[TowerTalk] Measuring Coax Loss
dan-schaaf at att.net
Thu Feb 11 16:44:36 PST 2010
Having been a calibration Engineer in my early days, I concur with Jim.
In addition, each coax connector inline introduces losses. They all add up.
Definitely worse in VHF/UHF. Some folks say you can't hear the difference,
but when you are trying to hear and work a weak signal in/above the noise,
every db counts.
"In the Beginning there was Spark Gap"
----- Original Message -----
From: "Jim Brown" <jim at audiosystemsgroup.com>
To: "Tower and HF antenna construction topics." <towertalk at contesting.com>
Sent: Thursday, February 11, 2010 7:34 PM
Subject: Re: [TowerTalk] Measuring Coax Loss
> On Thu, 11 Feb 2010 19:03:51 -0500, Mike Besemer \(WM4B\) wrote:
>>Still... a dB or so (or even 3) on a long run of coax isn't going to
>> make a hill of beans difference for most applications... an S-Unit
>>is ~ 6dB.
> Only by someone's definition, not in reality. If you actually MEASURE the
> response of the S-meters in REAL radios, you find that an S-unit may be
> to 6dB near S9, but is usually closer to 3dB at S5 and below.
> As to a hill of beans -- many hams have long runs of coax to their
> so loss can be a BIG deal, not a hill of beans. The loss in dB of verious
> style coaxes ranges by a factor of about 2:1 from the lowest to the
> Over the past several years, I've been doing a lot of little things to
> improve my station. A dB here, a dB there, they all add up. Smart
> know that. The difference between a 3-el yagi and a 4-el yagi of
> design is only 1-2 dB, and often double the cost. That doesn't stop a lot
> guys who have the space from putting up 4-el yagis!
> Somehow, the scientific method seems to be lost on some of those
> on my post. How do you know that the MFJ (or any piece of test gear) is
> accurate if you don't compare it to another measurement or test method of
> known accuracy? I'll bet a six pack of your favorite 807s that if I had
> posted loss measuerements made ONLY with an MFJ, someone would have pooh-
> poohed them because I didn't use equipment traceable to a calibration lab.
>>Forgot to mention... unless I actually SAW the 10' piece being cut off the
>>longer roll I was considering buying from, I wouldn't trust the
> Did it ever occur to you that sometimes ham stores, even the best known,
> the biggest ads in QST, may not have good data on what they're selling? I
> don't trust some of them any more than an anonymous vendor in a flea
> But there ARE some good deals out there, IF you have an open mind and know
> how to evaluate them. The point of my post was to show that you CAN get
> decent data from an MFJ259B that has been calibrated if you're measuring a
> sample that is long enough.
> How many measurements of coax loas have you actually made? How did you do
> How did you know that you had good data? At some point, you've got to know
> exactly how long that piece of coax is. RG8 is big and heavy, so
> enough of it to get a good measurement isn't always easy. You've got to
> with the length of cable sample you can get.
> Jim K9YC
> TowerTalk mailing list
> TowerTalk at contesting.com
More information about the TowerTalk