BNC pricing

You tickled me into measuring some connectors to see how good (or bad) they really are. I fired up my old S-52 -- S-6 TDR setup and tested various combinations of SMA, N, BNC, SMC and LEMO adapters and terminators.

The verdict: SMA is pretty damn good, BNC, N and SMC are quite honourable, although you have to look out who made them, and LEMO-00 is atrocious.

I threw together a little web page with the TDR plots. .

Jeroen Belleman

Huh? Those don't look like type N.

Den onsdag den 10. august 2016 kl. 02.33.11 UTC+2 skrev Jeff Liebermann:

they are soldered on both sides of the PCB

and you'd have to bend the end plates first

but of course it won't handle a 200 pound using a string of five adapters as a cheater bar

-Lasse

Sure, but if they're soldered equally badly on the other side of the PCB, and the nut securing the SMA to the front panel becomes loose, the solder connections might fall apart. Anyway, it's about what I would expect trying to solder to a connector that's attached to a rather large aluminum heat sink (front panel). The heat sink sucks away all the heat from the soldering iron resulting in a rather poor solder connection.

You also seem to have a gap between the back of the connector and the PCB. If this were an RF product using a microstrip, the gap would turn the exposed length of the center pin into an inductor. The gap also weakens the connection. The connector should be flush against the PCB.

I'm not sure, but it seems that you may have the lockwasher installed between the connector flange the back of the front panel. It really belongs between the front panel and the securing nut, to prevent the nut from vibrating loose. However, fixing that will change the PCB to front panel spacing and might not be possible (or worthwhile).

If the adapters are anything like the junk adapter I've been buying on eBay, they'll survive a few fractional Newton-meters before breaking in half. Instead, I use short jumper cables (pigtails) for adapters in mechanically sensitive areas.

Your choice of SMA connector is good. I use similar types for RF boards. However, I suggest you do something about the mounting and soldering.

Try what I described, sticking an N male into a BNC female. Works fine--nice tight shield connection.

Cheers

Phil Hobbs

I just tried this. Wow, I didn't know that! In fact, the fit is so tight I wonder if it was intentional.

Thanks.

Remember the old joke about the two hikers who ran into a grizzly bear:

One dropped his pack, whipped out his running shoes and started putting them on. His friend said, "What are you doing? You can't outrun a bear!"

"I don't have to outrun the bear. All I have to outrun is you." ;)

If the connector and mounting plate are stronger than the adapters, you're golden. Easiest with Type N, of course, but a good metal BNC with a bulkhead mount is close enough.

Cheers

Phil Hobbs

Lots uglier things than that happen in your average research lab. ;)

Cheers

Phil Hobbs

Then there are Jameco variants: 2077618 @$1.25 ea, 1952821 @$0.65ea,

"moving any equipment"??? NUTS! I will move the damn water cooler any time i see fit and you can jump in the lake.

What happens if you only give them 365 days notice? Or even just tell them that you did it after the fact?

My customer typically orders boards and wants them the next day and I haven't made them yet. So every couple of days they send me an email saying "please keep expediting". I do what I can but otherwise ignore the excess communications. The last order they want units by October! So no rushed schedule, but they send me an "order action report" to fill out and return. I ignore them.

Heck, this last order took a bit longer to give them an acknowledgement of delivery date because I wanted to save some bucks with a larger than needed PWB order. It took me an extra day or two to get all the numbers and commitments together. I got an email on the third day saying the supplier has 48 hours to confirm. I'm lucky I'm the only place they can get these boards.

Paul Neill designed the N connector and was one of the designers of the *NC connectors, probably intentional.

They were serious about it. I'm not sure how far down that went... trash cans?

They stole my design and did it themselves, so they were "potential" customers. Given the requirements they kept piling on, and the obnoxious people, I don't miss them. Truly rotten company culture.

I think it was Intel that imposed the Copy Exact requirement on their vendors. The idea is that they can debug one fab someplace in the first world, and clone it exactly in the third world, so they don't have to debug them all separately.

Makes a bit more sense at that level.

But the semi equipment business is neither nice nor comfortable.

Cheers

Phil Hobbs

Yes, but it gets pushed down the subcontractor chain. Mostly by "quality control" dweebs who love process control but don't actually care about quality. We make stuff for one big semi gear company that regularly sends in quality audit teams. I always ask them "Do you know what we make?" and they never do; I have to show them. They have no clue what our stuff is or does or what is the field failure rate. The failure rate is literally zero, so they gave us an "F" for quality. That leaves lots of room for "continuous improvement."

The Brat recently attended a big sorta-mandatory conference at the european headquarters. We made her a Vice President so they would be impressed with our enthusiasm and so I didn't have to go. The two themes were 1) System field reliability sucks, and it's all your fault and 2) You need to cut your prices in half.

Speaking of BNCs, this little issue has shut down some fab lines:

I think the early Silicon Valley folks, Schlockley and Grove and the boys, established the predatory culture. Many of the heavy-hitting people in the valley are truly repulsive.

The aerospace companies are the opposite. The people are mostly smart, friendly, and rational and buy stuff, or don't, without whining about price or process.

Yikes, that took talent. (It's also partly why I much prefer to present BNC females.) ;)

Good to know. Maybe they'd like to buy a bunch of photoreceivers. ;)

Cheers

Phil Hobbs

I've seen that before on radios, where someone programmed a robot to handle radio modules in an automatic test fixture. The robot had pressure and torque sensors to detect if too much force was being applied, but apparently they hadn't been set correctly, or not at all. The solution was to make a "floating" and self centering BNC receptacle, with less of a knife edge funnel entry and no keyway or ears to align. If only the ears are a problem, you could probably get away with just filing off the ears on the BNC receptacle. However, if alignment is the problem, then you'll need the "floating" connector and rounded funnel entry. I think such connectors are available commercially, but I couldn't find them with Google image search.

That preamp hangs on a pair of hermetic BNC females on a vacuum flange. A biggish cable exits the opposite end of the preamp box. That's it; no other mechanical support.

I'll ask. Send me a data sheet when it's available. Well, in a jet engine, low light levels may not exist. Fiber may be a possibility.

Holy Kdapt's liver on wheat toast!!! What did they try to mate those BNCs with - a rampaging brontosaurus?