PC Board Capacitors at HF
 

I want to make a couple of L networks to match a 20m delta loop to 50
ohm coax on 20m and 17m; 100W max. I know I can use a transformer on
20m, but I figure I'll save my ferrites if I'm building L networks
anyway, and to be economical of junkbox parts, I'd also like to save my
air variables.

Would capacitors made of PC board be decent transmitting caps at these
frequencies, provided they didn't break down? The board is hamfest
surplus special... so I don't know what it is, but it may be G10...
it's a dark green fiberglass material, about 2mm thick dielectric.
Comparing the capacitance of a 2 inch square of the material (88pF) to
a parallel plate cap with air dielectric of the same physical
dimensions (12pF) gives a dielectric constant of about 7.3.

I'm going to try it anyway, but I'd appreciate any input anyone has
about the suitability of unknown PC board as HF transmitting capacitor
material, primarily with respect to dielectric losses.

If it matters, the 20m network is a 1.14uH series inductor with an 88pF
capacitor shunting the antenna side. The 17m network is a 3.08uH
series inductor with a 29pF shunt. The 20m antenna impedance is about
250+j0 and the 17m impedance is about 950+j1200. I figure there might
be some voltage breakdown issues, but worst case, I smooth the edges of
2 or 3 2x27pF or 3x27pF caps and put them in series.

Thanks and 73,
Dan

PC Board Capacitors at HF
 

For the 20M case, the RMS voltage at 100 watts is only about 160.
Especially if you trim the copper back a bit from the board edge, it
shouldn't be a problem. The 17M case is about 500VRMS at 100 watts.
Even that probably won't be a problem. (I have an SO8 IC in front of
me that claims to be OK with 600 volts between adjacent pins: less
than half the thickness of your board.) Since the board area is so
small for that 29pf, why not just try it? The thing I'd be most
worried about is corona at the sharp edges of the board. If you were
to make the capacitor plates circular, that would help. In addition,
if you soldered a ring of, say, 12AWG bare copper wire to the edge of
the circular plate on each side, that would make a rounded edge. If
one side of the capacitor is "ground" anyway, just leave that side's
copper the size of the board, and trim back only the other side. The
outer edge of the PC board copper circle would just touch the "donut"
of copper wire. A cap like that would probably be good for several
thousand volts, assuming the dielectric extended out several mm past
the circle. Fiberglass-epoxy isn't the best dielectric for a capacitor
in terms of high frequency loss, but should be OK for your low power.
But is it really standard fiberglass-epoxy? If it's standard 1.6mm
thick, I'd expect no more than about 70pF with 4 square inches of plate
area. A relative dielectric constant of 7.3 is too big for standard
FR-4/G-10 material. You may need to put some RF current through it to
test to make sure it doesn't heat up too much.

I really don't think you'll need to put caps in series to get what you
want. In any event, instead of making three separate capacitors, you
could just add an air gap between two pieces of your copper-clad
material...or use some dielectric other than air if it's easier. But
some nylon screws and nuts (the nuts acting to space the boards), or
even steel ones out on the periphery away from the capacitor plates,
should work fine. You could even make it adjustable that way, by
adjusting the spacing...

Cheers,
Tom

wrote:
I want to make a couple of L networks to match a 20m delta loop to 50
ohm coax on 20m and 17m; 100W max. I know I can use a transformer on
20m, but I figure I'll save my ferrites if I'm building L networks
anyway, and to be economical of junkbox parts, I'd also like to save my
air variables.

Would capacitors made of PC board be decent transmitting caps at these
frequencies, provided they didn't break down? The board is hamfest
surplus special... so I don't know what it is, but it may be G10...
it's a dark green fiberglass material, about 2mm thick dielectric.
Comparing the capacitance of a 2 inch square of the material (88pF) to
a parallel plate cap with air dielectric of the same physical
dimensions (12pF) gives a dielectric constant of about 7.3.

I'm going to try it anyway, but I'd appreciate any input anyone has
about the suitability of unknown PC board as HF transmitting capacitor
material, primarily with respect to dielectric losses.

If it matters, the 20m network is a 1.14uH series inductor with an 88pF
capacitor shunting the antenna side. The 17m network is a 3.08uH
series inductor with a 29pF shunt. The 20m antenna impedance is about
250+j0 and the 17m impedance is about 950+j1200. I figure there might
be some voltage breakdown issues, but worst case, I smooth the edges of
2 or 3 2x27pF or 3x27pF caps and put them in series.

Thanks and 73,
Dan

PC Board Capacitors at HF
 

Hello Dan,

About your matching network for loop antennas:

Breakdown will not occur through the material in your case. Major issue
is dielectric losses.

As you know the antenna impedances, you are able to calculate the
apparent power for the capacitor (just Ic*Uc [rms values]). Both
Epsilon and loss factor for FR4 varies (depending on the manufacturer).
You can use 0.01 as a rule off thumb. So if your capacitor "consumes"
an apparent power of 500W (at 100W input), the loss inside the
capacitor is about 500*0.01 = 5W.

I hope this will help you.

Wim
PA3DJS

PC Board Capacitors at HF
 

wrote in message
ups.com...
I want to make a couple of L networks to match a 20m delta loop to 50
ohm coax on 20m and 17m; 100W max. I know I can use a transformer on
20m, but I figure I'll save my ferrites if I'm building L networks
anyway, and to be economical of junkbox parts, I'd also like to save my
air variables.

Would capacitors made of PC board be decent transmitting caps at these
frequencies, provided they didn't break down? The board is hamfest
surplus special... so I don't know what it is, but it may be G10...
it's a dark green fiberglass material, about 2mm thick dielectric.
Comparing the capacitance of a 2 inch square of the material (88pF) to
a parallel plate cap with air dielectric of the same physical
dimensions (12pF) gives a dielectric constant of about 7.3.

I'm going to try it anyway, but I'd appreciate any input anyone has
about the suitability of unknown PC board as HF transmitting capacitor
material, primarily with respect to dielectric losses.

If it matters, the 20m network is a 1.14uH series inductor with an 88pF
capacitor shunting the antenna side. The 17m network is a 3.08uH
series inductor with a 29pF shunt. The 20m antenna impedance is about
250+j0 and the 17m impedance is about 950+j1200. I figure there might
be some voltage breakdown issues, but worst case, I smooth the edges of
2 or 3 2x27pF or 3x27pF caps and put them in series.

Thanks and 73,
Dan

Use quality coaxial cable- high V breakdown and very low loss factor.

Dale W4OP