RG-58U for external antenna

What kind of loss can I expect if I use about 3 feet of RG-58U cable
to connect an external antenna to an 802.11b radio? I have a lot of it
lying around from a cable TV hookup and would like to use some it if
possible. I think the characteristic impedance of RG-58U is 50 Ohms,
so that seems good.

This link shows 38.9dB per 100 feet at 2400 MHz. Am I correct in
thinking that the attenuation would be 0.389 dB per foot. I've not yet
made complete peace with logarithms - is it ok to just divide here?
http://www.universal-radio.com/catal...erf.html#atten

Also, if I use some other kind of cable with a different impedance,
how would I match it to the antenna? What would the circuit look like
and how would I implement it?

Thanks in advance for any help,
Bruce

On Sun, 02 Oct 2005 12:45:21 -0400, bjs555 <(E-Mail Removed)> wrote:

>What kind of loss can I expect if I use about 3 feet of RG-58U cable
>to connect an external antenna to an 802.11b radio?

RG-58U (solid center) with foam dielectric is 0.57dB/meter
RG-58C/U (stranded center) with foam dielectric is 0.90dB/meter.
RG-58A/U (solid center) with polyethylene dielectric is 1.5dB/meter.

I'm too lazy to convert to feet so just assume 1 meter = 3ft. Close
enough. Don't forget to add about 0.5dB loss per connector pair if
using TNC connectors and about 3 times that for crappy BNC connectors.

For reference, 6dB loss is half your range. 12dB is 1/4th your range.

>I have a lot of it
>lying around from a cable TV hookup and would like to use some it if
>possible. I think the characteristic impedance of RG-58U is 50 Ohms,
>so that seems good.

Try again. CATV is 75 ohms. It's probably RG-59/u (crap) or RG-6/u.
If you search Google Groups, you'll see some of my postings on using
RG-6/u for wi-fi and for 10base2 ethernet. It works but there are
limitations. If it's RG-59/u, forget it.

>This link shows 38.9dB per 100 feet at 2400 MHz. Am I correct in
>thinking that the attenuation would be 0.389 dB per foot. I've not yet
>made complete peace with logarithms - is it ok to just divide here?
>http://www.universal-radio.com/catal...erf.html#atten

No logarithms involved. If it's 38.9dB/100ft, then it's 1/100th of
that per foot or 0.389dB/ft.

>Also, if I use some other kind of cable with a different impedance,
>how would I match it to the antenna? What would the circuit look like
>and how would I implement it?

It would be an odd multiple of a 1/4 wave electrical line section with
an impedance equal to sqrt(Z1 * Z2) where Z1 and Z2 are the load and
line impedances. For 50 ohms to 75 ohms, that's 61 ohms. At 2.4Ghz,
building one of these is tricky as the velocity factor of the matching
section is not very well defined for random pieces of junk coax.

If it's good quality RG-6/u satellite grade coax, just use it directly
with no matching section. Going from an F connector to a TNC is
possible with adapters. The 1.5:1 mismatch loss (2 * 0.18dB) will be
less than the extra connector (0.5dB) losses.

--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

Thanks again Jeff. Your knowledge and willingness to share it are very
much appreciated.

Bruce


On Sun, 02 Oct 2005 10:13:04 -0700, Jeff Liebermann
<(E-Mail Removed)> wrote:

>On Sun, 02 Oct 2005 12:45:21 -0400, bjs555 <(E-Mail Removed)> wrote:
>
>>What kind of loss can I expect if I use about 3 feet of RG-58U cable
>>to connect an external antenna to an 802.11b radio?
>
>RG-58U (solid center) with foam dielectric is 0.57dB/meter
>RG-58C/U (stranded center) with foam dielectric is 0.90dB/meter.
>RG-58A/U (solid center) with polyethylene dielectric is 1.5dB/meter.
>
>I'm too lazy to convert to feet so just assume 1 meter = 3ft. Close
>enough. Don't forget to add about 0.5dB loss per connector pair if
>using TNC connectors and about 3 times that for crappy BNC connectors.
>
>For reference, 6dB loss is half your range. 12dB is 1/4th your range.
>
>>I have a lot of it
>>lying around from a cable TV hookup and would like to use some it if
>>possible. I think the characteristic impedance of RG-58U is 50 Ohms,
>>so that seems good.
>
>Try again. CATV is 75 ohms. It's probably RG-59/u (crap) or RG-6/u.
>If you search Google Groups, you'll see some of my postings on using
>RG-6/u for wi-fi and for 10base2 ethernet. It works but there are
>limitations. If it's RG-59/u, forget it.
>
>>This link shows 38.9dB per 100 feet at 2400 MHz. Am I correct in
>>thinking that the attenuation would be 0.389 dB per foot. I've not yet
>>made complete peace with logarithms - is it ok to just divide here?
>>http://www.universal-radio.com/catal...erf.html#atten
>
>No logarithms involved. If it's 38.9dB/100ft, then it's 1/100th of
>that per foot or 0.389dB/ft.
>
>>Also, if I use some other kind of cable with a different impedance,
>>how would I match it to the antenna? What would the circuit look like
>>and how would I implement it?
>
>It would be an odd multiple of a 1/4 wave electrical line section with
>an impedance equal to sqrt(Z1 * Z2) where Z1 and Z2 are the load and
>line impedances. For 50 ohms to 75 ohms, that's 61 ohms. At 2.4Ghz,
>building one of these is tricky as the velocity factor of the matching
>section is not very well defined for random pieces of junk coax.
>
>If it's good quality RG-6/u satellite grade coax, just use it directly
>with no matching section. Going from an F connector to a TNC is
>possible with adapters. The 1.5:1 mismatch loss (2 * 0.18dB) will be
>less than the extra connector (0.5dB) losses.