[Prev][Next][Index][Thread]

Found it ! :o(( (was:What is going on?? )



Hello Gary, all,


Snipped myself

I wrote:
>Now for the next surprise:
>I used 18nF caps in a series/parallel arrangement and
>found the highest voltage of 654Vac achieved with a 50.2nF
>cap More ?!?!?. This, too, does NOT match the calculated
>value of a mains reso cap (31.8nF). However, the
>increase in voltage 68V vs. 654V is what I would expect
>from a resonance circuit.

Hehehehe. Stupid me!! I know what I forgot. Let me explain:
German NSTs are built somewhat different than U.S. NSTs.
The are open frame construction and have no potting compound
whatsoever.

Construction of the core:
------------------------------
(use courier font to view):
-----------
|-|-----|-|
|S|     |S|
|S|     |S|
| | |-| | |
| | |-| | |
|P|     |P|
|P|     |P|
|-|-----|-|
-----------
It is a twin I core, each I symbolized by "| |"
SS are the secondary coils (2 in series, ctīd)
PP are the primary coils (2 in series also)
- means core laminations

Further more, they have an adjustable shunt, so that you
can match the xformer to the neon tube (diameter, length). I
have a half dead NST, that I had used to make my horizontal
Jacobīs Ladder (see my website). As this xformer was half
dead anyway, I decided to use it as my testbed. Torturing
electronic and electrical components is my second nature, so
I did not feel guilty at all, while I removed the shunt completely.
The measured (!) secondary current was around 350mA. Even
though the NST is only rated for 75mA, it withstood the 10-
12 min run on the Jacobīs Ladder, w/o any further damage.
It did get plenty warm, but not dangerously hot (you could
still touch it). Well, what do you expect, with the copper
losses being 25 times as high as compared to normal
operating conditions (350mA vs. 75mA)? :o)) After seeing
this, I decided to modify the shunts for more, but safe
(longlivity of the NST) power. This is easy to do because
the shunts are made from small laminations, which are
simply riveted together. All I needed to do was drill out
the rivets, remove a few laminations and re-rivet the
whole thing. I reinstalled the shunts into the NST. Then
I shorted the secondary and let her rip. With 180mA
secondary current, the NST got only barely warm after
a 20 minute run. Satisfied, I decided to stop there. A 240%
increase in current is okay with me :o)). Then I counted how
many laminations I had removed and proceeded to do the
same with my good 7.5kV 75mA NST (which, in unmodified
condition, has gotten me the posted 125 cm sparks) for
more power. After finishing and testing it, I called it a day
and put the NST away. During the testing mentioned above,
I had forgotten about this modification. My shortend shunt
had been postioned so that the NST was delivering 130mA
and not 75mA. This is why there was a descrepancy
between the 31nF cap (calcīd) and the true reso cap
(50nF). Oh boy!!

As to the result of first experiment, I can only guess, but I
would assume my sig gen simply didnīt deliver enough
ummph (output current) to make a reliable assumption.

Gary:
I used the AC output on my signal gen. It has 50ohm,
600 ohm and pure AC (cap coupled) outputs. It has been
a while since I built this thing, but as far as I remember I
used a constant (user variable) voltage source and not a
constant current source for the output. My frequency
meter is a 9 digit HAMEG 1.9 Ghz counter with very high
accuracy. Using the 2Vac source, the NST (primary)
pulled ~140mA (w/o a cap attached) and roughly 200mA
at resonance.

Now, while I know what I did wrong, I find Jon Rosenstielīs
problem (30mA vs 48mA NSTs, yet exact same resocap) a
bit hard to understand, because I could actually adjust the
NST (via itīs movable shunt) to give me the best resonant
charging condition (i.e: I could match the NST TO THE CAP).
I tried various cap sizes and could always adjust the NST for
highest voltage rise (being ~650V in all cases, with 2Vac on
the primary side. I find that you will have to actually MEASURE
the output current (to find the proper resocap), rather than
trust the faceplate ratings. Strangely enough Jonīs NST(s) didnīt
seem to care what the real output current is. Any ideas ?
BTW Jon: Did you measure the 48 vs. 30mA?

As soon as I get my 50nF cap built, I will try out my newfound
NST power (+240%) and see what kind of spark output it gets
me.

Not scratchinī my head anymore :o))

Coiler greets from germany,
Reinhard