Hi,

I'm currently having to build some high-voltage capacitors for an auto-ATU
that I'm building. The capacitors are made from interleaving brass shim
with some kind of dielectric. It was suggested that the dielectric was made
from a triple layer of polyethylene sheet (ie freezer bag) stuck together
with DC4 grease to exclude any air. However, this seems a messy approach
and some engineer friends suggested that instead I use overhead projector
transparency. I found that a double layer of this transparency worked to
give me reasonable capacitor values with a reasonable number of plates per
capacitor.

Now, as the project requires, I fiddled the caps into a binary addition
series (1pF, 2pF, 4pF, ... , 1024pF) using a capacitance bridge with a 1kHz
oscillator. I was careful to measure the caps in situ so as to take into
account any stray capacitance on the circuit board. However, the tuner
doesn't appear to be matching as expected (its an L-match). Does anyone
have any views on using this dielectric? Is it possible that it has a
frequency-dependent dielectric constant so that the cap values as measured
at 1kHz are different at HF?

Any comments/help etc gratefully received!

73

Chris 2E0AYO


---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.788 / Virus Database: 533 - Release Date: 01/11/2004

When measuring capacitors at HF and UHF you get higher values of
capacitance than at audio frequencies. This is due to the inductance
and inductive reactance of the connecting leads between the capacitor
and the measuring instrument,

Furthermore, the capacitors themselves when connected into a practical
circuit, in series with their connecting wires, behave at HF and VHF
as if they have larger capacitances than their audio frequency values.

The actual values of capacitors do not change of course.

Large capacitors (1000 pF) are affected much more than small values
(10 pF).

The dielectric constants of low-loss plastic insulating materials are
constant between AF and VHF. Apparent changes in capacitance are NOT
due to changes in the dielectric constant.

On the other hand, what does it matter. When adjusting a tuner one's
eyes are focussed on the SWR meter and not on the capacitor values.
And the same applies to automatic tuners which don't have eyes.
----
Reg, G4FGQ

Now, as the project requires, I fiddled the caps into a binary addition
series (1pF, 2pF, 4pF, ... , 1024pF) using a capacitance bridge with a
1kHz
oscillator. I was careful to measure the caps in situ so as to take into
account any stray capacitance on the circuit board. However, the tuner
doesn't appear to be matching as expected (its an L-match). Does anyone
have any views on using this dielectric? Is it possible that it has a
frequency-dependent dielectric constant so that the cap values as measured
at 1kHz are different at HF?

Any comments/help etc gratefully received!
==========================
Have you tried to make capacitors from pieces of double sided fibreglass PCB
material ? Or even better the thin teflon material used for microwave
frequencies (expensive stuff). Because the dielectric material is thicker
,you might need a larger area to arrive at the wanted capacitance.

If you own or can borrow a MFJ 259B analyser ,you can measure the
capacitance at the operating frequency instead of at 1000 Hz.


Frank GM0CSZ / KN6WH

"Chris" wrote in message
...
Hi,

I'm currently having to build some high-voltage capacitors for an auto-ATU
that I'm building. The capacitors are made from interleaving brass shim
with some kind of dielectric. It was suggested that the dielectric was
made
from a triple layer of polyethylene sheet (ie freezer bag) stuck together
with DC4 grease to exclude any air. However, this seems a messy approach
and some engineer friends suggested that instead I use overhead projector
transparency. I found that a double layer of this transparency worked to
give me reasonable capacitor values with a reasonable number of plates per
capacitor.

I would be a little concerned about the loss factor of the transparency
material. The polyethylene originally suggested is an excellent dielectric-
the transparency material may yield reasonable capacitor values, but if the
loss factor is high, wach them smoke.
As suggested, teflon is also excellent.
For the lower values a good qulaity 50 Ohm coax will yield about 29pF/foot.

Dale W4OP

"Chris" wrote in message
...
Hi,

I'm currently having to build some high-voltage capacitors for an auto-ATU
that I'm building. The capacitors are made from interleaving brass shim
with some kind of dielectric. It was suggested that the dielectric was
made
from a triple layer of polyethylene sheet (ie freezer bag) stuck together
with DC4 grease to exclude any air. However, this seems a messy approach
and some engineer friends suggested that instead I use overhead projector
transparency. I found that a double layer of this transparency worked to
give me reasonable capacitor values with a reasonable number of plates per
capacitor.



Look for some Kayton made by Dupont. It is a high voltage plastic like film
similar looking to the mylar sheets. It is especially made for applications
like you want.

Ralph Mowery wrote:
"Chris" wrote in message
...

Hi,

I'm currently having to build some high-voltage capacitors for an auto-ATU
that I'm building. The capacitors are made from interleaving brass shim
with some kind of dielectric. It was suggested that the dielectric was

made

from a triple layer of polyethylene sheet (ie freezer bag) stuck together
with DC4 grease to exclude any air. However, this seems a messy approach
and some engineer friends suggested that instead I use overhead projector
transparency. I found that a double layer of this transparency worked to
give me reasonable capacitor values with a reasonable number of plates per
capacitor.




Look for some Kayton made by Dupont. It is a high voltage plastic like film
similar looking to the mylar sheets. It is especially made for applications
like you want.


sometimes sold as an insulator in sheets for transistors

On Mon, 18 Apr 2005 10:41:53 GMT, "Highland Ham"
wrote:

Have you tried to make capacitors from pieces of double sided fibreglass PCB
material ?

With traditional FR4/G10 materials you are only going to achieve
Q-factors of the order of 20-30 with a PCB capacitor even on HF, so
don't be upset, if the PCB material generates a lot of smoke :-) even
when exposed with only 100 W of RF power, especially if some high
impedances are involved.

Or even better the thin teflon material used for microwave
frequencies (expensive stuff).

Teflon has low losses on HF, tolerates high fields strengths and might
be quite usable material for HV RF capacitors, despite the lower
dielectric constant.

Paul OH3LWR

On Mon, 18 Apr 2005 13:29:05 GMT "Ralph Mowery"
wrote:

Look for some Kayton made by Dupont.

I suspect that you are thinking of Kapton.

-
-----------------------------------------------
Jim Adney
Madison, WI 53711 USA
-----------------------------------------------

"Jim Adney" wrote in message
...
On Mon, 18 Apr 2005 13:29:05 GMT "Ralph Mowery"
wrote:

Look for some Kayton made by Dupont.

I suspect that you are thinking of Kapton.


It could be. I copied the spelling from another artical as I could not
remember exectally how it was spelled or what it was called.
The Kapton really does sound right.

Yep, that is it.

http://www.dupont.com/kapton/