Coax Feedline Causing Loading Cap to Arc

[Todd, KA1KAQ]

The subject actually sounds reasonable when I read it, but it wasn't even a thought before.

Some may recall that last September the big rig took a hiatus when we were operating just before the Shelby 'fest, starting with a large arc in the RF compartment followed a while later by repeated line fuse consumption. Subsequent checking and testing revealed nothing initially, yet another annoying gremlin to chase down.

Eventually the problem was found to be a shorted splatter choke which I thought had been bypassed but still had HV on it through one side of the filter circuit. Johnny figures it went to ground after playing auto transformer for a while.

So after hipot testing through the HV supply and finally sorting this out, all seemed well.

Until last week, switching from TUNE to OPERATE while testing into the dummy load. Large zorches began to appear on the split stator loading cap. No problem in the TUNE position, but in Operate it would sometimes zorch on key up, other times on voice peaks. It wasn't liking the increase in voltage for some reason.

Checked the usual suspects: antenna relay was working, connectors were clean and tight, no critter carcasses across the plates of the cap, the large broadcast dummy load was even getting warm. A quick measurement shows it at 50.6 ohms. All meters on the rig were reading properly.

So a call returned from my RF expert pal WA2PJP (he always figures out the things that elude me) got things rolling. Joe was as confused as I was. We went through different scenarios including trying to tune up in CW mode to preclude parasitic oscillations. No joy. Then Joe suggests something pretty obvious - did you check the feedline from the rig to the dummy load? You mean are the connectors tight? No, the feedline itself - did you check it?

No. Why not? Same reason I always figure some major item burned up rather than checking the fuse first, I suppose. I'd built this jumper some years back from a piece of coax left over from a new roll I'd purchased. Belden was having a sale on it, 9913.

Anyone familiar with 9913 knows it gets its low loss characteristics from the air dielectric that replaces the foam used to insulate the center conductor in other coax. The spacing for maintaining the proper impedance is created by a swirly pigtail of flexible plastic traveling through the cable. Great stuff for VHF, sometimes referred to as Poor man's hardline.

Unfortunately, it's also prone to damage if it gets bent too sharply or even stepped on. And despite taking good care of this stuff over the years by using it in straight runs with wide bends an carefully coiling it in big rolls, this 20 foot jumper must've gotten damaged at some point. It never showed up when I was using the 32V into the dummy load due to the lower power level I guess. And since I soldered the teflon and silver plated PL-259s on it myself, I knew it was good and didn't even consider it to be a potential failure point.

Swapping in a piece of different coax yielded better results (no more arcing) and after hooking up the 80m dipole last night, I made my first 80 contact from NC back up to Long Island. Can't wait for the killer static conditions to subside for some good operating below 3800. Joe was 20-40db over S9 and at times it was hard to pick him out of the noise.

So the moral of the story is, don't discount those simple things. Especially coaxial feedline, wrapped up and sealed all nice and pretty. You can drive yourself nuts overthinking an otherwise-simple problem. I'm living proof, though the doctor sez I'm doing better...   

[WA1GFZ]

I would not use that kind of cable as patch or test cables, but then I would not use PL259s either. learned the hard way, just like you

[The Slab Bacon]

I would not use that kind of cable as patch or test cables, but then I would not use PL259s either. learned the hard way, just like you

I've burned up a few PL-259s myself. Burned to crispy ashes!! The phenolic ones are the worst. But I've always had good luck with the teflon ones if you buy GOOD ones and not the $.99 hamfest specials.

Running high powa and an extreemly low feedline impedance, I have always been leary of burning that little itty bitty pin off of type N connectors. (I fried one of them as well)

I have also seen coass do some wierd things as well.

Oh, yea, If you use a really old phenolic PL-259s for HV connectors, after a while they will go off as good as an M-80 at 3500v                 

[WA1GFZ]

I've put 5 KW through type Ns and more under pulse conditions.
Our 1 KW amp at work has a type N and we drive it into real high a vswr all the time. You have to use quality type Ns

[Jim, W5JO]

Bending radius of most feedline and, in fact, all lines should be 10 X the diameter of the cable.  If, for any reason you move cable like that you had, then make another to be safe and discard the old one.  Saves tune caps.  If you bend RG 8 or 213 excessively you distort the inner insulation and change the impedance at that point.  That isn't much of an issue at 3.9 meg but at 150 and up it is.

[w3jn]

High power tends to separate the men from the boyz with regard to substandard shack components 

[KB2WIG]

FWIW

dont drag yer coas on the ground


klc

[ke7trp]

Glad you found the problem!

I dont like 9913.  It has issues with water getting in.  Its pretty well documented.  I remember a ham club used brand new 9913 for a station.  Every year, For fun, They tested the actualy loss of this feedline when they set it up. After 3 years, It was Shot. Huge losses in the line.  did it work? Yes. Would they have noticed? probably not.  But the chart from new to 5 years really told the story. They determined that water was leaching into the 9913 during rainy days and this eventualy showed up in the loss chart.  They are now using LMR and the chart only has the new test so far.  I guess in a few years we will know how this compares

For the shack cables, I use LMR 400 ultra flex for most of them.

[The Slab Bacon]

High power tends to separate the men from the boyz with regard to substandard shack components 

Yea, what he said! ! 

[The Slab Bacon]

I dont like 9913.  It has issues with water getting in.  Its pretty well documented.  I remember a ham club used brand new 9913 for a station.  Every year, For fun, They tested the actualy loss of this feedline when they set it up. After 3 years, It was Shot. Huge losses in the line.  did it work? Yes. Would they have noticed? probably not.  But the chart from new to 5 years really told the story. They determined that water was leaching into the 9913 during rainy days and this eventualy showed up in the loss chart.  They are now using LMR and the chart only has the new test so far.  I guess in a few years we will know how this compares

9913 sucks! I have actually seen water drip out of the connectors at the shack end !

[W3SLK]

I got to chime in here since I use 9913. I've had coax with the jacket cut and placed tape over it. Now mind the biggest power generator I have is only capable of 600 watts. And this stuff is MIL-SPEC. I have to replace it before this winter.
With regards to N type connectors. Eric, WB4VVI(SK), didn't like them because he claimed the reciever for the pin would expand and contract with temperature.

[WA1GFZ]

Right on John, Then there is lightning, God is the master of lightning

[Steve - WB3HUZ]

Plain old 213 is FB for in the shack jumpers. Using low loss coax for short runs on the lower end of HF can be an expensive overkill.

A properly constructed UHF connector is rated for 500 Volts peak working voltage. At 50 Ohms, this is 5 kW. The dielectric breakdown voltage is 1500 VDC.

Working voltage rating for the N connector is 1500 Volts. The dielectric breakdown voltage is 2500 VDC. For corrugated types the voltages drop to 707 RMS and 2000 VDC respectively.

[Todd, KA1KAQ]

9913 sucks! I have actually seen water drip out of the connectors at the shack end !

It's actually pretty good stuff if used properly(which wouldn't include HF). Water is a HUGE issue with the hollow center acting as a pipeline. Inevitably, it can be traced to either an improperly installed (sealed) connector or a nick in the jacket somewhere. The trick I was shown for sealing it up is to take a small amount of silicone sealant and use it to plug the open end before installing the connector. Worked well for me, never had water issues until I hit the cable with the mower years later and cut the outside jacket. Taping and sealing did no good, it still managed to travel between the braid and jacket.

Plain old 213 is FB for in the shack jumpers. Using low loss coax for short runs on the lower end of HF can be an expensive overkill.

Absolutely, and a pain in the ass to work with, too. This cable was purchased back in 1987 when I erected my first tower, and was intended to eventually serve as a VHF line once I had that station in place. In the meantime, it would serve as the primary HF feedline from the house up to the tower on the back hill, straight shot with a flexible jumper at the top of the tower. The jumper was one of those things that happened as a result of needling something fast, and having a chunk left on the spool. Convenience vs. Common sense.

The problem comes in when time passes and you 'forget' the shortcomings of such specialized cable. It becomes just another piece of coax, with less flexibility. As contrary as it might sound, I've learned a lot since buying that cable ad making that jumper. Seems I've forgotten just about as much, too. 

A lesson (re)learned, for sure. Will be using 213 from here on out. Might save the 9913 for 6m, might not. Using what's handy might be expedient, but that doesn't mean it's a good idea.

BTW Steve, the 40m foam coax had really degraded too. When I repaired a damaged area last year, I noticed the insulation had become discolored. So I'm thinking a similar impedance/load issue is what caused that massive arc when you were operating the rig last fall. The flash certainly came from that area of the RF deck. The fuse issue didn't happen for another hour or so, and was likely unrelated, just a coincidence of timing. So the 40m aerial is scheduled for a rebuild too, with KC1BT's bullet-proof mil surplus wire used on the 160 and 80 dipoles.   

[W2VW]

The current is the killer at QRO at 50 ohms. That being said 22 kw peak visual TV channel 4 through a type N for 1/2 hour no failure.

Plain old 213 is FB for in the shack jumpers. Using low loss coax for short runs on the lower end of HF can be an expensive overkill.

A properly constructed UHF connector is rated for 500 Volts peak working voltage. At 50 Ohms, this is 5 kW. The dielectric breakdown voltage is 1500 VDC.

Working voltage rating for the N connector is 1500 Volts. The dielectric breakdown voltage is 2500 VDC. For corrugated types the voltages drop to 707 RMS and 2000 VDC respectively.

[Steve - WB3HUZ]

True dat. The center pin of a UHF connector is larger than an N. Think about that.   

UHF connectors are specified up to 165 degrees C. Most ham grade coax would have problems well below this temperature. IIRC a good UHF connector has loss on the order of 0.01 dB or less, at least at HF. At 1500 Watts, that's about 3 watts of loss (if I did the math correctly). Anyone want to calculate the temperature rise for this?

Also remember for coax (and coax connectors), maximum power handling capability occurs at 30 Ohms. Guess what the impedance of a UHF connector is?  30 Ohms.

[W2VW]

Also remember for coax (and coax connectors), maximum power handling capability occurs at 30 Ohms. Guess what the impedance of a UHF connector is?  30 Ohms.

I have some special PL-259s with carbon dielectric. Those have variable impedance depending of wheather it's producing a whisp of smoke or a steady stream.
They started life as the brown ones.

[Steve - WB3HUZ]

Those are especially good in damp environments.

[WA1GFZ]

Yup 30 ohms all you 2 meter guys.

[Steve - WB3HUZ]

Doesn't matter. The connector represents less than 0.01 wavelength on 2 meters or about 4 degrees. Plot a 30 Ohm line 4 degrees long on a Smith chart and see how much you deviate from 50 Ohms. The SWR will be about 1.1:1. Hardly worth worrying about.

Yup 30 ohms all you 2 meter guys.

[Todd, KA1KAQ]

Sheesh, everyone who's anyone in the world of high power knows the only connectors to use are BNC. They're at their best with RG-58/U.

[WA1GFZ]

Yea they are impressive at -130dBM

[Steve - WB3HUZ]

For quick on and off (like the BNC), try the GR874. I don't know what power they can handle but they are larger. Don't have to worry about male and female too. Bye bye barrels and adapters.

[WA1GFZ]

I bet they handle plenty of power but I have never seen one on a cable, only adapters.

[W2VW]

Sheesh, everyone who's anyone in the world of high power knows the only connectors to use are BNC. They're at their best with RG-58/U.

BNC and type N are close enough that you can jam a type N male into a BNC female jack.