I take back all the bad stuff I've ever said about EE Times. Sorry.

On Thu, 13 Jul 2006 16:43:11 -0700, John Larkin Gave us:

SMA is where it's at. Positive connections, high frequency capacity. Doesn't merely pull off. Doesn't count on some spring tension for retention.

But they are such a pita to mate and unmate, and you can't put them very close together, and the right-angle pcb versions are all taller than everything else. But below about 40 picoseconds, they are electrically mighty fine, virtually invisible on a 20 GHz TDR.

I really like MCXs. They snap in pretty solid, are small and rugged, good to at least 6 GHz, and there are tons of cables cheap on ebay.

John

Hello John,

But it did feel manly to work with a wrench when repairing the impedance analyzer a few weeks ago. Like working on that old T-Bird (not that I have one...).

On Sat, 15 Jul 2006 11:28:51 -0700, John Larkin Gave us:

Are your products meant to be fussed with so incessantly that it is a PITA, or is it just a PITA to develop using them because engineers in a lab DO mate and un-mate so often !!! (errrrr.... sick!!!)

The semi-rigid cable and a straight connector can be hand shaped to a bend radius that yields a lower exit height than any Right angle connector out there... almost. A lot cheaper too!

mighty fine... I like that.

That limit keeps them out of my stuff.

umm... how *big* a wrench?

Yeah, sometimes I feel like Real Men should be out dynamiting dams or cutting down trees or something. Then a plague of environmentalists shows up, and it gets cold and wet for a few months, and I am content.

John

On Sat, 15 Jul 2006 19:17:29 GMT, Joerg Gave us:

So you're the guy that over torques all the connectors and causes failure in the field!

Sick? There are some awfully cute female engineers around these days. Mostly software engineers, of course, but great for embedded applications, and most of them do work well close together.

Actually, I prefer to do my mating at home, and avoid frequent un-mating because of the expense. But yeah, when you're developing a board, or when the test folks are doing production test, all that SMA screwing around gets tedious. I put "connector savers" on all my sampling scope inputs, because SMA/3.5 threads will wear out if they're used a lot or hook up with a bad mate.

John

Hello John,

Well, I have to confess I used the "SMT version" of an adjustable wrench. Main reason was that it has a hole in the back where I could connect a spring scale so I don't over-torque after the repairs.

Oh, we do that on occasion (not the dams, of course). Man out here has got to have a chain saw, a truck and at least sometimes cut and split his own firewood. Or gitten his farwood, as some would say. Ideally the truck should be all decked out with horns, big mud flaps (you know the kind...), monster tires and some impressive arials. And a Cummins turbo.

Or is it "in-bedded" applications ?:-)

Oh, Lord, I apologize for that ;-)

...Jim Thompson

Hello Roy,

Well, hold the tomatoes but that's the reason why I used a little, ahem, adjustable wrench. They have a hole in back so I could pull via spring scales and not torque them too much.

On Sat, 15 Jul 2006 13:28:36 -0700, Jim Thompson Gave us:

He doesn't care. All that matters is what you say as you are on your way out.

On Sat, 15 Jul 2006 20:29:32 GMT, Joerg Gave us:

That's funny. A good, MORE accurate torque wrench isn't all that expensive, and if one IS working with SMA, it seems hardly cost prohibitive to have the right tool. Doing hundreds of them a day would make the spring measurement method more than a little labor and time intensive as well.

Hello Roy,

Sure. However, I generally do not work with SMA. The only time that happens is when I have to repair lab equipment like in the case I mentioned.

In an analyzer such as the HP4191A you'd need a really small torque wrench since the stuff is so close together. Often you have to take a few rigid coaxes out just to gain access to the one you really want.

On Sat, 15 Jul 2006 22:49:07 GMT, Joerg Gave us:

I routinely work with banks of 16 SMAs inline at a spacing of just over 1.25 times their diameter.

There are two tricks (at least two). Grind the sides down on the open ended wrench such that it will still function, yet provide maximal clearance (one or two mm wall thickness is typical). The canted head angle tool makers use is essential for six sided fastening.

The second is to take a closed end of a wrench (box end, 12 point), cut the head of it off, and grind the resulting "ring" down for thin wall thickness, and then cut a slot in it to gain access to an attached cable's hex nut, and then weld the resulting piece onto the end of a standard screwdriver (vertically) that has a hex shaft at the top so that the torque wrench can still be applied. This yields a fine close access tool. One can even find screw drivers that have the same size hex drive shaft that the SMA connector uses.

Also, one should always also incorporate the use of the small quarter inch wrench to keep the center nut from turning, thereby keeping any and all such stresses out of any particular coax that one happens to be using. This is important.

Kowabunga, I think I *will* use MCXs on this gadget I'm doing now.

John

On Sat, 15 Jul 2006 20:27:09 GMT, Joerg wrote in Msg.

And a gun rack.

robert