[Top] [All Lists]

Re: [TowerTalk] Power limitation for N-Connectors

Subject: Re: [TowerTalk] Power limitation for N-Connectors
From: Ian White GM3SEK <>
Reply-to: Ian White GM3SEK <>
Date: Thu, 20 Apr 2006 08:26:21 +0100
List-post: <>
Ron Stordahl wrote:
>The consensus is that N connectors are fine at 1500 watts up to 30 MHz
>(and probably beyond but my use is 30 MHz or lower).  My problem with
>N-connectors in the past, and it's a distant past, is in assembly..maybe
>I had some bad ones.  And I am from the old to use
>solder.  But if the compression connection for the shield is good enough
>and long lasting..apparently it must be..then fine.
>The cable will be LDF5-50A which is foam.
>Is there a certain connector which has a 'captive center pin' as you
>My plan would be to use female connectors on the heliax, then male
>connectors on the more flexible jumpers at the ends, probably short
>length of 213.
>Am I on the right path?

Looks good to me.

The centre conductors in the Andrew cables are anchored against endways 
movement by the foam, and also their connectors are captive pin; so no 
problems there.

Likewise the RG213 jumpers have a captive centre conductor, so you can 
use any N connectors on those, provided they're assembled correctly. As 
Gary says, cutting lengths for the end of the cable are critical. At the 
factory, they use jigs to cut and assemble everything just right... but 
we don't. If we solder the centre pin too long on the plug, or not quite 
straight, it may split the mating pin on the socket. If we leave it too 
short, it won't mate reliably.

The 'improved' connectors are SO much better than the original MIL 
pattern! They have a pressure sleeve clamp with is waterproof, requires 
no solder and really does >>clamp<< the cable. Cutting couldn't be 
simpler - just one square cut, right down to the centre conductor, and 
no need to trim the jacket back or comb any braid. The captive centre 
pin is self-locating on the cut end. Solder that pin, tighten the clamp 
with a pair of wrenches and it's done.

As others have already said, there are several suppliers of 'improved' 
connectors in the N, BNC and UHF series. I'll leave it to US 
contributors to recommend specific sources. Try them, and I'll bet 
you'll never go back.

The captive pin will also anchor the floating centre conductor of 
9913-type cables, and greatly reduce the problems due to endways 
movement. In a long vertical run, you still need to support the hanging 
weight of the centre conductor (for example by using a 360deg loop as 
Steve recommends). If you don't, then over time a floating centre 
conductor will drop, and even a captive centre pin can't handle that 
much end force.

73 from Ian GM3SEK         'In Practice' columnist for RadCom (RSGB)

>Ian White GM3SEK wrote:
>> Steve Katz wrote:
>>> I'd say they're wrong.
>>> Type Ns flashover at >3000V at STP, and that's a lot of voltage.  The center
>>> pin can handle 6A continuously, probably 12A intermittently.  1500W is 5.48A
>>> in a 50 Ohm system.  1500W is 273.9V in a 50 Ohm system.
>>> I don't see the issue.  I've used type Ns at 1500W output power on 70cm for
>>> eme work, and never had one fail...
>> There is no problem with correctly assembled N connectors at these power
>> levels; but there is a risk of failure if the centre pin is not mated
>> correctly.
>> This can happen when N connectors with a floating centre pin are used
>> with  semi-airspaced cables whose centre conductor is free to move
>> inside the cable. A hanging length of cable, and/or some flexing in a
>> rotator loop, can sometimes pull the centre pin partly out from its
>> socket, causing high resistance (especially at UHF where skin depths are
>> smaller) and failures at high power.
>> The solution is always to use connectors that have a captive centre pin,
>> and to take care when installing cables that have a floating centre
>> conductor.




TowerTalk mailing list

<Prev in Thread] Current Thread [Next in Thread>