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


Original poster: "Gerry  Reynolds" <gerryreynolds@xxxxxxxxxxxxx>

Hi Christoph,

The Terry filter is to keep HF and VHF off the transformer - not the fundamental resonant frequency of the coil. These higher frequencies are created when the spark gap fires and when the topload arcs (especially to ground). At 5KVA, you can probably reduce the 1Kohm resister and increase the shunt cap value to keep the same frequency response of the filter. With a 1K resister and 400ma, the power dissipation will be 160 watts. For example, you could use a 100 ohm 50 watt resister and increase the shunt capacitance by 10x and still have a 3x power safety margin with the same frequency response.

A string of caps giving you 5.5nf will keep the frequency response of the filter the same with a 100 ohm resister as a string of caps giving you 0.55nf with a 1000 ohm resister (original Terry filter design). Intermediate solutions will also work (just a matter of how much power you want to dissipate in the series resistance). If your homemade transformer is under oil, the importance of this filter is probably not as great as with a NST (with tar). The real question is if you can do without the shunt caps and if you can't, would you be able to change the response curve to only filter higher frequencies (ie, use less than the 5.5nf with 100 ohms). The series resistance is probably still a good idea because of parasitics and any transmission line effects.

Gerry R.

Original poster: "Christoph Bohr" <cb@xxxxxxxxxxxxxxx>

Hello DrR, folks.

> Even PT and pole xmfrs need a protection circuit if you want to keep
> the high freq currents out of the secondary windings --- you only
> want 60 Hz current there.  The HF currents form "spiders" tracking
> across the HV windings and can lead to break down of the sec
> winding.  These are caused RF which likes to "track" across insulator
> surfaces.  Not good inside a power xmfr.  Pole xmfrs will tolerate
> some abuse but good design says to eliminate the problem before it
> occurs --- prevention.
OK, with my homemade transformer, I obviously got away without a filter
over the last two years, but your comment sounded my alarm bells.
Now, how would I size a filter for my xformer, measured at 13500V ca 400mA out,
usually running at around 5KVA.( single ended )
I am not the math guy, but I tried to do some calcs and found,
that with a lets say 1nF doorknob and 1KOhm resistance the filter
will cut off around several hundred KHz, isn't that way to high? I would have
aimed for something over several KHz.... am I missing something.
What layout would you suggest, I hoped to get away without another MMC in the
filter, would rahter use some doorknob stype cap there...
Any suggestions are appreciated....

best regards

Christoph Bohr