## Re: coax capacitor

Expand Messages
• I knew that. I just didn t think about it :-) Old age is the pits! Bill W5EC ... capacitance per ... foot of
Message 1 of 16 , Nov 10, 2005
• 0 Attachment
I knew that. I just didn't think about it :-)
Old age is the pits!
Bill W5EC

--- In loopantennas@yahoogroups.com, Daniel Reynolds <aa0ni@y...>
wrote:
>
> Hi Bill,
>
> There's not really a formula. Coax has a constant amount of
capacitance per
> foot.
>
> The coax you refer to is easily looked up on belden cable's website
> (www.belden.com). RG-8X has approximately 25 pF of capacitance per
foot of
> cable. If you need 100 pF ... just cut a length a little longer
• Just a question: is the use of coax as a capacitor not limited to lengths that are (very) short in relation to the wavelength involved? regards, Cor
Message 2 of 16 , Nov 10, 2005
• 0 Attachment
Just a question:
is the use of coax as a capacitor not limited to lengths that are (very) short in relation to the wavelength involved?

regards,

Cor Beijersbergen van Henegouwen

[Non-text portions of this message have been removed]
• capacitance is: RG6 16.2 pF per ft RG 59 16.3pF RG58 30.8pF However, co-ax is NOT just capacitance: the series inductance comes into play (which is where the
Message 3 of 16 , Nov 10, 2005
• 0 Attachment
capacitance is:
RG6 16.2 pF per ft
RG 59 16.3pF
RG58 30.8pF

However, co-ax is NOT just capacitance: the series inductance comes into
play (which is where the impedance comes from) and so unless less than 1/10
of a wavelength long a cable cannot be considered to be a lumped
capacitance: this severely limits the application of coax as a variable cap.

Alan Ainslie

----- Original Message -----
From: "C. Beijersbergen" <c.beijersbergen@...>
To: <loopantennas@yahoogroups.com>
Sent: Thursday, November 10, 2005 9:26 PM
Subject: Re: [loopantennas] Re: coax capacitor

> Just a question:
> is the use of coax as a capacitor not limited to lengths that are (very)
> short in relation to the wavelength involved?
>
> regards,
>
> Cor Beijersbergen van Henegouwen
>
> [Non-text portions of this message have been removed]
>
>
>
>
>
> For uploading images, I prefer the Files section since Photos only allows
> everyone (except the uploader and moderators) to see a max of 300x400.
> http://groups.yahoo.com/group/loopantennas/files
>
> Put them in the appropriate folder, or create one.
>
>
>
>
>
>
>
• You are correct, Alan, however in order to get more capacitance you can put several small pieces (
Message 4 of 16 , Nov 10, 2005
• 0 Attachment
You are correct, Alan, however in order to get more capacitance you can
put several small pieces (<1/10 wavelength) in parallel with each
other. As you know, capacitors in parallel add for the total
capacitance. The "impedance" of the coax isn't important if you are
planning to use it as a capacitor rather than a feedline. For the
example capacitances listed below maybe 3 pieces of RG58 that are 1'
long in parallel (connect all the shields together and all the center
conductors together at one end) should be about 92.4 pF if a single 3'
long piece is not practical otherwise.

Also, if the coaxial cable capacitor is going to be used for Tx
applications, pay attention to the voltage breakdown specification.
Maybe a larger size coax would be more suitable, say RG8 or RG213 or
even 1/2" or 7/8" hardline possibly if high power transmitting is
planned. Another thing to keep in mind is that the voltage would like
to most easily flash over at a flush cut end, so you may want to remove
part of the shield for about 1" from the end to give it extra breakdown
voltage capability.

Mark, NK8Q

Alan Ainslie wrote:

>capacitance is:
>RG6 16.2 pF per ft
>RG 59 16.3pF
>RG58 30.8pF
>
>However, co-ax is NOT just capacitance: the series inductance comes into
>play (which is where the impedance comes from) and so unless less than 1/10
>of a wavelength long a cable cannot be considered to be a lumped
>capacitance: this severely limits the application of coax as a variable cap.
>
>Alan Ainslie
>
>----- Original Message -----
>From: "C. Beijersbergen" <c.beijersbergen@...>
>To: <loopantennas@yahoogroups.com>
>Sent: Thursday, November 10, 2005 9:26 PM
>Subject: Re: [loopantennas] Re: coax capacitor
>
>
>
>
>>Just a question:
>>is the use of coax as a capacitor not limited to lengths that are (very)
>>short in relation to the wavelength involved?
>>
>>regards,
>>
>>Cor Beijersbergen van Henegouwen
>>
>>[Non-text portions of this message have been removed]
>>
>>
>>
>>
>>
>>For uploading images, I prefer the Files section since Photos only allows
>>everyone (except the uploader and moderators) to see a max of 300x400.
>>http://groups.yahoo.com/group/loopantennas/files
>>
>>Put them in the appropriate folder, or create one.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
>
>
>
>
>
>For uploading images, I prefer the Files section since Photos only allows everyone (except the uploader and moderators) to see a max of 300x400.
>http://groups.yahoo.com/group/loopantennas/files
>
>Put them in the appropriate folder, or create one.
>
>
>
>
>
>
>
>
>

[Non-text portions of this message have been removed]
• ... An interesting question. If you had lossless coax, or didn t care about the loss (which reduces Q), the length need not be very short compared to a
Message 5 of 16 , Nov 13, 2005
• 0 Attachment
> is the use of coax as a capacitor not limited to lengths that are (very)
> short in relation to the wavelength involved?

An interesting question.

If you had lossless coax, or didn't care about the loss (which reduces Q),
the length need not be very short compared to a wavelength.

However, one of the things to be aware of, is that the effective input
capacitance is NOT a linear function of length, nor is it constant vs.
frequency! The DC / very low frequency capacitance (*much* shorter than a
wavelength) is indeed proportional to length, but as the length starts to
become not so short electrically, the effective capacitance starts going up
faster than the length does.

Thus, three 1-foot long pieces in parallel, has different capacitance than
one 3-foot long piece. Except at 0 Hz, or at frequencies much less than a
wavelength.

When the length reaches a quarter wavelength long, capacitance becomes
infinite (resonance), and above that, it's inductive.

I guess this only matters if you measured some length and expected a
specific capacitance. Or if you needed a truly frequency-independent
capacitor.

I've not heard of using coax as a variable capacitor before. Not sure I
understand the mechanics.

Regards,
Andy
• This post is wrong. The other many posts that explained that coax is a transmission line with distributed inductance and capacitance are correct. The poster
Message 6 of 16 , Nov 13, 2005
• 0 Attachment
This post is wrong. The other many posts that explained that coax is a
transmission line with distributed inductance and capacitance are correct.
The poster who explained to use short pieces in parallel to overcome this
was insightful.

on 11/13/05 3:40 PM, Andy at ingraham.ma.ultranet@... wrote:

> is the use of coax as a capacitor not limited to lengths that are (very)
> short in relation to the wavelength involved?

An interesting question.

If you had lossless coax, or didn't care about the loss (which reduces Q),
the length need not be very short compared to a wavelength.

However, one of the things to be aware of, is that the effective input
capacitance is NOT a linear function of length, nor is it constant vs.
frequency! The DC / very low frequency capacitance (*much* shorter than a
wavelength) is indeed proportional to length, but as the length starts to
become not so short electrically, the effective capacitance starts going up
faster than the length does.

Thus, three 1-foot long pieces in parallel, has different capacitance than
one 3-foot long piece. Except at 0 Hz, or at frequencies much less than a
wavelength.

When the length reaches a quarter wavelength long, capacitance becomes
infinite (resonance), and above that, it's inductive.

I guess this only matters if you measured some length and expected a
specific capacitance. Or if you needed a truly frequency-independent
capacitor.

I've not heard of using coax as a variable capacitor before. Not sure I
understand the mechanics.

Regards,
Andy

For uploading images, I prefer the Files section since Photos only allows
everyone (except the uploader and moderators) to see a max of 300x400.
http://groups.yahoo.com/group/loopantennas/files

Put them in the appropriate folder, or create one.

antenna
sales

" on the web.

To unsubscribe from this group, send an email to:
loopantennas-unsubscribe@yahoogroups.com
<mailto:loopantennas-unsubscribe@yahoogroups.com?subject=Unsubscribe>

<http://docs.yahoo.com/info/terms/> .

[Non-text portions of this message have been removed]
• ... I don t know why you say my post is wrong. I agree completely that coax is a transmission line and has distributed capacitance and inductance. I was just
Message 7 of 16 , Nov 14, 2005
• 0 Attachment
> This post is wrong. The other many posts that explained that coax is a
> transmission line with distributed inductance and capacitance are correct.
> The poster who explained to use short pieces in parallel to overcome this
> was insightful.

I don't know why you say my post is wrong.

I agree completely that coax is a transmission line and has distributed
capacitance and inductance. I was just referring to the effective terminal
capacitance when looking at a length of coax from one end, i.e., using a
length of coax as a 2-terminal circuit element.

This effective capacitance, at any (non-zero) frequency F, is *not* equal to
the capacitance-per-unit-length times the length, because of those
transmission line effects and the interaction with the distributed
inductance. It will be somewhat greater, as it turns out (unless the length
exceeds a quarter wavelength).

This is borne out by transmission line theory, and I have shown it to myself
and others over and over again using circuit simulations over the past 20
years.

If the length is very much smaller than a quarter wavelength, then Cin will
be very close to (C/length * length), as you might have expected it to be.

But if the length is not quite so small, then Cin != (C/length * length).

Indeed, that's another reason why using several short pieces in parallel
instead of one long one, can help. You get closer to what you calculated;
not something different because of those transmission line effects.

Long coax lengths do work too, but they don't give you the capacitance you
thought you had. And generally that's why you want to avoid them.

Regards,
Andy
• We may be wavering from the original post here, but I agree with Andy. Coax has both inductance and capacitance as does any other transmision line. Consider
Message 8 of 16 , Nov 14, 2005
• 0 Attachment
We may be wavering from the original post here, but I agree with Andy.
Coax has both inductance and capacitance as does any other
transmision line.

Consider the geometry of a piece of coax.

The capacitance is generated by the surface area of the shield verses
the surface area of the center conductor with the dielectric constant
of the material between the two. This is analogous to the 2 plates of
a capacitor with its dielectric sandwiched between, with an electric
field across the two). The difference is that the coax capacitance is
long compared to a traditional fixed or variable capacitor.

The inductance part of a transmission line is produced by the length
of the coax (current travelling along a conductors produces a magnetic
field).

Because the coax is uniform in its construction, a relatively simple
mathematical formula expresses the two quantities and relates them to
the native impedance of the particular coax.

The idea that the capacitance per unit length is not constant stems
from the non-linear relationship between the capacitance and
inductance of a transmission line.

Here is a link to a simplified explanation:
http://tinyurl.com/7m2us

Dave WA6YSO

--- In loopantennas@yahoogroups.com, "Andy"
<ingraham.ma.ultranet@r...> wrote:
>
> > This post is wrong. The other many posts that explained that coax
is a
> > transmission line with distributed inductance and capacitance are
correct.
> > The poster who explained to use short pieces in parallel to
overcome this
> > was insightful.
>
> I don't know why you say my post is wrong.
>
> I agree completely that coax is a transmission line and has distributed
> capacitance and inductance. I was just referring to the effective
terminal
> capacitance when looking at a length of coax from one end, i.e., using a
> length of coax as a 2-terminal circuit element.
>
> This effective capacitance, at any (non-zero) frequency F, is *not*
equal to
> the capacitance-per-unit-length times the length, because of those
> transmission line effects and the interaction with the distributed
> inductance. It will be somewhat greater, as it turns out (unless
the length
> exceeds a quarter wavelength).
>
> This is borne out by transmission line theory, and I have shown it
to myself
> and others over and over again using circuit simulations over the
past 20
> years.
>
> If the length is very much smaller than a quarter wavelength, then
Cin will
> be very close to (C/length * length), as you might have expected it
to be.
>
> But if the length is not quite so small, then Cin != (C/length *
length).
>
> Indeed, that's another reason why using several short pieces in parallel
> instead of one long one, can help. You get closer to what you
calculated;
> not something different because of those transmission line effects.
>
> Long coax lengths do work too, but they don't give you the
capacitance you
> thought you had. And generally that's why you want to avoid them.
>
> Regards,
> Andy
>
• The capacitance per unit distance is what it is independent of frequency. No way of knowing, but I m guessing you measured input impedance at some frequency
Message 9 of 16 , Nov 14, 2005
• 0 Attachment
The capacitance per unit distance is what it is independent of frequency.
No way of knowing, but I'm guessing you measured input impedance at some
frequency and it was lower than you would calculate for that frequency.
That's the distributed capacitance and inductance tanking.

on 11/14/05 9:32 AM, Andy at ingraham.ma.ultranet@... wrote:

> This post is wrong. The other many posts that explained that coax is a
> transmission line with distributed inductance and capacitance are correct.
> The poster who explained to use short pieces in parallel to overcome this
> was insightful.

I don't know why you say my post is wrong.

I agree completely that coax is a transmission line and has distributed
capacitance and inductance. I was just referring to the effective terminal
capacitance when looking at a length of coax from one end, i.e., using a
length of coax as a 2-terminal circuit element.

This effective capacitance, at any (non-zero) frequency F, is *not* equal to
the capacitance-per-unit-length times the length, because of those
transmission line effects and the interaction with the distributed
inductance. It will be somewhat greater, as it turns out (unless the length
exceeds a quarter wavelength).

This is borne out by transmission line theory, and I have shown it to myself
and others over and over again using circuit simulations over the past 20
years.

If the length is very much smaller than a quarter wavelength, then Cin will
be very close to (C/length * length), as you might have expected it to be.

But if the length is not quite so small, then Cin != (C/length * length).

Indeed, that's another reason why using several short pieces in parallel
instead of one long one, can help. You get closer to what you calculated;
not something different because of those transmission line effects.

Long coax lengths do work too, but they don't give you the capacitance you
thought you had. And generally that's why you want to avoid them.

Regards,
Andy

For uploading images, I prefer the Files section since Photos only allows
everyone (except the uploader and moderators) to see a max of 300x400.
http://groups.yahoo.com/group/loopantennas/files

Put them in the appropriate folder, or create one.

" on the web.

To unsubscribe from this group, send an email to:
loopantennas-unsubscribe@yahoogroups.com
<mailto:loopantennas-unsubscribe@yahoogroups.com?subject=Unsubscribe>

<http://docs.yahoo.com/info/terms/> .

[Non-text portions of this message have been removed]
• ... I think we are in agreement. While capacitance per length is frequency independent, the effective input capacitance seen looking into a fixed length of
Message 10 of 16 , Nov 15, 2005
• 0 Attachment
> The capacitance per unit distance is what it is independent of frequency.
> No way of knowing, but I'm guessing you measured input impedance at some
> frequency and it was lower than you would calculate for that frequency.
> That's the distributed capacitance and inductance tanking.

I think we are in agreement. While capacitance per length is frequency
independent, the effective input capacitance seen looking into a fixed
length of coax is not. That's was what I was saying; it's why a coax
capacitor ought to be electrically short, to work the way you expected.

It also means that the simple formula on the webpage previously given
(http://www.westbay.ndirect.co.uk/capacita.htm) is not correct, because they
forgot to include the effect of frequency (and inductance).

As the frequency approaches that where the length is a quarter-wavelength,
with an open circuit on the far end, the impedance looking into the cable
approaches zero. It's a capacitive reactance (below resonance). That means
the effective capacitance looking into the stub is increasing, rapidly. The
impedance (reactance) approaches zero or the effective capacitance
approaches infinite at the quarter-wavelength frequency, even though the
length is not infinite.

This effect also makes a transmission line act like a kind of capacitance
multiplier. For example, a length of PC board trace with a distributed
capacitance of 5 pF, that has a 20 pF capacitor hanging on the end of it (or
an IC with a 20 pF pin capacitance), looks like 5 + 20 = 25 pF at low
frequencies only, but much more than 25 pF as the frequency goes up and
the electrical length of the trace becomes significant. Again, a frequency
dependent capacitance.

Above the quarter-wavelength frequency, the reactance goes inductive.

Regards,
Andy
Your message has been successfully submitted and would be delivered to recipients shortly.