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

Re: Reply to And anotherone bites the dust



Hi Viv,

Glad to hear that we could brighten up your spirits!
Keep at it until it works!!
Here are a few more comments:

>No I don't have any resistors only two 3mH chokes.  After these
>problems I intend filters, resistors etc etc.  I need to SEE whats
>going on, so getting a high voltage probe to try scope this stuff
>is a new objective..

DONīT uses chokes. If they start to ring (for whatever reason), they
will be defeating the original purpose (to kill spikes) and will only
add to the stress the xformer is seeing. I am using no protection
whatsoever, except for a safety gap directly on the NST. Did those
chokes ever get hot? This might be a sign that they are ringing.
Where they air or ferrit chokes? Ferrit chokes are NOT lossy enough
to provide adequate protection.

>I had my previous Neon for two years using a series static gap then I
>tried a sync Rotary and my problems started :-(

Quite obviously *the* problem is being aggrevated by the rotary. If I
remember correctly, you are running a 2x Fmains SRSG, right? So your
"failure time" is 10ms. What is the output current of your NSTs and what
cap are you using? You will want to size both, so that the voltage
doesnīt rise >>> above the 10kV rating of your NSTs in these 10ms.
Try bigger caps and stay away from mains reso circuits. You DONīT
need the higher voltage, period.

>The Neons were Identical units.  They were single output only - no
>center tap.  One side was permanently earthed to the case.  I ran
>them in anti-phase with a common ground. I do not consider them
>being in series from a HV stress point of view in this configuration.

Okay, so the ct theory wasnīt quite on the mark. I will assume you have
both ground (cases) connected together (Donīt feed the high voltage
terminal from one xformer to the case of the next. Instead reverse
the primary leads for anti-phase operation). If this connection is
earthed, you shouldnīt be overvolting the neons. (effectivly they will
be delivering +-5kV above ground level). If the "ct" isnīt earthed, you
will be feeding a lead, which is usually near or at ground level with
5kV. The insulation between core and wire (= nearest core =
ground) is NOT good enough for this feat. Have you torn one apart
and had a look, where it failed? Where in the winding did it fail
(start/end/middle)? Did the insulation fail due to overvolting or
overcurrent? Is it always the same xformer (i.e on the same side
of the circuit)? Is there one or more places in the insulation, where
it failed? Give Richieīs and my idea of using a 200bps SRSG a
thought. You would only need to use 2 as many rotating electrodes
and you will be decreasing the "dangerous" failure time to 5ms.

>So is 1/4" with rounded brass electrodes suitable for 5KV safety gap ?

This is a hard question to answer, because 5kV will jump different gap
widths (depending on ROC and altitude, humidity, etc). I would
disconnect the tank circuit. Make the gap very small (1/16"). Short one
xformer (they are current limited, so no harm will be done). Fire up
the NST and slowly (power down in between of course) increase the
gap distance until the NST ALONE just doesnīt fire across the gap.
This will be the safest distance. Repeat this for the other side.

Once the SRSG is set up correctly, you can further REDUCE this
distance (for safer operation) because the SRSG will depleat the main
cap of itīs voltage (i.e: it will fire) long before the cap can reach
the SSG firing voltage. Are you using a single cap or an Equadrive
setup (one cap in each HV leg)? The Equadrive presents NO
advantages at all and WILL provide you with headaches, because
the two caps MUST be sized 100% equally and you will need to
discharge each cap seperately.

I donīt think the three piece gap killed your xformers. If you grounded
the connection between xformer #1 and #2 (i.e: the "ct"), this can NEVER
rise above zero volts (okay maybe a few, but certainly less than 100V).
Letīs look at the four cases possible:

1.) None of the gaps fire. Assuming your gap is set correctly, this
     means your xformers are not being overvolted. All is well.

2.) Both sides of the gap fire. This means excessive voltage is
     present on both "sides" of the circuit. While the gap doesnīt
     act as a full short (arc resistance), it is close enough to be
     viewed as such. In the case of no coil attached, the voltage
     drops (on my 7500V setup) to about 250-450V per side. In
     coiling usage, the voltage might be higher because you are
     feeding the gap with high power (from the cap(s)). Following
     Ohmīs Law, the voltage drop across the resistor (i.e.: your
     arc) will be higher. However, it can never (unless your are
     running ĩF of capacitance) reach the 5kV level. The maximum
     voltage (drop) is used up in ĩsecs. My double enameled wire
     (0.85mm), used to wind my secondary, will withstand ~5kV with
     no additional insualtion inbetween. I assume your xformer wire
     is (almost) the same, so this canīt be the reason either (over-
     volting for ĩsecs time).

3/4.) Only one (left or right) side fires. In this case only one
        side of the tank circuit is experiencing overvolting. Once
        the gap fires, it is essentially the same as case #2. The
        execessive voltage is drained from the leg to ground. The
        other side (xformer) sees no excessive voltage, because
        the >>voltage from the first side is clamped to ground.
        If it sees any overvolting at all, it is surely less than a few
        hundred volts. Certainly not several thousand. The worst
        that could happen is that the SSG from the second xformer
        starts to fire and clamps the total setup to ground.


Coiler greets from germany,
Reinhard