Comparing Coaxial Loss in RG-58 and RG-316

by Eric, WD8RIF

I recently decided that I would operate portable during the upcoming ARRL 10 Meter Contest and would combine this operation with a pair of Parks on the Air (POTA) activations, one on the Saturday of the contest and one on Sunday. While putting together my field station for this, I realized I wanted to see how lossy my RG-58 and RG-316 feedline cables were on 10m. Yes, I could have resorted to online charts of feedline losses, and I know that the RG-316 is lossier than the RG-58, but I thought it would be fun and more enlightening to make actual measurements, to empirically determine the losses in my particular coaxial cables.

The two coaxial cables in question are both 25′ in length and both are equipped with BNC male connectors on each end. The RG-58 cable is a high-quality cable that was originally manufactured to be a 10Base2 ethernet cable, but it’s probably now over twenty years old. The RG-316 cable was recently purchased from Tufteln (link) and includes an RF choke near one end.

To make my measurements, I used my Elecraft KX2 (link) to generate RF into an Elecraft DL1 dummy load with RF detector (link), using a digital multimeter to measure the voltage at the DL1’s measurement points. I used the formula that came with my DL1 assembly instructions to calculate the measured wattage:
P = (((V x 1.414) + 0.15))^2)/50

I made measurements in the CW portion of the 10, 15, 20, 40, and 80m bands, with the KX2 set at 5 Watts output.

My first set of measurements was made with the DL1 connected directly to the KX2’s antenna jack using a BNC union:

Direct (no feedline)
Band Volts Watts
10 11.62 5.498
15 11.59 5.470
20 11.61 5.489
40 11.63 5.508
80 11.63 5.508

My immediate observation was that the KX2 appears to be generating more than 5w when it is set to be producing 5w but also that the KX2 output is essentially same from 10m through 80m.

My second set of measurements was made with the DL1 connected to the KX2 through the RG-58 cable:

Direct RG-58
Band Watts Watts
10 5.498 4.809
15 5.470 4.844
20 5.489 5.004
40 5.508 5.184
80 5.508 5.276

My third set of measurements was made with the DL1 connected to the KX2 through the RG-316 cable:

Direct RG-316
Band Watts Watts
10 5.498 4.190
15 5.470 4.322
20 5.489 4.507
40 5.508 4.774
80 5.508 4.959

I was pleased to see to see that both the RG-58 and RG-316 behaved as I expected them to: the loss increased with increasing frequency.

Using the magic of MS-Excel, I created a table of Loss in Watts for both cables, relative to direct connection and to each other:

Direct Difference (Loss) (Watts)
Band Watts RG-58 vs
RG-316 vs
RG-316 vs
10 5.498 0.690 1.309 0.619
15 5.470 0.626 1.148 0.522
20 5.489 0.485 0.982 0.497
40 5.508 0.324 0.734 0.410
80 5.508 0.232 0.549 0.317

In looking at this table, it was immediately obvious that RG-316 is much lossier than the RG-58, particularly on 10m. By looking at the row for 10m, one can see that I am losing nearly 0.7 watts in the RG-58 but I am losing over 1.3 watts in the RG-316.

For completeness, I added columns for Loss in dB to the spreadsheet:

Difference (Loss) (dB)
Band RG-58 vs direct RG-316 vs direct RG-316 vs RG-58
10 0.582 1.180 0.598
15 0.528 1.023 0.495
20 0.402 0.856 0.454
40 0.263 0.621 0.358
80 0.187 0.456 0.269

This exercise showed me that for my upcoming ARRL 10 Meter Contest POTA outings, I would do best by connecting my antenna directly to my transceiver, if possible, without using either coaxial feedline. If conditions at the operating site require me to use feedline, I will chose the RG-58 over the RG-316.

The tables also tell me that RG-316 is pretty lossy regardless of the band; for my regular field operating, unless I’m planning to do bicycle- or pedestrian-portable operations where weight and bulk is a consideration, I’ll carry RG-58 instead of RG-316. (I purchased the RG-316 specifically for bicycle- and pedestrian-portable operations, and I plan to continue to use the RG-316 for those applications.)

At some point, I will repeat this exercise with RG-8X, a feedline that is very close in size to RG-58, is less lossy, but is also heavier and stiffer.

40 thoughts on “Comparing Coaxial Loss in RG-58 and RG-316”

  1. This is awesome as I’ve often wondered what the true losses are in RG-316. I use it quite often for the convenience of compact size. It’s not great but at the same time not horrible. The thing is this, is the station going to notice the loss on the other end? In reality they probably won’t notice that .619 watts/.598db of loss between the 2 coaxes. When you compare 1 S-Unit which is 6db of loss I really doubt the stain on the other end is going to notice.

    1. Thanks for the comment, Mark.

      As I said in the article, I’ll continue to use the RG-316 for bicycle-portable and pedestrian-portable operations, when weight and bulk are concerns. I’ll use the RG-58 for car-portable operations.

      I should have said in the article that for nearly all of my field operations, I use no feedline whatsoever, with the antenna connected directly to the transceiver.


    2. What this also means is: given a 25′ RG-58, and a 10′ RG-316, the RG-316 would actually be (slightly) better.

      If going on foot, or installing in a car, the smaller distances make the performance hit even less relevant.

      1. Indeed. My car VHF/UHF antenna is fed with approximately 12′ of something very much like RG-316 and I’ve noticed no problems with it on either transmit or receive. (I’m scared to run the numbers for RG-316 on 70cm, though!)

  2. “Pretty lossy”…. Really? Going by your figures, there’s only 0.6dB difference between the two at 28 MHz! My take on those results is that, for portable use (SOTA etc), the difference is negligible at those frequencies and cable lengths. The 316 more than makes up for its very minor losses by being considerably lighter and easier to pack than the 58.

    The choice really boils down to whether or not the cable needs to be rugged or lightweight.


    1. Hello, Ronan.

      Thanks for your comment. I will certainly use the RG-316 when weight and bulk are concerns. However, when weight and bulk aren’t a concern, why wouldn’t one choose to the use the slightly less lossy feedline?


  3. Eric-

    A good study and a fine write up!

    That first set of measurements shows a surprisingly ‘flat’ sensitivity over frequency. My casual tinkering with RF detectors years ago would show a considerable falloff on the high bands. Not surprisingly, Elecraft has done it well.

    I used to primitive-camp in the Green Mountain National Forest in Vermont. I had a grassy site with two conveniently-placed trees. They hosted a 20M dipole up 35 feet. It was fed with a 50-foot length of RG-174. Jameco offered that cable complete with male BNCs at each end.

    Lossy? Sure- about 2dB on 20M. I knew that going in, but worth it to me for its convenience. I still made plenty of contacts in the QRP contests of that era. We all learn as we go, and I’d do it differently now.

    Best of luck with the bicycle and pedestrian-portable activities! 73- K1SWL

    1. Thanks for the comment, Dave.

      I, too, started doing field operations using single-band dipoles fed with RG-174. In my case, the feedline was only 30′ long.


  4. Hi Eric – I congratulate you on a nice piece of work. I believe that coaxial cable losses get worse when I deploy a non-resonant antenna (an EFRW, for example). That makes your conclusion about direct connection to the KX2 even more relevant.

    1. Hello, Brian.

      Yes–feedline loss rises dramatically with increasing SWR. I made no attempt to measure this loss since I’ll be using coax feedlines only with (nearly) resonant antennas. I always connect my non-resonant end-fed antennas directly to the ATU or ATU-equipped transceiver. For non-resonant doublets or loops, I use parallel-line feedline which is less lossy than coax in high-SWR situations.

      The nifty online tool ( mentioned below by Jeff allows one to calculate feedline loss with SWR.


  5. Great job Eric.

    I would be curious to see how RG174 does in the same test.

    But like Ronan, for me it boils down to weight(size) versus ruggedness and how much room is in the go kit. When operating around the yard it’s no problem carrying a heavier cable outside.

    I’m not sure if the loss would be that noticeable on the remote side but I’m thinking it would all depend on how bad the bands are. And we all know that every bit helps.


    1. Hello, Marshall.

      I don’t have a long RG-174 feedline and, thus, can’t do direct measurement. However, the tool mentioned below by Jeff,, allows one to compare calculated loss for RG-174 and RG-316. Using this tool configured for 25′ of RG-174 and RG-316 feeding matched loads on 28MHz, the calculated loss for RG-174 is 1.082dB and the calculated loss of RG-316 is 1.148dB, so the two feedlines are very close, at least on 10m.


  6. Great work.
    The difference between the 2 may come down to being heard or not.
    I have some high quality RG-174 that I use on short runs, but generally speaking use RG-58.
    EFHW I tend to connect straight to my KX2.


    1. Hello, Jeff.

      Thanks for the link to the online calculator. I’m going to be sure to bookmark this tool.

      According to what I’ve read, and this is confirmed by the online calculator, loss in coax rises dramatically with increasing SWR. I wouldn’t feed any high-SWR antenna with coax except, perhaps, in an emergency. My end-fed random wire antennas connect directly to the output of an ATU or ATU-equipped transceiver, and my non-resonant balanced antennas are connected to the ATU through balanced line.

      As mentioned in the article, for portable operations when weight or bulk are concerns, when I’m using resonant antennas, I’ll happily use RG-316, even on 10m. For car-portable operations, I’ll chose RG-58 when I use resonant antennas.


  7. Indeed, I’m not sure anyone can hear the difference in a half-db weaker signal. However, that may increase when you have a real portable antenna that is not going to be exactly 50ohms and zero reactance like a dummy load. I’ve thought about doing this type of experiment with an antenna and the tinysa measuring field strength a fixed distance away. Lots of error in that settup maybe, but it would account for losses in a real antenna and feedline configuration.

    If it really came down to a half-db, you are better off worrying about things like feedpoint height or adding more radials to a vertical. There will be other places where the effort might be better spent.

    1. Hello, Jason.

      I’d be interested in seeing the results of your real-world tests if you’re able to do them.

      If I were doing a regular POTA activation, when I’m running a frequency, I wouldn’t worry much at all about the increased loss of the RG-316 over RG-58–or, indeed, of the RG-316 over a direct connection–since the POTA hunters will really work to make the QSO with the activator. However, this coming weekend, for ARRL 10 Meter Contest, my plan is to hunt for QSOs and, since I’ll be competing with other hunters running 100w or even 1kw, I want the best signal I can get with the equipment I’m using. (I have doubts that I would be able to hold a frequency while running QRP into a portable antenna during a major contest.)


  8. Eric – Tnx for a data oriented write up. You and I are on the same page for this weekend. We’ll do portable QRP CW camping out at Kerr Lake State Recreation Area K-3845 (NC-VA border).
    Wonder how RG-174 (the smallest I use) compares to RG-316?
    You mentioned the RG-58 was about 20 yrs old. I just replaced old RG-58 at the beach and noticed a nice improvement. The coax had signal loss from UV damage.
    Hpe to work you this weekend.
    73 de K4RLC Bob

    1. Hello, Bob.

      Good luck at Kerr Lake this weekend!

      I used the tool provided by Jeff, above, to compare RG-174 and RG-316 (see my comment, above). The calculator showed RG-174 and RG-316 to be very close in loss at 28MHz.

      I have wondered if my old-but-unweathered RG-58 is significantly lossier than new RG-58 would be. I plan to make a new 25′ RG-58 feedline and, after I do, I will measure its loss. (I already have the coax; I’ll order BNC connectors this week or next.)


    1. Hello, Paul.

      You know, I actually have a nanoVNA, but I’ve not yet used it enough to have gotten comfortable with it. Learning how to use the thing–and learning what the thing is capable of doing–is on my list of things to accomplish 🙂


      1. It’ll measure almost anything 🙂

        The measurement of loss is best displayed in the following link. Just do it for 3 to 30 MHz instead:

        Have fun! With software attached you can capture the loss graph with ease. If your coax is old and water damaged you can not only measure the loss but also what it’s done to the impedance of the cable. With TDR testing you can see where a cable is damaged and cut out the bad bit….

  9. Thanks, that’s a nice comparison. I did a test of my once when I forgot to show my RG8X into the truck and had to link together two 25 ft pieces of 316 to reach the antenna. Yikes! Whispering into the wind that day. My favorite these days is to use antennas that hook directly to the radio to eliminate feed lines. Lately, experiments with open feed line has been doing well
    Thanks for your work on testing this part of the equation. 73 FB

  10. Many THANKS for the experiment. I too like to see these sorts of real life measurements. For some “way down in the noise” QSOs the loss might make a difference in being heard, but we never know, do we?
    In the end, your experiment matches almost exactly the loss specifications shown for those coax types at places that sell them. I often wonder if all those who rave about RG-316 really know how lossy it is. You’ve shown it with real comparisons.

  11. Wow Brian, you opened up a can of worms….

    Good to see so many comments. I applaude your curiousity and methodologies. And, I’ve enjoyed reading the varied opinions.

    I liked what Roman had to say in respect to dB’s. I like RG316, and usually only run 15’ or less…

    Antenna deployment has a more direct impact on signal strength. But to each his own… Thanks for the post!

    de W7UDT

  12. Great job Eric,
    As I mentioned earlier, I really like real life experiments. Now, since you have both 25′ cables on hand, can you tell us what each of them weighs?

    Is it really the case that people who are in good enough condition to climb hills find 25′ of RG-58 too much to carry?

    1. Hello, Bob.

      That’s a very interesting question! I’ll weigh the two cables right now.

      RG-316: 5-3/8oz or 154g
      RG-58: 10-3/4oz or 310g

      So, my RG-58 feedline is roughly twice as heavy as my RG-316 feedline.

      My RG-316 feedline includes an RF choke on one end; this will add a bit of weight to it, so an RG-316 feedline without the choke will weigh even less.

      I don’t hike as much with my station as I’d like, and I haven’t yet done any serious hiking up mountains, so I won’t be one to say that the RG-58 is too heavy for a serious SOTA climb.

      I will say that the RG-316 rolls up into a much less bulky roll than the RG-58 does, and if I were building a complete but tiny station for bicycle-portable use–as I will be doing once my QMX arrives–I’ll be using the RG-316 because the smaller feedline packs so much smaller.


  13. If I lessen or eliminate my RG-316 will that affect the resononance of my EFHW? I have often thought that the feedline is an important piece to making an EFHW resonant, am I wrong? Thanks for all of the insight, I love RG-316! Not only is it light an easy to carry, it is also very very easy to close a car door on top of it. 73 de Jon K7CO

    1. Hello, Jon.

      I have only limited experience with the EFHW antenna but I can say that I purchased my 25′ RG-316 feedline with integrated choke specifically to use with an EFHW and a QMX transceiver that doesn’t have an internal antenna tuner, a QMX that, unfortunately, won’t be delivered until June, I’m afraid.

      It’s my understanding that the EFHW needs some sort of a “return” wire for best performance, and most operators rely on the outside of their coax’s shield for this purpose. It’s also my understanding that in most cases the EFHW can be deployed without a coax feedline if a dedicate counterpoise wire is used to provide the “return”. I plan to test this myself at some point, perhaps after I receive my QMX.

      Perhaps Thomas, K4SWL, can provide additional thoughts on deploying an EFHW without coaxial feedline.


  14. Eric, congratulations on the excellent analysis (and resulting many super comments). Let me toss in one more:

    For context, my wife KB1IFZ and I love one-watt QRPing – whether on DXpeditions, POTA activations, or from our home QTH. I am a long-time (since ’74) ham and expert antenna designer and modeler, and hold numerous antenna patents (and have licensed designs to mfgs in our fine hobby).

    When every fraction of a watt counts, there are very few reasons to use coax of any type. Use ladder line.

    For the sake of all readers’ eyes, I won’t turn this into a 1,000-word research paper. The following are three top-line points for quick reference and further reading:

    (1) Maximum antenna-system efficiency. For an easy numbers’ benchmark, consider a 100-ft run of RG-8X vs 450-ohm ladder line and an antenna fed from the TX with 5 watts. Take a look at what actually gets to the antenna (numbers rounded):

    3.8 MHz, 1:1 SWR —
    RG-8X (4.4 watts), 450-ohm Ladder Line (4.9 watts)

    3.8 MHz, 6:1 SWR —
    RG-8X (3.6 watts), 450-ohm Ladder Line (4.8 watts)

    14.2 MHz, 1:1 SWR —
    RG-8X (3.9 watts), 450-ohm Ladder Line (4.9 watts)

    14.2 MHz, 6:1 SWR —
    RG-8X (2.8 watts), 450-ohm Ladder Line (4.7 watts)

    29.0 MHz, 1:1 SWR —
    RG-8X (3.4 watts), 450-ohm Ladder Line (4.8 watts)

    29.0 MHz, 6:1 SWR —
    RG-8X (2.2 watts), 450-ohm Ladder Line (4.5 watts)

    As others have commented here, the following fine tool by KV5R gives hams the ability to perform quick comparisons among lots of combinations of feedline type, power, and SWR:

    (2) This excellent piece also by KV5R dispels common ladder line myths:

    (3) Use a 1:1 current balun (not 4:1 or others) as close to the rig as possible (usually as a transition to a short, few feet, run of coax going to the rig). Us hams should not try to use the balun’s design ratio to match the ladder line’s impedance with the rig’s or coax’s impedance. The following is a superb piece as to why. Scroll to the part that begins: “DO NOT match the balun impedance to the transmission line impedance.”

    Again, very nice work in your meticulous research and analysis. I really enjoy when fellow hams roll up their sleeves and do fundamental experiments like this. It’s what our wonderful hobby is all about!

    I hope my info is helpful. Happy QRPing!

    73, Eric KB1EHE

    1. Hello, Eric.

      Thank you for your comments and for providing those excellent resources.

      I’m a big fan of window-line. My main hamshack antenna is a 135′ doublet fed with 450-ohm window-line all way to a venerable EF Johnson Matchbox balanced antenna tuner on my radio desk. This antenna system provides good service on all the bands from 10m through 80m.

      In the field, I tend to avoid using dipoles or doublets because they’re slower to deploy than an end-fed wire. The times I have deployed a non-resonant doublet in the field, they have been fed with parallel feedline in order to provide efficient multiband service.


  15. All,

    I’m not sure why my previous comment didn’t come out with paragraph breaks! What a mess! It looked fine when I pressed the POST COMMENT button.

    I hope the content is decipherable!

    73, Eric KB1EHE

    1. Your comment looked fine in my browser (Chrome) except the lines in the table were forced to wrap. I’ve reformatted the table somewhat to improve readability.


  16. This is outstanding! You could have just looked this up in a book or on the Internet, or pushed a couple of buttons on an instrument. Instead you designed and conducted an experiment, and collected, analyzed, and published data. This is true Amateur spirit in action!
    A while back I was curious as to how much of my KX2’s 10 watts was making it to the other end of my 25 foot piece of RG-174. However, I just used my antenna analyzer to do a return loss measurement. (I never would have thought to do that had it not been for one of Dave Casler’s columns in QST.) In my case my reaction was quite different from yours, though: “That’s ALL I’m losing, even at 30 MHz? Cool!” 😉

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.