I'll have to re-wind,
least it won't fry the insulation. Would 60 mils of polyethylene do the job?
only one layer, so at least that won't be a problem. I'm putting about 5 24
windings on top of each other and paralleling them, so current won't be an
issue. 32 AWG, according to my information, can take an amp or so. As for
limiting, I've done the math, and thanks to the oil, it shouldn't fry even
if it's shorted. I'm not aware of oil increasing arc-overs until it's
mistaken? It wouldn't depend on the oil to prevent arcovers-it's only to
increase thermal conductivity to the outside."
First of all I would use one wire (no paralleling) to wind the pri and sec
(You have to rewind anyway). Iīll tell you why. First of all itīs a lot less
lossy. Secondly, if you have a voltage difference in the windings (because
itīs pretty hard to wind 5 wires in layers, each layer exactly on top of the
other) you might have an arcing problem in case of extreme VPL difference.
You say you have done the math on current limiting. The transformer you are
building is NOT (no, nope, none, not at all) current limited. If you short the
secondary it will pull more and more current until either the primary or
secondary burns up. If you take a MOT for example and short the heater
filament winding (rated at 3.5V and 13.5A) the filament winding will deliever
in execess of 50(!)A, no problem AND MOTīs are usually current limited. The
only way a non shunted transformer will "limited" the shorted current it can
dish out is through core saturation. Usually one of the windings give way
before this happens (on long runs).
The oil surves two purposes. First it will prevent arcovers (to a certain
degree) and secondly it transfers the heat of the windings to the outside. The
nice thing is that the oil cools the inner windings, too. Pure air convection
doesnīt help the inner windings at all. The oil does NOT increase the output
potential your transformer has !!! It will allow you to push it harder and
further for a length of time, but it will NOT increase the capability. (This
is one reason why oil filled pigs and PTīs can be overrated for short
durations, but it has nothing to do with their rated capabilities) The rated
capabilities are simply a question of wire and iron core gauge. (i.e.
saturation and VA properties). If the oil would prevent burning up the
windings, I could throw a 500W transformer in a big tub of oil and pull 5kVA
from it. Nope that wonīt cut it. The oil will only allow a longer run time at
a specific loading, but not a higher rating. Itīs important to understand the
Now to the thickness of your poly. Poly is good, altho you can use Kraft
paper, too.(Tell ya another secret: Use thick grocery bags instead of Kraft
paper. Itīs almost the same and realllllllly cheap). The thickness you need
depends on the volts per layer. Example: If you have a TPV rating of 1 and
wind 100 turns on a layer, you will have a VPL of 100. This means the
difference between turn #1 and turn #101 will be 100 volts, so your insulation
will have to withstand at least 200 volts (for a 100% safety margin, which is
not too high in hv design) in this case. The insulation between the core and
the secondary will have to withstand the full voltage, unless you ground one
of the hv secondary terminals to the core and ground. But then you will have
to design the primary insulation to withstand the full output voltage, so
youīre not off any better.
Understanding the VPL "rating" is important. Itīs better to wind a relativly
small width and use more heigth than it is to wind a long layer and use less
layers (heighth) to achieve the same output voltage. In the second case you
VPL will be higher and this means your insulation will have to withstand a
much higher voltage. Plus as you build up heigth (instead of width) you also
get further away from the core and this is of great help, too.
You also should stay away from the edges of the insulation, to prevent arc
overs along the edge of the insulation layer.
I have done some extensive poly "high voltage destruction testing" and have
found that 1 mil thickness will withstand about 2.0kV-at-50Hz (for a longer
length of time) before PUNCHING holes into it. This means to be safe you can
say 1kv per mil AT 50/60Hz (! very important to note the 50/60 Hz !). At TC
frequency I consider a safe poly voltage rating to be 450-500 volts per mil
which comes pretty close to values I have seen on the list or in old RQ
archives(I believe these where around 600V/mil).
I did these test to find a good value for poly caps. All my test caps are
built with these specs in mind and I run em dry (no oil) and no safety gap. I
have only punctured one cap (this was built with lots of layers of 1 mil PE).
The problem was not really the voltage rating I used, but this cap got so hot
that it actually melted the PE. As a matter of fact I got burns on my fingers
(after about 2 minutes cool down time) when I ripped it apart to see what
happend. I can reduplicate this heating effect at will. It only happens when
I use the 1 mil PE sheets. My other caps (built with 6.7 mil PE sheets) donīt
get warm at all Not even on longer runs or under higher power. I donīt have a
real physical explantion for this, but I feel that it is do to inductive
heating inside of the cap.
BTW: There was no corona, so thatīs not the reason. I checked this in a
My "real" cap will of course be oil filled and have a safety gap. I just
wanted to see how far you can push it before it dies and goes to hv heaven.
That way I can design a cap that will really hold up in TC use. (Yes, I did
have tank kickback and no, it didnīt kill the cap). Kickback was another
reason for my destruction testing.
Hope I could be of help,
Coiler greets from germany,