Losses in shunt-fed towers
 

Hi Wim,

despite my wire is only 4mm in diameter, SWR does not vary at all applying a
steady full-power carrier (not even some small oscillations of the reflected
power meter needle). I then presume that no big corona effect takes place (also
because I see no corona light at dark). I am not sure whether a light corona
effect (i.e neither visible nor influencing SWR) could take place, but if so I
presume loss would anyway be very low.

I will check again under rain conditions (test not done so far), but I am not
sure whether rain increases susceptibility to corona, with respect to hot summer
days (with umidity close to 80% sometimes). Any idea at that regard?

73

Tony I0JX

Hello Tony,

It is not the voltage that will introduce loss (as the field lines will go
into air and frequency is low), but it is (corona) discharge. Whether this
occurs, depends on presence of sharp edges and especially sharp (double
curved) surfaces with small radius, think of bolts, threaded rods, etc that
point away from the tower into the air.

A thin end of an antenne element can also cause breakdown as it points far
into the air (away from other solid structures) and the voltage maximum will
be at the elements of your HF antenna.

In case of your 8 kVrms (that is 12 kVp), an antenna element with 10mm
diameter and spherical (smooth) end may result in 2.4kVp/mm at the end. If the
end of the element is just straight (just cut aluminium), you will exceed
3kVp/mm and corona will occur.

If (corona) discharge occurs, you will notice a sudden change in VSWR with
increasing power.


--
Wim
PA3DJS
www.tetech.nl
Please remove abc first in case of PM

Losses in shunt-fed towers
 

On 9/13/2011 8:06 AM, Antonio Vernucci wrote:
HV loss is hard to estimate. You could make your shunt wire a shunt
tube or rod and basically eliminate corona (A rule of thumb is that
10-15 kV/cm radius will have virtually no corona.. so for your 6kV, a
1cm diameter tube is in the right ballpark) Skin depth at 3.5 MHz in
Aluminum is .043mm, and the usual rule of thumb is to make the tubing
wall thickness 3-5 skin depths. 0.12-0.20 mm seems about right. Copper
could be thinner wall (skin depth is less)

I presume that the corona effect should be visible at dark. So far I
have seen none, but we had not a single day of rain since I mounted this
antenna (incredible summer season...). So, I must verify when rain will
come, in a few days from now they say. Or do you think that corona may
not be visible?


Nope.. you can't usually see the corona discharge. The professionals
use a special camera with a narrow band solar blind filter.




I am using a 4-mm diameter wire so, if I will really have corona
problems, I could insert the (vertical) wire into a 1-cm aluminum tube
connected at the very bottom of the wire (i.e. at the antenna feed
point). Probably a 2-meter long tube could be sufficient (RF voltage
gradually diminishes getting away from the antenna feed point).


Sure..

Losses in shunt-fed towers
 

Antonio Vernucci wrote:

Hi Wim,

despite my wire is only 4mm in diameter, SWR does not vary at all applying a
steady full-power carrier (not even some small oscillations of the reflected
power meter needle). I then presume that no big corona effect takes place (also
because I see no corona light at dark). I am not sure whether a light corona
effect (i.e neither visible nor influencing SWR) could take place, but if so I
presume loss would anyway be very low.

I will check again under rain conditions (test not done so far), but I am not
sure whether rain increases susceptibility to corona, with respect to hot summer
days (with umidity close to 80% sometimes). Any idea at that regard?

What happens if you move the slant-wire up or down on the tower?

Losses in shunt-fed towers
 

On Sep 9, 3:10*pm, "Antonio Vernucci" wrote:
For 75 and 80 meters, I use a shunt-fed tower (58 feet high, with a fairly big
HF yagi on its top). The home-made tuner, placed at the tower base, has a
motorized variable vacuum capacitor in series and a (properly selected) fixed
capacitor in parallel. I can so remotely tune the variable capacitor across the
3500 - 3800 kHz range obtaining a perfect match to 50-ohm everywhere in the
band. The antenna works fine and I receive good reports.

The only thing which worries me a bit is the very high voltage that develops at
the antenna end (about 6,900 Vrms at 3.500 MHz with 2kW applied, growing to
about 8,600 Vrms at 3.800 MHz). I am not sure whether such high voltage could be
a source of significant losses.

To solve my doubt, I started by precisely measuring the capacitance of the tuner
capacitors, and I could then easily calculate the antenna impedance at 3.500
MHz, which resulted to be (18.2 + j 656) ohm.

ABOUT CURRENT
With 2 kW applied, one can easily determine that the RF current flowing through
the antenna is about 10.5A, quite higher than the 6.3A figure one would get
should the antenna resistance be 50 ohm instead of 18.2 ohm. So, I must expect
some more loss in the conductors due to the fairly high current.

ABOUT VOLTAGE
The 10.5 A current flowing through the big 656 ohm antenna reactance causes the
antenna RF voltage to get up to 6,900 Vrms at 3.500 MHz. Touching the wire with
a (well insulated) screwdriver, you would see a nice Tesla-like arc. I am
wondering whether such a high voltage could constitute, by itself, a source of
extra loss. Please note that the wire coming down from the tower is connected
directly to the vacuum capacitor terminal, with no stand-off insulator. So, I do
not see a place where power can get dissipated due to the high RFvoltage, other
than perhaps in the humid air (?).

I could probably avoid this situation by changing the tap height on the tower,
but I would run the risk of not being any longer able to tune across the whole
3500 - 3800 kHz band adjusting one of the two capacitors, and not both.

Any idea on whether the high RF voltage present on the antenna could cause some
significant loss?

Thanks and 73

Tony I0JX
Rome, Italy

I would think the losses as a result of 10 amperes current at the
feed point not to be excessive unless your shunt feed is made of small
wire. This is usually not the case if only designed for sound
mechanical integrity.

Jimmie

Losses in shunt-fed towers
 

On Sep 9, 3:10*pm, "Antonio Vernucci" wrote:
For 75 and 80 meters, I use a shunt-fed tower (58 feet high, with a fairly big
HF yagi on its top). The home-made tuner, placed at the tower base, has a
motorized variable vacuum capacitor in series and a (properly selected) fixed
capacitor in parallel. I can so remotely tune the variable capacitor across the
3500 - 3800 kHz range obtaining a perfect match to 50-ohm everywhere in the
band. The antenna works fine and I receive good reports.

The only thing which worries me a bit is the very high voltage that develops at
the antenna end (about 6,900 Vrms at 3.500 MHz with 2kW applied, growing to
about 8,600 Vrms at 3.800 MHz). I am not sure whether such high voltage could be
a source of significant losses.

To solve my doubt, I started by precisely measuring the capacitance of the tuner
capacitors, and I could then easily calculate the antenna impedance at 3.500
MHz, which resulted to be (18.2 + j 656) ohm.

ABOUT CURRENT
With 2 kW applied, one can easily determine that the RF current flowing through
the antenna is about 10.5A, quite higher than the 6.3A figure one would get
should the antenna resistance be 50 ohm instead of 18.2 ohm. So, I must expect
some more loss in the conductors due to the fairly high current.

ABOUT VOLTAGE
The 10.5 A current flowing through the big 656 ohm antenna reactance causes the
antenna RF voltage to get up to 6,900 Vrms at 3.500 MHz. Touching the wire with
a (well insulated) screwdriver, you would see a nice Tesla-like arc. I am
wondering whether such a high voltage could constitute, by itself, a source of
extra loss. Please note that the wire coming down from the tower is connected
directly to the vacuum capacitor terminal, with no stand-off insulator. So, I do
not see a place where power can get dissipated due to the high RFvoltage, other
than perhaps in the humid air (?).

I could probably avoid this situation by changing the tap height on the tower,
but I would run the risk of not being any longer able to tune across the whole
3500 - 3800 kHz band adjusting one of the two capacitors, and not both.

Any idea on whether the high RF voltage present on the antenna could cause some
significant loss?

Thanks and 73

Tony I0JX
Rome, Italy

Please forgive me but it seems a little odd that someone who can make
the calculations you have made can not compute the IR losses in the
feed. You are almost there.

Jimmie

Jimmie

Losses in shunt-fed towers
 

Hi Jimmie,

I am not sure of what you mean to say.

With regard to the IR loss (by the way, I presume that "IR" means I squared
times R), there is little difficulty to estimate them once you know the RF
current.

But my point was to estimate any extra losses caused by the very high voltage
that develops at the antenna terminal. I would not know how to estimate those.

73

Tony I0JX


Please forgive me but it seems a little odd that someone who can make
the calculations you have made can not compute the IR losses in the
feed. You are almost there.

Jimmie

Jimmie

Losses in shunt-fed towers
 

Nope.. you can't usually see the corona discharge. The professionals use a
special camera with a narrow band solar blind filter.

That is interesting.

But I presume that, increasing voltage, at a certain point the corona effect
becomes visible.

73

Tony I0JX

Losses in shunt-fed towers
 

El 13-09-11 17:30, Antonio Vernucci escribió:
Hi Wim,

despite my wire is only 4mm in diameter, SWR does not vary at all
applying a steady full-power carrier (not even some small oscillations
of the reflected power meter needle). I then presume that no big
corona effect takes place (also because I see no corona light at
dark). I am not sure whether a light corona effect (i.e neither
visible nor influencing SWR) could take place, but if so I presume
loss would anyway be very low.

I will check again under rain conditions (test not done so far), but I
am not sure whether rain increases susceptibility to corona, with
respect to hot summer days (with umidity close to 80% sometimes). Any
idea at that regard?

73

Tony I0JX

Hello Tony,

It is not the voltage that will introduce loss (as the field lines
will go into air and frequency is low), but it is (corona)
discharge. Whether this occurs, depends on presence of sharp edges
and especially sharp (double curved) surfaces with small radius,
think of bolts, threaded rods, etc that point away from the tower
into the air.

A thin end of an antenne element can also cause breakdown as it
points far into the air (away from other solid structures) and the
voltage maximum will be at the elements of your HF antenna.

In case of your 8 kVrms (that is 12 kVp), an antenna element with
10mm diameter and spherical (smooth) end may result in 2.4kVp/mm at
the end. If the end of the element is just straight (just cut
aluminium), you will exceed 3kVp/mm and corona will occur.

If (corona) discharge occurs, you will notice a sudden change in
VSWR with increasing power.


--
Wim
PA3DJS
www.tetech.nl
Please remove abc first in case of PM


Hello Tony,

As long as your wire with D=4mm is not close to other structures (say
0.25...0.5m), you will not exceed 1 kVp/mm when the wire itself is
carrying 12kVp. You still have some margin before getting close to 3
kV/mm.

At ground level, influence of HF frequency on dielectric breakdown
voltage of air is negligible. If the wire is coated, the diameter
increases, hence reducing the E-field at the plastic-air interface.

When VSWR doesn't, change (with increasing power), the loss due to
corona discharge will be minimal. Of course when you lose 10W from
2kW, you will not notice this probably. However if you get a discharge
between metal and plastic insulation, the plastic will fail in the end
(due to spark erosion).

conducting Droplets (rain) do reduce the breakdown voltage as it
distorts the field. This introduces 'hot spots'. Unfortunately I don't
have data on this.

I could imagine that moisture/water layer on the wire would introduces
some loss (as you get a current through the moisture layer). But this
is not due to corona effect.

It can be calculated, as you know the voltage and you can calculate
the capacitance to ground (so you know the capacitive current that
"leaves" the wire.


--
Wim
PA3DJS
www.tetech.nl
Please remove abc first in case of PM

Losses in shunt-fed towers
 

Hello Tony,

As long as your wire with D=4mm is not close to other structures (say
0.25...0.5m), you will not exceed 1 kVp/mm when the wire itself is carrying
12kVp. You still have some margin before getting close to 3 kV/mm.

At ground level, influence of HF frequency on dielectric breakdown voltage of
air is negligible. If the wire is coated, the diameter increases, hence
reducing the E-field at the plastic-air interface.

When VSWR doesn't, change (with increasing power), the loss due to corona
discharge will be minimal. Of course when you lose 10W from 2kW, you will not
notice this probably. However if you get a discharge between metal and plastic
insulation, the plastic will fail in the end (due to spark erosion).

conducting Droplets (rain) do reduce the breakdown voltage as it distorts the
field. This introduces 'hot spots'. Unfortunately I don't have data on this.

I could imagine that moisture/water layer on the wire would introduces some
loss (as you get a current through the moisture layer). But this is not due to
corona effect.

It can be calculated, as you know the voltage and you can calculate the
capacitance to ground (so you know the capacitive current that "leaves" the
wire.


--
Wim
PA3DJS

Thanks for info. When things go down to the physics level, it is not so easy to
understand phaenomena, unless one has specific knowledge on them.

73

Tony I0JX

Losses in shunt-fed towers
 

On 9/13/2011 10:55 AM, Antonio Vernucci wrote:
Nope.. you can't usually see the corona discharge. The professionals
use a special camera with a narrow band solar blind filter.

That is interesting.

But I presume that, increasing voltage, at a certain point the corona
effect becomes visible.

Yes.. but that's usually a LONG way from where there's detectable RFI or
noise for instance. When you can see visible corona from some meters
away, you're probably in the 10s or maybe low 100s of kV range.

For instance, I can run my small 14" Van deGraaff generator with no
visible corona in a semi-darkened room, but, judging from the copious
ozone production, there's quite a bit of corona (as well as the
characteristic crackle sound).

Maybe people have seen noticeable corona at 2-3kV, but that's usually
indoors in a dark room, and you're fairly close and well dark adapted.
Corona from aged ignition wires (around 15-20 kV) also often visible, in
the dark.