Low cost coax connectors

NEITHER item is the place to scrimp on a coax connector.

Putting a 50 cent RCA jack into such a nice product as that is an insult to the client!

It's about the disposable. No, the connector is not RCA in that one but it was meant as an example where you cosnstantly turn every cent around. If you don't the competition eats your lunch.

"MAKE a little cutter"? You're nuts. Try BUY a cutter already made for the purpose from MANY makers, such as Ideal, or Leviton.

If your client is shopping around, you should be charging more for your NRE, among other things.

I have seen clients use us for what they claim are going to be 10M orders, then get a design, and go shopping for someone else to build it in China, or the like. Funny, since we had two China facilities and had the cost of the item down as low as it could get. They don't understand that the US made protos are going to be higher in price, and until they commit to the high Mpiece orders, they are not going to see the lower per piece price.

As long as you refuse to actually name a characteristic, we're not.

"100 MHz" isn't a characteristic, it's a parameter.

John

I was just suggesting that you mention a few _actual_ products with these constraints, not proprietary info.

Now wait just a minute. You're not telling us those products have 20-year lifetimes and can't use 18-cent parts vs 5-cent parts? I really don't think so.

You're an idiot.

Like you saying "fine at 100MHz", idiot.

Here, hand hold retard! The ELECTRICAL characteristics of operation at

One example of which would be how much RFI your suggestion spews into the local environment compared to mine.

Ok, here are several others and this kind of stuff _is_ cost sensitive:

Among the things I've designed on this one is the analog front-end. The patient interface (not in picture) became rather small and is very thoroughly cost optimized. Yes, discrete circuits and all that, pretty much like what I usually do. Not that it was paramount here but by that time it was a natural habit to me. Marketing was rather happy about it because this part is often squished, banged, dropped, hit, crushed or scrubbed to death. So their customers can now enjoy quite reasonable replacement costs.

The high cost electric barrier parts that were in the previous model all got the boot, some semiconductor and other sales guys were mighty miffed, and as an added benefit it can now withstand a defibrillator hit without as much as a flinch. Impressed the heck out of the TUEV engineer who tested it all.

Joerg hath wroth:

Like everything else, using one of those removal tools takes some practice. There are also a few tricks involved. Once the connector is loose, the cable can be swung around the tool instead of the other way around. I'm somewhat clumsy, but I've seen an experience tech quickly move a termination panel full of cables to an adjacent rack without requiring any profanity.

That's why I carry a scissors and cutters in my war chest.

Lacing cord is a bit dated. I don't think I know too many people that know how to properly lace a wire bundle. These daze, it's done with Ty-Wraps. I've been to a few headends, in some manholes, and into some pedestals, and have never seen any real lacing cord. There's also a rule not to use ty-wraps as the soft foam dielectric on the RG6a/u tends to cold flow causing shorts and slight impedance bumps. It won't happen with quad shielded cable, but might with single foil shielded coax. Long parallel runs of coax cables is also a bad idea, especially when laced or ty-wrapped together. The shielding may be

Anyway, you didn't reveal previously that you wanted cheap connectors for internal (PCB?) connections. I've been assuming external. The easiest way is to forget about coax cables and connectors and use parallel wires. They don't radiate if they're balanced, properly impedance matched, and away from any other metal objects. If you have the space, use 300 ohm twin-lead. Tin the leads and shove them into a pair of PCB mounted probe test jacks.

Oh-oh. I just found a roll of waxed lacing cord. Brings back fond nightmares of rebuilding wire harnesses in mobile radios, telco hardware, and teletype machines. Yech.

I only cost NRE :-)

(there is no production here, clients take care of that)

My clients don't have a problem with that. The cost of the prototype doesn't matter. What matters is the BOM total including labor and machine times. Many have their own fab capabilities in China.

This is why the slotted socket works better than a device with a handle on it. For one thing, no one needs to apply so damned much torque on them that a handle is needed to remove them, and secondly, the socket spins freely around the cable with no handle to get in the way. The device the cable installer use is like this with a bail wire type torque arm for adding torque or loosening a tight fitting.

The right way to use the handle device is to make 180 to 270 degree turns, and pull back off the fitting, and reset the handle angle and repeat the operation. The F fitting only has about six full turns on it anyway.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Nicely put :-)

My last run-in with cable ties: Bought several (new!) linear supplies. Unpacked, two broken cable ties fell out. Nah, I don't trust them.

It has to be coax in this case. PCB placement is necessary because of cost (machine placed).

But this stuff lasts forever. There is a reason why it still has to be done that way in mission critical apps. One of the posters here (Dave M) got me a fresh spool from a Florida supplier after mine ran dry.

Absolutely not. Still in wide use. Particularly where a lot of hands are going to be in the area. Tie wraps are notorious for cutting up hands and catching on things.

But you can't put it into the DeWalt and push a button ;-)

Use flat style cat 6 cabling. Great stuff. eSATA cable is pretty nice if you want shielded runs.

Specifically, the waxed type is the type that is NOT mil spec.

BTW, both are still addressed in IPC-A-610 manual. So is wire wrapping, if you can believe that.

The they should be able to do their own costing. It would certainly be different than yours.

Rich Grise hath wroth:

That doesn't work. A bit more detail on the procedure:

1. Flush cut the coax cable.
2. Strip back only the outer jacket to expose the braided shield.
3. Tin the shield by either dipping in a solder pot, or using a metal solder brush. You gotta be really good and fast with the pot, but it produces the best looking soldering. Dipping only works well with PTFE dielectric.
4. Use a rotary blade stripper to trim back the tinned braid. Do not strip the center wire. The rotary stripping is fairly tricky as variations in solder buildup will cause corresponding variations cut depth. The trick is to tweak the die so that the guides do not rest on the blobs of solder that always accumulate at the end of the wire.
5. Strip the center wire with a thermal stripper.
6. Tin the center wire. There are some other tricks to this that I don't wanna get into.

If you try to strip back both the outer jacket, and the center conductor to the final lengths, the solder pot will make a huge mess of the exposed insulation and center wire. You could do it with the solder brush, but there's always a blob of solder that forms at the end of the shield, which has to be trimmed. That's difficult to do with the rotary trimmer after the insulation has been removed.

Not much different. I usually calculate considering Chinese production. Except when production is in the US or Europe where for reasons I have a hard time understanding even SMT placement is much more expensive so the whole design strategy has to be different. But I can adapt.