Saturday, August 29, 2015

"Barely Room To Swing A Wrench!"

     Or, "What I did Thursday."
      And on a tiny scale -- those are SMA connectors.  That's a 5/16" wrench.  And the the whole mess is overhead.  I had to replace the N-to-SMA bulkhead adaptor.  The other side is even worse to get to -- and that's where the nut that holds the thing is located.

     As for the nut who decided it was okay to build this widget in a box this small, I don't know.  I hope he has either learned better or retired.


The Neon Madman said...

No, he didn't. He went to work for General Motors instead.

rickn8or said...

Looks like plenty of swing-flip-swing-flip there. And a crow's foot wrench would be no help at all.

Did you have to scrunch down or stand tippy-toe to reach it?

Makes my arms and shoulders hurt just looking.

Raz Raxxaffian said...

In my years of retrofitting these Chinese-puzzle subassemblies, I always wondered where they found people small enough to put them together.

Pixies or child labor maybe?

Rey Brandt said...

Memories of a long since retired(like the sailor that worked on it)weapons system. I remember your pain.

Roberta X said...

It was no fun -- tiptoe, on a plastic chair of uncertain strength. Got it done.

I gave up trying to get an N connection on the end of foam 1/2" Heliax (a semi-rigid concentric line) lined up with the other side. The other four co-axes cross over within a foot of the box and it's all but impossible to bring the center one in square. Installed a low-loss 90-degree adaptor pointed down, threaded on the Heliax jumper and ran it over to the 1-5/8" line three feet away, where there was room to get everything aligned.

rickn8or said...

"Improvise, adapt and overcome."

markm said...

Sometimes the wire lengths must be short, so it all has to be crowded into a small box. I have worked on aircraft radar systems where six inches of coax would lose a few dB. More often, either things were added after the original simple, uncrowded circuit was built, or the smaller and cheaper box was specified by an engineer that neither has to build his designs, nor (since it's a one-off, I presume) has to listen to feedback from the factory. Given that Roberta added three feet to one wire and it still worked, I don't think it was the first reason.

But you will also see messes like this in mass-produced items. E.g., in a Chevette I once owned, the recommended way of changing the starter was to remove the master brake cylinder first. There actually was a lot of room around the starter, much more than on most cars. It was easy to unscrew it without removing anything else, but then none of the gaps between the engine, firewall, suspension, and master brake cylinder were quite big enough to fit the starter through. But no way was I going to remove any of those items. I'm an electronics technician and an EE, not an auto mechanic. So I bought a rebuild kit for the starter and repaired it in place.

In cases like that, the engineers did have to think about assembly, but not about maintenance. So the starter went in before the master brake cylinder, and it was all easy - provided that taking it apart again was someone else's problem. It's possible that this electrical box was designed so if you started in the correct corner and put everything together in the prescribed order, nothing was too difficult to reach. But it's unreasonable to expect the tech to remove half of the wires for better access to the one bad jack, so it's a poor design.

RandyGC said...

A line from a Heinlein novel goes approximately: "I never knew an engineer that didn't stop when he had a pretty picture and think that someone would have to actually use the thing".

Roberta X said...

Markm: I didn't add three feet -- I installed a jumper shorter than the one that was in there, and replaced a sharp bend at one end with an adaptor at the other.

This is 2 GHz. Extra coax does matter. We use the good stuff, but you still don;t want any more length than necessary.

rickn8or said...

Yes, markm, in my years of working on modified Navy airplanes, I've more than once said ugly things about an engineer's mother.

Eric Wilner said...

I sometimes get to watch other people dealing with aircraft radar modules that are packed with boards, tubes, wires, waveguides, semirigid coax, and flexible coax... all packed as tightly as possible, and sometimes more so, to fit the customer-specified envelope. Fun!
My role in this is designing the digital controller boards that go in the boxes with all the RF and power stuff. I've learned to include as much diagnostic capability as I can get away with, to reduce the need to disembowel these units and trace signals when something's not working. Still, the guys who do integration usually end up disassembling and reassembling the first unit of each new model several times.
I still haven't persuaded the Power guys (or their management) to include diagnostic capability in their subsystems, so the controller could read the status over RS-485 or CANbus or something, instead of a tech having to try to figure out what's going on inside an epoxy brick.

0007 said...

I still have my kit with the tools and plastic totes full of OSM and OSSM connectors for doing stuff to the MK-92FCS and other systems. About half the tools were "homemade" from other tools and the RF connectors and adapters cost ~$40-50 apiece back then - early '80's/'90's. I figured out one time that there was probably about 3 grand worth of parts in one of those totes between adapters and connectors.
Had to declare the value of the tool box(20X16X42) once when we went to Taiwan. $4K and it got held in customs for a couple of weeks while "they" debated over who was going to pay the duty, heh, heh, heh.