Towertalk
[Top] [All Lists]

[TowerTalk] Connectors, was DX Engineering

To: towertalk@contesting.com
Subject: [TowerTalk] Connectors, was DX Engineering
From: "Roger (K8RI) on TT" <K8RI-on-TowerTalk@tm.net>
Date: Thu, 15 Oct 2015 05:18:56 -0400
List-post: <towertalk@contesting.com">mailto:towertalk@contesting.com>
Regardless of connector make, be it generic, ot Amphenol, I have run into bad and good. If I ask for good reliable connectors from the major suppliers, I get them. I worked with a few suppliers to find a good UHF connector for LMR-600 and found a few good looking ones that were clinkers. "I think" Amphenol's QA is good. They use good materials and the designs are sound, but I've had a few with missing threads, or half a center pin. It's rare with Amphenol, but as the saying goes, "Shit happens". 100% QA just isn't economically viable, but automation is improving the checking.

I retired as a computer systems project manager. A lot of the laboratory instruments were automated, so test results went directly into the database. The ops just put what ever in pushed a button, flipped a switch, or stepped on a pedal/switch. They did little of the actual testing. There were still many tests done by lab personnel. Electronic companies are moving to the same approach, but it's not cheap. The connectors can be run through a device that can measure virtually all sizes, and electrically test the connector, but the number of connectors may be per second, rather seconds per connector. Problems often come when starting or stopping production lines, even on these very sophisticated lines. Who does what? I have no idea. Eventually the price of labor and production volume drives a process from a cottage industry to automation. You can only pay so much before automation, or out sourcing becomes your only options. OTOH outsourcing will eventually lead back to automation.

Another chink in the armor is domestic companies often out source parts and only test the lots by random selection. Many companies have reputations to uphold, but they still have to remain competitive by price and quality in what is now a truly global market.

I often see remarks about the cheap connectors at ham swaps, but take a close look at cables and connectors from big name companies with big reputations. I've found the very same, cheap thermo plastic miniature DIN connectors for some well known and expensive equipment. They can machine solder those connectors on in seconds with no problems, but those connectors were never meant for hand soldering 8 conductors in a spot no more than a 1/4 or 5/16ths inch across. An action that is now well beyond these old eyes and shaky hands <:-)

 End of soap box  <:-)

73

Roger (K8RI)


On 10/15/2015 2:20 AM, Thomas Noel wrote:
John,

There was nothing wrong with my technique.

Since it hit the TowerTalk listserve again yesterday, I was nudged to find an 
answer. I cut another of the DXE-8XDX006 jumpers in half, prepped the cable as 
per their instructions. The center conductor measures 0.055 inches just as it 
should. Nothing wrong there.

I then used machinists’ drill bits to measure the center conductor hole in the Amphenol 182115-10 
connectors shipped to me. The center hole will not pass a #55 drill bit which measure .052 inches. It 
should be able to pass a 1/16” bit (.0625”) as the spec calls for .063”.

So the problem is a batch of the Amphenol connectors which appear to be out of spec. 
Interestingly, I bought another batch from Dave’s Hobby Shop on EBay, and the 
measurements are the same. The connectors are not marked, and could be a cheap 
knock-off.

Interesting learning experience. I don’t think this was something within the 
control of DXE.

Let it die.

Thomas Noel
KF7RSF


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
<Prev in Thread] Current Thread [Next in Thread>