the difference between straight and reverse polarity antennas

so the connector fittings match up.

It's not the polarity of the antenna. It's the polarity of the connector. The FCC demanded that manufacturers use hard to find connectors for wireless so that the antennas could not be changed. When the rule was written, reverse polarity connectors were almost impossible to find. Now, they're as common as the normal polarity.

For the details, see the photos at:

normal connector and the reverse polarity connectors have center pins with a different polarity (pin or socket).

Forgive my novice questions, but if I'm connecting a miniPCI wireless card with a u.fl connector to an antenna with an N type connector through a couple of pigtails do I simply have to make sure I have connectivity from the centre pin of the u.fl on the wireless card to the centre pin on the antenna or is it more complicated than that.

My other question is would it still work if the centre pin of the u.fl connected to the outer part of the antenna?

Thanks in advance.

Keep the center path straight. I'm not sure how you could even swap them with coax.

I am at a loss why you would have a couple of pig tails. I go from the card to a N female. Use a N female on your antenna. Whatever cable you need should be N male to N male, and low loss. Because the N connector is large, you can easily mate it to low loss cables, most likely RG-8. The pigtail acts as a strain relief for the wifi card. you need to insure the N male to N male cable doesn't pull on the pigtail.

John Stubb> Forgive my novice questions, but if I'm connecting a miniPCI wireless

You really only need one pigtail. If you're going from u-FL to TNC and then to N connector, you can probably do batter with the pigtail and a TNC to N adapter instead of a second pigtail.

Well, the absolute minimum for proper operation is a connection between the u-FL connector on the card and the antenna. You have to have continuity of both the center conductor and the outer shield. Also check the cables for an open circuit between the center and the shield. Do these continuity tests with just the cables, as some anteannas and a few cards have DC shorts to ground. Al

No. The RF must remain inside the shield or it will radiate (leak) away.

Many thanks for everyone's replies.

The reason I am asking these questions is that I have a problem.

I have built several systems using a WRAP board inside a NEMA case.

See

The reason that two pigtails are required is that 1 pigtail goes from u.fl to N TYPE [with O ring seal] onto the NEMA case which seals the case.

You can then either connect an antenna directly or run another pigtail.

The system I built for testing just had a small 7dBi antenna attached directly.

It just works. I scan for SSID's using wicontrol -i ath0 -LL I get about

15 networks in the office complex where I work. It works, week in week out, not even the hint of a problem.

The OS is a cut down version of freeBSD running m0n0wall off a CF card.

The wireless card uses an atheros chipset.

The systems which have been deployed [not by me, or built by me] should be identical.

The problem is that when they scan for SSID's sometimes it doesn't show anything at all when their laptops see up to 25 networks. Reboots don't help. When another system [not mine] is connected it works fine. A day later my system works again.

Done the obvious, reseated and changed network cards. Complete new systems. Spoke to manufacturer's etc.

The only difference is the number of pigtails and antenna.

They had several other pigtails installed 'professionally' using LMR800. Up to 80ft long.

The cables I'm told have been swept for VSWR using some kind of meter.

The antenna's used have been 4ft to 8ft marine antenna's from

and They tell me this all works with another system. However I cannot check any of this as the systems are on the other side of the world, but assuming they have done everything correct my thoughts are

The wireless card could get hot and fail to work inside the NEMA case. This would explain why it works again after a while.

The N TYPE connector attaches to the NEMA case. The case is bolted to a steel boat. This means I think that the outer 'shield'[don't know what it's called] of the antenna/N TYPE/u.fl is attached to the whole boat.

Could this affect the antenna or wireless card?

Could something else in the boat zap the wireless card temporarily?

Should the 'shield' attach to ground at all?or even to the case?

My case in my office just stands on a wooden bookcase.

There is something about wicontrol or freeBSD I don't know like it fails to show any networks if there are more than 25.

Note that the only way I can get no networks is to disconnect the u.fl connector, even with just the pigtail I can see a couple of networks.

Any experience of similar issues or wacky ideas gratefully received...

"John Stubbings" hath wroth:

Thanks for the complete explanation. I have a guess which points to a problem with the u.FL connector. There are two connector types which look very much like a u.FL connector but are not. They are easily distinguished by looking down into the receptacle (on the miniPCI wireless card). One has a sloping side inside the connector. The other has a flat bottom around the center pin. Mixing connector types is a great way to destroy the plug.

Flat bottom (real u.FL)

Tapered sides: more later

Thanks for the reply Jeff

We have used two different wireless cards.

The first definitely has a u.FL connector. It is 802.11a/b/g.

It works fine for me and pretty much matches what my laptop sees when I scan for SSID's.

The second card was this one

This has both a MMCX and a u.FL connector.

I used the MMCX connector on this card on my test kit.

I had no issues except the card with the MMCX consistently saw two less networks which was surprising considering the card cost three times the price of the other card.

I put this down to coverage or loss in the MMCX connector however I didn't test it against the u.FL connector on the same card. Wish I had now, but at the time I was focussed on cutting out the u.FL connector.

The other guys did and they reckoned they saw twice as many networks using the u.FL connector over the MMCX.

Until it saw nothing again that is.